KI'A 560 6-76-016
TECHNICAL AND MIGROECONOM1C
ANALYSIS
TASK III - ARSENIC AND ITS COMPOINDS
APRIL 12, 1976
FINAL REPORT
I'.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF TOXIC SUBSTANCES
WASHINGTON , IXC. 20460
-------
-------
REVIEW NOTICE
This report has been reviewed by the Office of Toxic Slubstances, EPA,
and approved for publication. Approval does not signify chat the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
The role of arsenic Cand Its compounds) in the environment and in the
economy of the United States was studied, to evaluate the need for and the pro-
jected effect of controlling its production/ use, dissipation, and emission.
The occurrence, chemistry, and toxicology were reviewed; the prevalence of ar-
senic as an impurity in commercial raw materials, processes, and products was
systematically documented; the intentional commercial flow of arsenical pro-
ducts was quantified; the sources of pollution were identified and characterized;
and the health hazards were evaluated.
The intentional production and use of arsenic and its compounds is greatly
exceeded by the quantities unintentionally mobilized by industrial activities.
The arsenic currently in food and water presents no identifiable health hazard,
and the present controls on arsenical products, by a number of Government agencies,
appear adequate. Emissions to the air from high-temperature processes are large,
particulate collection devices appear largely inadequate, and the dangers pre-
sented are of serious concern.
11
-------
TABLE OF CONTENTS
Page
I INTRCDUCTION 1
Objectives of the Study 1
Previous Studies of Arsenic 2
Scope of this Study and Eeport 3
Constraints Upon this Study and Report ........ 5
II CONCLUSieNS 6
Societal Flow of Arsenic 6
Industrial Sources of Arsenic 8
Commercial Flow of White Arsenic and Its
Derivatives 11
Dangers to Man and the Environment 14
Control Alternatives Suitable for Reducing
Dangers 17
III OCCURRENCE AND CHEMISTRY OF ARSENIC 21
Natural Occurrence of Arsenic 21
Chemistry of Arsenic 26
Similarity to Phosphorus 28
Determination of Arsenic 28
Inorganic Compounds 29
Organic Compounds 34
Arsenic Adsorption and Coprecipitation 38
White Arsenic Refining 39
Chemistry of Arsenic in Fresh Water 41
Chemiscry of Arsenic in Soil 46
Arsenic Removal from Soils 48
Effects of Phosphorus 50
IV ARSENIC PRODUCTION AND USES 52
Data Collection and Use Trends 52
Pesticides 57
Wood Preservatives 65
Feed Additives 69
Nonferrous Arsenical Alloys 70
Glass 72
Electronics 73
Animal Dips 74
Arsenic in Dip 74
Note on Inorganic Arsenic Production 75
Environmental Emissions Resulting from Arsenic
Uses 75
Non-Arsenical Alternatives 78
111
-------
OF CONTENTS (Con^t)
V INDUSTRIAL SOURCES OF ARSENIC M3BTT.TZAT3ON 83
The Primary Zinc industry 84
The Primary Lead Industry 88
The Primary Copper Industry . 93
Other Primary Nbnferrous Metals 102
Arsenic in Nonferrous Metal Products 103
Phosphate Rock 109
Sludges from Municipal Sewage Treatment and
Municipal Water Treatment 112
Sulfur Deposits 113
Borax and Boric Acid 113
Iron Ore 116
Manganese Ores 119
Fossil Fuels 122
Geothennal Energy 127
VI ARSENIC TOXICOLOGY 128
Exposure Standards . . 129
Acute and Chronic Effects 130
Mode of Action 144
Oxidation State vs. Toxicity 146
Organic vs. Inorganic Arsenicals 147
VTI ASSESSMENT OF HEALTH HAZARD 150
Arsenic in the Air 152
Arsenic in Water 158
Arsenic in Food 162
Arsenic in Soil 163
VIII THE MARKET FOR ARSENIC 169
Domestic Arsenic Supply 169
World Arsenic Supply and Total U.S. Supply ..... 175
Demand for Arsenical Insecticides 177
Demand for Arsenical Desiccants and Defoliants . . . 180
Demand for Arsenical Herbicides for Weed Control . . 183
Demand for Arsenical Soil Sterilizers 184
Demand for Arsenical Wood Preservatives 185
Demand for Arsenical Feed Additives 186
Demand .for As203 in Glass Manufacture 187
Demand for as203 in Miscellaneous Uses 187
Summary of Demand for As2O3 187
-------
TABUS OF CONTENrS (Oon't)
IV roEtOTFICATION AND SCREENING OF CANDIDATE CONTROL
AtflERNATIVES 190
Existing Control Programs 190
Control Alternatives for Specific Emissions or
Dissipations .... 191
Control Alternative Aimed at the Ccnrnercial Use
of White Arsenic 198
Needs for Additional Research 199
X COSTS OF ALTERNATIVE REGULATIONS ....... 201
Bans Upon White Arsenic Use 201
Estimation of Foregone Benefits (Long-Run Costs) ... 207
Estimation of Disposal Costs for Excess As 0
(Long-Run) 208
Estimation of Short-Run Costs for As 0 Use Bans ... 208
Summary of the Costs for Banning While 3 Arsenic
Use 211
Costs of Controlling Industrial Arsenic Emissions
to the Atmosphere 211
Costs of Controlling Arsenic Emissions from Fossil
Fuel Combustion Stationary Sources .......... 214
Costs of Safe Disposal of Land-Destined Wastes .... 215
XI REFERENCES 218
-------
LIST OF TABLES
1. Arsenic Supply and Use 1968-1974 53
2. Arsenic Content of Zinc Concentrates 85
3. Primary Zinc Refined in the U.S., Metric
Tons/year 86
4. Arsenic Jn Wastewaters from Zinc Smelting 89
5. Arsenic Content of Lead Concentrates 90
6. Primary Lead Refined in the U.S., Metric
Tons/Year 92
7. Primary Copper in the U.S 94
8. Arsenic Content of Copper in Various Stages
of Refining 98
9. Other Primary Nbn-Ferrous Metal Ores Mines in
the U.S 104
10. Arsenic in Commercial Phosphate Rock 110
11. Production, Conversion, and Consumption of
Phosphates Ill
12. Manganese Ore Statistics 121
13. Normal Arsenic Content of Human Tissues and
Fluids 131
14. Arsenic Content of Various Foods 133
15. Toxicities of Various Organic and Inorganic
Arsenical Compounds 135
16. Maximum Permissible Levels of Arsenicals in Animal
Feeds and Maximum Permissible Levels of Arsenic in
Animal Tissue 149
17. Summary of Available Ambient Data for Arsenic .... 157
-------
LIST OF TflBT-py;
Page
18. Supply Statistics for Arsenic Trioxide ....... 179
19. Production of Arsenic and Copper by Country
in 1972 ...................... 178
20. Estimated U.S. Demand for White Arsenic ...... is 8
21. Economics of Arsenical Derivative Products ..... 210
22. Sunmary of Costs of Selected and Total Bans .... 212
-------
LIST OF FIGURES
Page
1. Calculated vs. Reported U.S. Production of
White Arsenic 172
2. Domestic Production of White Arsenic 173
3. World Supply of White Arsenic 176
4. Domestic Market for White Arsenic 181
5. Foregone Benefits 203
vail.
-------
ACKNOWLE1XSMENIS
This report was prepared by the staff of Versar Inc., Springfield,
Virginia. Mr. Robert P. Burruss, Jr. was the principal investigator and
Mr. Donald H. Sargent was the Program Manager for Versar. Mr. Robert Carton
was the Project Officer for the Environmental Protection Agency/ Office of
Toxic Substances.
Appreciation is extended to the many individuals who participated in
this effort. The organizations cooperating were:
Government Agencies
. Department of Agriculture, Agricultural Research
Service
. Environmental Protection Agency, Office of Pesticide
Programs
. Pood and Drug Administration
. Department of Interior, Bureau of Mines
. Department of Tabor, Occupational Safety and
Health Administration
Trade Associations
. American Wood-Preservers1 Association, Washington, D.C.
. Lead Industries Association
. National Agricultural Chemical Association, Washington,
D.C.
. National Cotton Council, Memphis, Term.
. National Paint and Coatings Association, Washington,
D.C.
-------
Private Cong>ani,fts
. Abbott Laboratories, North. Chicago, HI.
. American Smelting and Refining Co., New York, N.Y.
. Ansul Co., Marinette, Wise.
. Battelle Laboratories, Columbus, Ohio
. Buckman Laboratories, Inc., Memphis, Tenn.
. Chevron Chemical Co., Richmond, Cal.
. W.A. Cleary Corp., New Brunswick, N.J.
. Ccranercial Chemical Co., Memphis, Tenn.
. C.P. Chemicals, Inc., Sewaren, N.J.
. E.S.B., Inc., Philadelphia, Pa.
. Fleming Laboratories, Inc., Charlotte, N.C.
. Kbppers Co., Inc., Pittsburgh, Pa.
. Los Angeles Chemical Co., South Gate, Cal.
. Pennwalt Co., Bryan, Texas
. Salsbury Laboratories, Charles City, Iowa
. Thompson-Hayward Chemical Co., Kansas City, Kansas
. Ventron Corp., Beverly, Mass.
. Voluntary Purchasing Group, Bonham, Texas
-------
SECTION I
lOTKBUCTIQN
Objectives of the. Study
Efforts by various parts of the U.S. Environmental Protection Agency,
EPA contractors, other Government agencies and other workers in the field are
making increasingly apparent the present and potential dangers to man and the
environment from unrestricted production and use of certain toxic chemical sub-
stances. For many of these substances, there is ample evidence that the sub-
stances are in fact toxic. However, these substances have, in general, bene-
ficial uses and are of value to the private and public seccors of the U.S.
economy. Hence, the posture of the EPA with respect to these substances is
neither a blanket endorsement of current and projected practices as presenting
no real danger; nor is it, at the other extreme, a total and immediate ban of
the production and use of these substances. Realistically, for many of these
toxic substances, a careful assessment is required of the dangers and of the
options reasonably available for reducing the dangers.
This report is the partial result of a study specifically intended to
provide such objective data for several toxic chemical substances. The sub-
stances covered in this report are elemental arsenic and arsenic compounds.
The specific objectives of this study of arsenic (and its compounds) are:
1. To objectively and quantitatively evaluate the real dangers
(both present and projected to man and to the environment,
without the implementation of new and specific control
measures.
2. To make an accounting of how, where, and how much arsenic
is entering the environment in accessible (and possibly
dangerous) forms.
3* To identify control alternatives which may be techno-
logically and economically feasible, and to evaluate
the effectiveness of each of these control alternatives
in reducing the overall danger of arsenic to man and the
environment.
-------
4. To delineate the present and projected role of arsenic
(and its compounds) in the. U.S. economy, and to evaluate
the impact of each, of the control alternatives upon the
economy.
Previous st'VHgg of Arsenic
Much, has already been published on the various aspects of arsenic and the
environment. Various investigators over the years have separately reported on
the physical, chemical and biological properties of arsenic and its compounds;
on the natural abundance and polluted levels found in air, water, and food; on
the toxicology and estimated human dose rate ranges; and on the movement and
effects of arsenic in the ecosystem. Much less has been reported on the uninten-
tional mobilization of arsenic (as an impurity) by industry; on the flow of ar-
senic in society Ci.e. , in the economy) ; on the potential for substitutes in com-
mercial applications; on the identification of pollution sources and of abatement
practices; and on the costs of abatement and of use restrictions.
Study of these aspects of arsenic and the environment have been severely
hampered by the fact that no authoritative U.S. production or consumption data
have been published since 1959, when the American Smelting and Refining Company
(ASAROO) became the sole U.S. producer of white arsenic. The U.S. Bureau of Mines
has since then withheld these data to protect the proprietary interests of ASAROO.
Arsenic as a minor constituent of industrial wastewaters and of industrial
land-destined wastes has received much less attention than the heavy metals in the
many recent EPA studies on an industry-by-industry basis for effluent guidelines
development and for hazardous waste practices. A possible explanation is that
while atomic absorption is a rapid and economical analytical technique for the
determination of heavy metals, it requires more modification of technique and
matrix correction for arsenic determination so that alternate separate and specific
methods are usually preferred when the determination of arsenic is mandated.
There have been several recent publications which cover more than a narrow
aspect of the subject of arsenic and the environment. Among these are the
-2-
-------
publications of Sullivan, ^ Davis, Whitacre and Pearse, ' and Wood.
However, these were for the most part still addressed to only a portion of the
subject, and none were intended to be a comprehensive and deb-ailed encyclopedia
0
•for arsenic.
u
Scope of This Study and Report
In light of what already has been published and what hiis not, this study
and report attempts to provide a resource analysis for arsenic and its compounds
with as complete a breadth of coverage as was practicable within time and bud-
getary constraints. It was felt that an appreciation of all aspects of the com-
mercial and environmental flow of arsenic was needed to realistically assess any
dangers and to formulate and assess options for reducing the dangers.
This resource analysis of arsenic may be divided into four major subjects.
First is a detailed review of the occurrence and chemistry (Section III and of
the toxicology (Section VT) of arsenic and its compounds. These are the areas
which have received considerable attention from other investigators but which,
to our knowledge, have not been assembled before in a comprehensive fashion
suitable for achieving the objectives of this study. Included in Section IV
are natural occurrence, chemistry of the element, analytical determination and
coprecipitation, white arsenic refining, chemistry in fresh water, chemistry in
soils, removal from soils, plant uptake, biological transformation and the ef-
fects of phosphorus on arsenic transport. Section VT includes! exposure stan-
dards, acute and chronic effects, levels in foods and in tissues, modes of toxi-
cological action, oxidation state vs. toxicity, organic vs. inorganic arsenicals,
and the metabolism of arsenical animal feed additives.
The second major subject of this report systematically covers, for the first
time (to our knowledge), the many commercial raw materials, processes, and pro-
ducts in which arsenic and its compounds are involved as an inpurity or byproduct.
A stated intent of this effort (in Section V and in part of Section VTII) was to
quantify the commercial mobilization of arsenic. In a few cases, adequate data
were found to generate rather precise estimates. In many other cases, the esti-
mates were made to the best of our judgement despite a lack of consistent or
-3-
-------
verified data; the entry "no available data" purposefully was never used. Our
intent in going on record with, estimates was frankly to invite controversy,
hopefully to solicit constructive critieism of these estimates which should even-
tually lead to a set of data with a much-improved confidence level". Section V
includes, for each commercial occurrence of arsenic, the quantified fate of this
arsenic through our economy and especially into our environment. Section VIII
treats the potential for the commercial occurrences of arsenic becoming sources
for commercial arsenic.
The third major subject of this resource analysis of arsenic is the in-
tentional commercial flow of arsenic and its compounds (as opposed to the unin-
tentional flow of Section V). This subject, in Section IV and in Section VIII,
is usually based, for other commodities, upon comprehensive historical data
gathered and published by the Bureau of Mines and by the Bureau of the Census.
In the case of arsenic, however, such data has not been published for the past
16 years, in order to protect the interests of ASARCO, the sole U.S. producer
of white arsenic. Hence, quantifying the intentional" cotimercial flow 'of' arsenic
was an exercise in detective work and in estimation. As in the "commercial
mobilization" effort, estimates were always made; no entry was left blank or
given such a wide span which would have made the matrix useless for the project
objectives. We again invite criticism of our estimates. This analysis had one
less degree of freedom, however: the independently-derived estimates of the
total white arsenic supply and demand were made to balance each other. In addi-
tion to the quantification of the commercial flow of arsenic and its compounds,
Sections IV and VIII discuss the quantities released to the environment at each
step of processing, transfer, and use; the substitutes available in each use
category, the price of each arsenical product relative to its arsenical ingredient
and relative to its replacements, and the price elasticity of its demand (how
its use would vary with the price of its arsenical starting material).
The fourth major subject in this report is an assessment of the first three
subjects in relation to each other and in relation to the objectives of this pro-
ject. Section VII assesses the health hazard (both present and projected) from
arsenic and its compounds resulting from intentional and unintentional commercial
-4-
-------
;
mobilization, production, conversion, consumption, and disposal; without the
implementation of new and specific control measures. Section IX presents and
evaluates control alternatives for reducing the health hazard, and screens out
those which, are not reeded, not feasible, not effective, or too costly on an
a priori basis. Those control alternatives passing the screening process of
Section IX are analyzed in Section X for their estimated impact, upon the economy.
Constraints Upon This Study and Report
As alluded to before, this investigation proceeded without access to the
specific white arsenic production and consumption data as gathered by the Bureau
of Mines, the Bureau of the Census, or other Government agencies.
This investigation did not have the time, funds, or mandate to generate
any new experimental data.
This study, and the conclusions and recommendationE resulting from this
study, was intended to assess the role of arsenic in the U.S. economy and in the
general environmenc; i.e., the exposure of the general population to the overall
environment. It was not intended to substitute for other Governmental activities
in much more specific areas of interest. This study did not deal with arsenic
regulations for the vrork environment, as this is the province of the Occupational
Safety and Health Administration of the Department of Labor. This study did not
deal with arsenic regulations which are the province of the Fbod and Drug Adminis-
tration. This study only marginally touched upon the province of the Office of
Pesticide Programs of the Environment Protection Agency, mainly because the major
commercial use for arsenic is in pesticides; any appearance that this study was
for the purpose of influencing pesticide registration is purely unintentional.
-5-
-------
SECTION IX
CONCLUSIONS
Societal Flow of Arsenic
The table on the succeeding page is a quantitative summary of where arsenic
is found, produced, converted, used, and inadvertently altered. Of the arsenic
in the cotmercial flow in the United States, this summary table presents estimates
of the amounts dissipated in end products, of tfie amounts dissipated to land,
and of the amounts accessible to the environment via air, water, and land dis-
charges. The differentiation between the arsenic dissipated to land and the
arsenic in land discharges is that the former means a general distribution over
wide areas of the country, whereas the latter means a deposit of a waste material
in a bounded
-------
EU-H1W OF U.S. A'OENIC FIO-J, IKyiJlPATICWS, AMI ITIISSICHS
KCTRIC TOTS PiS Yii-Jl (1974 BASIS)
Conmercial CemnerciAl Dissipation in Dissipation Ateljomc! Katerbome Land'Dcstined
liicojpta Shintt..iits End Product!! to land Bnisciona Effluents Wastes
Printiry Zinc
In ?.n Concentratrii 525
LoS'.xiB in fjioltiny 190 0.4
Residues to Pb Smelters 210
IVitaii-.trl in Zn Products 5
Primary Ussd
In Pb CuTccntrates
In Residues from Zinc
Losses in Srelting
in Pb Products
PrJJrary Coiner
In Cu Co:iccntrabos
Losr.cs in Suiting
In Ac- id Loach Rssidues
In Collected i'luo Dusts
RctniJiod in Cu Products
Other Primary Non-rcrrous Metals
In Ore Concentrates
Losu&s in Sidting
Foe piyncnts, etc.
lead Alloys
Arrr.ii.ic lK7x>rted for Alloyinrj
teolc'.inis! firm Products
LocM-i in Processing
Contained in Products Sh.tpp>xl
Copper
Aisc'nit; Jrrfortod for Alloying
R?claJixd from Products
ConLriUiOd jr. Products Ship^tx!
In Cirauirxl Pl«s|>hnto llock
In I'Y'itilrscr
In AiJn.il Food and Other Products
In DFitL-rt'tnts
Krora Product Purification
and l.'a.stev.Tter Treatment
(tetnr 'jieatnent Sludges
Wastcw.itcr Traatrannt
Borax an! lioric Acid
In Rjfiiica Uorax
iii Eerie Acid Production
in Products
Manganese Ore
In Mi Ore Conccr.tralcs
lossns in Sao] ting
In Ferroalloys to Iron & Steel
In Batteries a.Td Other Products
Iron and Stool
In Ircn Ore
In Ferroalloys
jn Sti??
rJ in Utccl Predicts
in Cost lro;i ftnu^lrifS
in Cas^. Iron Product!;
Cool
In Coal
losses in Ocnirjstion
Petrolcun
In Pctroleuri
losses in Gcr&jstion
In Non-Fuel Products
Mute Arsenic Production & Conversion
From Ucxcstic Cu Flue Dusts
Irpartcd
Pesticide Production
l*xd Preservative Production
Feed Additive Production
Glass Additives
Misc. uses
DSe of Khite Arsenic 6 Derivatives
Pesticides
Hood Preservatives
Feed Additives
Glass Minufecture
Misc. Uses
TOBUS
850
210
35,000
150
580
855
107
75
555
22
4,000
54,060
2,160
2,450
120
8,300
8,550
12,790
1,400
407
1.805
320
103,000
(Net)
20
8,800 ,
2.JO
.4,800 32
8,300
50
1,105
182
30
130
293
32
65
2
110
3.9
2,160
18
400
35,500
3,300
1,080
36,000
12
12,790
1,400
407
1,805
320
1,400
1,595
320
23,980
10,490
407
57,170
10
12
20
650
1C 8
130
. 2
2,300
210
8,850
120
800
3,400
9,600
50
U
300
90
350
1,250
100
1,800
150
17,860
-------
Industrial Sources of Arsenic
*
1. The quantity of arsenic recovered for comnercial sale from
copper smelting is less than 25 percent of the arsenic
estimated to be in the copper ore concentrates. Most of the
remaining arsenic reports in slags, sludges, waste flue dusts,
and acid plant residues, all eventually disposed of on land.
Very little of the arsenic is discharged in wastewaters or
is retained in refined copper products. However, it is esti-
mated that 14 percent of the arsenic originally in the copper
ore concentrates is emitted to the atmosphere; this quantity
amounts to 4,800 metric tons per year and is more than all
other sources of airborne arsenic emissions put together.
2. The historical basis for the large quantities of arsenic emitted
to the air from copper smelters is related to the emissions
of sulfur oxides. The practice in past years was that sulfur
oxides capture (for sulfuric acid manufacture) was limited to
converter flue gases, which contain tsro-thirds of the sulfur
originally in the ore concentrate. The sulfur in the flue gases
from the'prior process steps of roasting and smelting was too
dilute for economical recovery. However, the arsenic partition
is exactly opposite: two-thirds of the arsenic is volatilized
in the roasting and smelting operations. When flue gases are
used to make sulfuric acid, cold-gas cleaning (wet scrubbing
as well as dry dust collection) assures arsenic removal. When
sulfur-bearing flue gases are emitted, dry dust collection
techniques such as cyclones, "balloon flues", electrostatic pre-
cipitators, and baghouses are only partially effective in cap-
turing arsenic (as explained below).
3. New emission standards for sulfur oxides from copper smelters,
aimed at 90 percent overall capture of sulfur, are resulting
in process changes such that considerably more of the arsenic
-8-
-------
Cand cadmium, lead, etc.l is being captured as well as
sulfur*
4. In the primary copper industry, in other non-ferrous
primary metals industries, and in coal combustion at
electric power generation stations, one-third to one-half
of the arsenic in flue gases escapes dry dust collection
devices despite nominally-high particulate collection ef-
ficiencies for these devices. Asr)_ does not condense
below 295°C and then only slowly (the particle nucxLeation
and growth processes are relatively slow). Conversely,
electrostatic precipitators and baghouses are routinely
kept above the dew'point of the flue gases, electrostatic
precipitators are run at elevated temperatures where the
gas resistivities are more favorable, and collected flue
dusts in the non-ferrous metals industries are commonly
recycled, providing more opportunities for arsenic loss.
5. Airborne emissions of As203 from all sources amount to
as much as the domestic commercial production of this
material.
6. Except for the arsenic in phosphate detergents, and some
small loss via wastewaters from copper smelters, the water-
borne effluents of arsenic are virtually zero. The stan-
dard trea'onent of wastewaters containing arsenic and otlier
metals is lime addition, with a flocculent such as ferric
chloride, and sedimentation. In the non-ferrous rnetals
industry, such treatment is required and justified for
the removal of heavy metals; the cost of this treatment
is not borne by the necessity to remove arsenic.
7. Much of the arsenic in commercial materials reports in land-
destined industrial wastes. Much of this arsenic is in a
relatively insoluble form, as complex arsenates in slags.
-9-
-------
However, a substantial portion ia subject to further
mobilization via wind-*dispersion of collected flue dusts,
and via leaching and runoff of sludges. Sulfide sludges
are particularly vulnerable to leaching.
8. Very little of the arsenic in non-ferrous metal ores and
concentrates is retained in the refined non-ferrous metal
products. The smelting and refining processes either
* <
vaporize the arsenic, remove it via a basic flux into a
slag, or leave it in electrolysis residues.
9. Arsenic occurs as a minor constituent in a great many
commercial crude materials at concentrations which are
highly variable but which are commonly two to four orders
of magnitude greater than the average crustal concentration
of 2 to 5 ppn. Two types of such enriched minerals are
prevalent: in sulfide ores such as copper, lead, zinc and
other non-ferrous metal ores; and in sedimentary deposits
where arsenic had been originally coprecipitated by hydrous
iron oxide. Significant quantities of arsenic are found in
such sedimentary materials as phosphate rock, borax, manganese
ore, and iron ore. The concentrations of arsenic in iron ore,
pig iron, and steel and cast iron products were estimated, but
were not extensively verified. Because of the huge commer-
cial quantities of ferrous metals, however, the quantities
of arsenic are correspondingly huge. It is estimated that
the arsenic in iron ore is more than that in all non-ferrous
ores, and more than the total arsenic in all other commercial
materials put together.
10. The arsenic in iron ore is retained through the blast furnace
process as stable and non-volatile iron arsenides. Basic
steelmaking processes remove the bulk of the arsenic as an
arsenate. The huge quantity of steelmaking slags containing
arsenic is used commercially for many purposes.
-10-
-------
11. While the arsenic concentration in coal is about the
average crustal concentration, the arsenic quantities
mobilized are large because of the magnitude of the coal
industry. This quantity is expected to grow dramatically.
Phosphate rock, is another growth industry where arsenic
is involved.
12. Searles Lake brines contain large quantities of arsenic
which conceivably could be recovered.
13. Three net/ technologies for energy production have important
arsenic implications. Early data on coal gasificaition in-
dicates that two-thirds of the arsenic is volatilized. Oil
shale may mobilize more arsenic by 1990 than is presently
mobilized by the copper and other non-ferrous metcil industries.
Geothermal energy development could also mobilize large
quantities of arsenic.
14. Metallic arsenic is an alloying element for lead eind copper
in several important uses. Much of these arsenical non-
ferrous alloys are recovered, however, in the secondary
metals industry; the arsenic in reclaimed metals is as much
as the quantity of new arsenic used for alloying. There are
significant losses, however, in the processing of reclaimed
metals.
Commercial Flow of White Arsenic and Its Derivatives
1. It is estimated that the U.S. production of white arsenic
is only 7 percent of the arsenic in all crude commercial
materials, and that the total quantity of arsenic potentially
available as a supply source should grow to be much larger
in the near future. Much of the present and future arsenic
resource should be recoverable by hydrometallurgical processes.
-11-
-------
2. The potential, supply of white arsenic, in the United
States and world-wide, far exceeds the'current or
potential demand. Arsenic and its derivatives are
consequently low-priced commodities.
3. The domestic production of white arsenic by the single
manufacturer increases as both white arsenic price and
domestic copper production increase, on a year-to-year
basis. Both factors are of approximately equal impor-
tance in affecting the production level. It is expected,
however, that several neV and important factors are
changing this relationship: the increase in copper ore
leaching, the process changes brought about by tighter
SO regulations upon copper smelters, and (most important)
A
the proposed changes in OSHA standards. Alternate sources
for arsenic supply also potentially exist.
4. Arsenical products compete directly with petrochemicals
in most use categories. The large price increases in
1974 and 1975 for arsenicals were likely the result of
large price increases for petrochemicals in these markets.
The demand for arsenicals in the future is to a large ex-
tent dependent upon the price and availability of its
petrochemical competitors.
5. The future for arsenical products lies to a great measure
upon actions to be taken by a number of Government agencies.
The Occupational Safety and Health Administration, the U.S.
Environmental Protection Agency's Office of Pesticide Programs,
and State agencies have the mechanisms for banning, severely
restricting, or otherwise drastically influencing the demand
for arsenicals or for their market competitors. The very
threat of yuch Government actions has inhibited commercial
activity on both the production and consumption sides.
-12-
-------
6. The 1974 demand for white arsenic was estimated to be
24,000 metric tons, broken down as follows:
Insecticides 23 percent
Herbicides (Waed Control) 24 percent
Dessicants and Defoliants 15 percent
Soil Sterilizers 18 percent
Waod Preservatives 6 percent
Animal Feed Additives 2 percent
Glass Additives 10 percent
Miscellaneous Uses 2 percent
While the general category of pesticides includes 86
percent of the total white arsenic demand, the above
breakdown indicates that no one specific use dominates
the market.
7. Alternate (organic) insecticides are generally available,
and in fact have taken over this market in which arsenicals
were once dominant. The two remaining important applications
for arsenical insecticides are for pest control on apples
and for mosquito control.
8. Alternate organic herbicides for weed control are generally
available. The two important markets for arsenical herbicides
are for weed control on cotton lands and on turf.
9. The demand for arsenical dessicants in cotton harvesting in
the Texas-Oklahoma region is growing, and there appear to
be no totally-adequate substitutes.
10. Arsenical soil sterilizers are being used less frequently.
Organic alternates exist.
11. Arsenical wood preservatives are increasing in demand, and
there does not appear to be an adequate alternate in many
applications.
-------
12. Arsenical feed additives are important in the poultry
industry; the antibiotic alternates are noicn. more-
expensive.
13. White arsenic consumption in the glass industry has
drastically decreased; its remaining uses are minor
and specialized.
Dangers to Man and the Environment
1. The greatest threat to human health is the inhalation
of airborne trioxide. The recent studies of airborne
arsenic in.the workroom, conducted relevant to the pro-
posed revisions in OSHA standards, have resulted in the
consensus that arsenic trioxide is a carcinogen, with
lung and lymph cancer mortality rates for exposed workers
6 to 7 times the expected rates.
2. The major sources of arsenic pollution of the air outside
of the workroom are the 40 to 50 primary non-ferrous metal
smelters, particularly copper smelters. At distances of
10 to 15 miles from smelters, levels of arseni^ in the
air exceed the newly-proposed standards for the workroom.
Dusts which have settled from the ai± near smelters contain
hundreds of ppm of arsenic. Within the context that the
areas influenced by smelter discharges represent only a
small proportion of the Nation and of its population, the
arsenic pollution of the air from smelters represents a
public health hazard apart from the workroom considerations
Of 03HA.
3. Other than airborne emissions from primary non-ferrous
smelters, important sources include secondary lead
smelters, the many coal-burning electrical power genera-
tion stations, the production plants using white arsenic
as a raw material, the emissions to the air from the use
-14-
-------
and application of pesticides, and the incineration of
cotton trash. Ml o£ these sources put together emit
less arsenic than copper smelters, but these sources are
much more dispersed in our population than the smelters.
4. Arsenic ingested via food, even in high concentrations
in some sea foods, does not present any health threat
yet identified. Although biomagnification of firsenic
occurs in the food chain, the organic forms of arsienic
in food are excreted within four days, with no identified
hazard to humans. Arsenical feed additives for poultry
and swine cause little if any accumulation of arsenic in
the tissues of these animals; the Food and Drug Adminis-
tration has set standards and monitors arsenic levels.
The largest hazard from arsenic via foods appears to be
inorganic arsenic on the surface of fruits and vegetables,
either as insecticide residues on apples and some other
fruits, or fallout from industrial and camercial point
sources of air pollution.
5. Arsenic in water constitutes no current threat to the
public health. Municipal water treatment plants eire
effective in reducing the arsenic content of raw water.
The arsenic in fresh waters (resulting from natural or
man-made erosion, from geothermal natural sources,, from
point sources of pollution, and from runoff from agricul-
tural or suburban lands) becomes either locked into highly
insoluble soil or sediment complexes where it is effectively
removed as an environmental hazard, or it moves to the
oceans. Very few public water supplies exceed the recom-
mended maximum arsenic standard of 10 ppb.
6. The inorganic pentavalent forms of arsenic are 10 to 60
times less toxic than the inorganic trivalent forms.
Moreover, organic compounds of arsenic are 10 to 100 times
less toxic than inorganic compounds.
-15-
-------
7. The. vise of arsenical pesticides and animal feed additives
results in large quantities dissipated to land. These.
quantities are augmented by the arsenic in phosphate ferti-
lizers, the fallout from sources of arsenic air pollution,
and the large quantities of steelmaking slag that are used
for various base and fill applications (although the ar-
senic in slag is likely fixed and insoluble as ferric
arsenate).
o
8. Of the mobile arsenic dissipated to the land, chemical and
bacterial actions serve to oxidize the arsenic over a period
of time to the pentavalent state. Much of the pentavalent
arsenic becomes bound as insoluble arsenates to iron oxide
c
and aluminum oxide sites in clays. Some, as in the case of
defoliated and dessicated cotton, is removed from the land
via crop harvesting. Some is washed from the soil into sur-
face waters, and some is leached and transported deeper into
the soil. There is evidence that some organic arsenic is
microbialJ.y changed to methylarsines, which volatilize from
the land (and are subsequently oxidized to As-Oj . As a
cumulative result of these mechanisms, there is data to
show the reduction with time of both total arsenic and avail-
able (soluble) arsenic in the soil after application of an
arsenical.
9. Caoodylic acid is more resistant to oxidation tlian the
sodium salts of methanearsonic acid (MSMA and DSMA). How-
ever, the microbially-aided oxidation of cacodylic acid is
enhanced in "adapted" soils. All of these organic arsenicals
are less toxic than inorganic arsenicals.
10. Competition of phosphorus with arsenic for available sites
in soil renders arsenic relatively more soluble. Arsenic
uptake by plants, and arsenic transport deeper into soil,
is enhanced by phosphate fertilizers.
-16-
-------
11. Since the largest.uses of arsenicals are for non-food
crops (cotton), for turf, and for other non-food applica-
tions, plant uptake is not a threat to human health. The
use of arsenical insecticides on apples and other fruits
has not resulted in arsenic levels which, present a hazard.
The tolerance of humans to organic arsenicals in foods,
in combination with the above factors, negates the potential
for a health hazard by arsenic in the food chain.
Control Alternatives Suitable for Reducing Dangers
Based upon the analyses in Sections III through VIII of this report, alter-
natives for controlling the emissions of arsenic and for reducing the hazards
to health were formulated and are presented in Section IX. Also included in
Section IX is an evaluation of these alternatives; several were screened out
and rejected because they were not needed, not feasible, not effective, or too
costly, on an a priori basis. The control alternatives passing this screening
process were then evaluated from a cost standpoint.
A summary of the control alternatives passing the screening process,
each with a concise statement of feasibility, effectiveness, and cost, is in-
cluded at the end of Section II.
-17-
-------
SUMMARY OF CONTROL ALTERNATIVES
Control Alternatives
Feasibility
Effectiveness
Cost
Requiring effective (99-1- %) removal
of As,03 from flue gases emitted to
the atmosphere from copper smelters,
other non-ferrous mstal shelters,
cotton trash incinerators, glass
plants, and other industrial sources
with As203 in high-temperature process
gas streams.
Technology of removal using
high-pressure-drop ventun
scrubbers is demonstrated in
gas-cleaning sections of by-
product sulfuric acid plants.
Should reduce As O emissions '
from 6,300 kkg/year by 99+ per
cent, except for gases lost via
leaks and spurious emissions.
Other hazardous constituents
would also be controlled.
Estimated capital cost
of $8.3 million, plus
$1.0 million/yr operating
cost. Total cost is about
$300 per kkg of As203 con-
trolled.
oo
Inquiring environmentally-adequate
land disposal of arsenic-containing
slags, sludges, and collected flue
dusts from industrial sources.
Arsenic-bearing wastes from primary
copper ir-dustry are 3/4 of all such
wastes.
Technology of approved land- .
fills, secured landfills, waste
treatment, encapsulation, etc.,
has been demonstrated and is in
practice.
Should protect 15,000 kkg/yr of
arsenic in industrial wastes9
from migrating into environment
via leaching, runoff, and wind
dispersion. Other hMmn^raig
constituents in thpse wastes
Would aim be <««'il |T>11«v^T
Estimated total cost for
copper industry is $2.1
million/yr; of which
$280,000/yr may be appor-
tioned for control of
arsenic. Cost is equiva-
lent to $22 per kkg of As
.controlled.
3an on the use of arsenical insecti-
cides (calcium and lead arsenates,
paris green).
Petrochemical alternatives
available, but relative health
hazard may be equal or greater,
and relative insecticide effec-
tiveness may be less.
Would prevent the dissipation
of 5,500 kkg/yr of As2O3 equiva-
lent; only the portion airborne
during spraying application is
hazardous.
Estimated costs would be
$3.7 million first year,
$3.4 million/yr next 4
years, $2.9 million/yr
thereafter ($680, $630,
and $530 per kkg of As2O3
diverted.
Ban on the use of arsenical desi-
ccants and defoliants (arsenic
acid).
Petrochemical alternatives do
not appear to be adequate sub-
stitutes for .Texas-Okla. cotton
use, and relative hazard may be.
equal or greater.
Would prevent the dissipation of
3,500 kkg/yr of As2O3 equivalent;
the portion airborne during
spraying and the portion emitted
via incineration of cotton wastes
are hazardous.
Estimated costs would be
$2.4 million first year,
$2.2 million/year next 4
years, $2.0 million/year
tnereafter ($680, $640,
and $570 per kkg of As20}
diverted)....
-------
Control Alternatives
Ban on the uce of arsenical
herbicides.for weed control (riSMA
and DSMA).
Feasibility
Petrochemical alternatv.ves are
generally available, but relative
hazard on cotton and turf nay be
equal or greater.
Effectiveness
Would prevent the dissipation of
5,800 kkg/yr of As2O3 equivalent;
the portion airborne during
spryaing and the portion emitted
via incineration of cotton wastes
are hazardous.
Cost
Estimated costs would be
$5.2 million first year,
$4.3 million/year next 4
years, $2.9 m:llion/year
thereafter ($890, $740,
and $500 per kkg of As203
diverted).
Ban on the use of arsenical soil
sterilizers (sodium arsenite).
Ban on the use of arsenical wood
preservatives (CCA & PCAP).
Petrochemical alternatives avail-
able, but use of arsenicals is
highly selective. Relative
hazards may be equal or greater.
Would prevent the dissipation of
4,200 kkg/yr of As2O3 equivalent;
this quantity has not been shown
to be hazardous.
Estimated costs would be
$2.5 million/year first 5
years, $2.3 million/year
thereafter ($600 and $560
per kkg of As2O3 diverted).
There do not appear to be ade-
quate alternatives for many
applications.
Would prevent the dissipation of
1,550 kkg/yr of As2
-------
Control Alternatives
Feasibility
Effectiveness
Cost
Ban on the use of As2O as an
additive for glass (except for
highly specialized infrared or
scientific glasses).
Substitutes are available for
oxidizing and fining.
Would prevent the dissipation
of 2,400 kkg/yr of As2O3; only
the portion emitted to the air
(280 kkg/fyr) during glass manu-
facture is hazardous.
Estimated costs would be
$1.1 mining/year for all
years after ban ($460 per
kkg of As O diverted).
?
Total ban on the use of white
arsenic and its derivatives
(except for highly specialized
and small-volume uses).
See feasibility of individual
use bans.
See effectiveness of individual
use bans. Would prevent the
riiggipaj-iom of 24,000 kkg/yr of
As203.
PaMmafrcrl cOStS would be
$20 mill-inn first year,
$16 million/yr next 4
years, $13 million/year
thereafter ($830, S665,
$550 per kkg of As.O.
diverted). * '
Requiring effective (99+ %) removal
of As^O3 from flue gases emitted
to the atmosphere from coal-burning
electric power generating stations
and other stationary sources.
Technology is similar as that
for industrial sources of
As203 emissions, but the con-
centration of As203 in coal-
burning flue gases is much
lower.
eliminate air-
of 650 kkg/year
Should
borne
of arsenic, or 860 kkg/year of
As2O3. Other hazardous constit-
uents would also be controlled.
Ext i mated costs are $335
million/year. If iatal is
constituents, the cost for
arsenic is estimated at
$39,000 per kkg of As2O3
controlled.
-------
0 SECTION III
OGCUF^ENCE AND CHEMESTPY OF ARSENIC
Natural Occurrence of Arsenic
The adjective most often used to describe the occurrence of arsenic is
ubiquitous. The average crustal abundance is about 5 ppm (5 mg/kg, 0.0005
percent); ' ' 'it is one of the less abundant elements (14th in abundance among
trace elements), about on the same order of average crustal abundance as tin.
(2)
Virgin soils usually contain only a few ppm of arsenic, ' but soils having
(4\
natural concentrations as high as 500 ppm have been reported. Ferguson and
Gavis list concentrations of arsenic for the following rocks:
igneous rock 1.8 to 2.0 ppm
shale 6.6 to 10.0
sediments 10.0
(deep sea)
sandstone and 1.5
limestone
The greatest concentrations of arsenic occur with ores of copper, lead,
cobalt, nickel, iron, and silver, either alone or with sulfur. Lead, copper,
and gold ores contain amounts of arsenic measured from trace amounts up to 5
percent.(1'2)
Three of the 15 copper smelters in the U.S. process ores having high
arsenic content. The ASARCO smelter in Tacoma, Washington, processes ore
containing 5.2 percent arsenic (52,000 ppm); the ASARCO plant at El Paso pro-
cesses ore having an arsenic content of 0.96 percent, and the Anaconda smelter
at Anaconda, Montana, processes ore containing 0.8 percent arsenic. The re-
maining copper smelters all process ore containing less than 0.2 percent
arsenic.(19)
The arsenic content of zinc, lead, and copper ores is discussed in the
section dealing with primary nonferrous metals. The arsenic content of zinc and
lead concentrates from five foreign sources (data for American Ores is not avail-
able) averages 565 ppm for zinc concentrates K50% Zn content) and 944 ppm for
-21-
-------
lead concentrates (65% Pb content). The arsenic content of copper concen-
trates have been measured at up to 16,000 ppm (Butte, Montana). * ' Unpro-
cessed copper ore from Butte, Montana, has been measured to contain as much as
(21 22)
1000 ppm and 3700 ppm arsenic. ' Thus arsenic is significantly concen-
trated above its average crustal abundance of 5 ppm in the ores of zinc, lead,
and especially copper.
Gold ores in Sweden contain 7 to 11 percent, and copper ore from the
now-depleted Boliden deposit in Sweden contained an average of 10.8-percent
arsenic - versus only 2 percent copper. According to Swain, "not all sulfide
ores contain arsenic, but wherever arsenic has been a source of ^trouble (e.g.,
pollution from smelting), sulfur has been present to aggravate it." 3he Boliden
ore body contained about 30 percent sulfur.
Over 150 .arsenic-bearing minerals have been' identified, of which the meet
common are the magnetic sulfides such as arsenopyrite (also called mispickel,
FeAs2'FeS2), loellingite (Fe,+ As. ), enargite {SCuS-As-SJ, realgar (AsS), and
orpiment (As.S,). Magnetic sulfide ores contain an average of 2000 ppm of
Veij
(1,3)
(3)
arsenic. Veins of native arsenic have also been found in a number of
localities.
In sea water, according to Schneider, the "normal" concentration of
arsenic is 0.003 mg/1, or 3 ppb. Lansche places the concentration at 20 ppb
and says that the arsenic exceeds the concentration of iron in sea water.
(2)
Sullivan cites 10 to 100 ppb as the arsenic concentration in seawater. Ferguson
and Gavis estimate the average concentration to be 2 ppb, "though measured values
,(5)
range from 0.15 to 6 ppb".(5)
Arsenic Content of Oceans
Concentration (ppb)
English Channel 2-4
Pacific Coastal Water 3-6
Northwest Pacific 0.15 - 2.5 (avg. 1.2)
Indian Ocean 1.3 - 2.2 (avg. 1.6)
Southwest Indian Ocean 1.4 - 5.0 (avg. 3.0)
-22-
-------
Probably the single greatest source of arsenic in the earth's crust and
in sediments and sedimentary rocks is the combined contributions of hot springs
(22)
and volcanic activity. According to Reay, hot springs in the Wairakei (New
Zealand) geothermal field "are litely to be an important source of arsenic",
because they are an important source of magnetic chlorine - which occurs in a
fairly constant ratio with arsenic throughout the Pacific region. Reay calculated
the natural output of arsenic in the Wairakei area to be on the order of 22 kkg/yr.
Also, he rioted that the bores for a geothermal power plant at Wairakei produced
190 kkg of arsenic in 1964, and "this can be expected to remain more or less
constant" .
Marine organisms tend to concentrate arsenic in their -tissues. In sea
water containing 0.05 to 5 ppb of arsenic, marine plants have been reported to
contain between 1 and 12 ppm of arsenic (dry weight) , while marine animals con-
tain concentrations of 0.1 to 50 ppm. Arsenic in shrimp and lobsters, probably
as trimethylarsine, has been measured as high as 200 ppm - a 100,000-fold increase
over the average sea water concentration of 2 ppb.
Arsenic occurrence is "very common in the freshwater of the western United
States" ; and in one part of the world, New Zealand, the naturally occurring arsenic
in freshwater is reportedly sufficient to be lethal to animals (44 mg/animal kg) .
Ferguson and Gavis report freshwater arsenic concentrations fear various rivers
and lakes throughout the world as follows:
Arsenic Content of Fresh Waters
Concentration
Lakes in Greece
Lakes in Japan
Tak«as in Wisconsin
Rivers and lakes in U.S.
Rivers in Sweden
Rivers in Japan
Elbe River, Germany
Columbia River, U.S.
1.1 - 54.5
0.16 - 1.9
2-56
10 - 1100
0.2 - 0.4
0.25 - 7.7 (weighted avg. 1.7)
20 - 25
avg. 1.6
-23-
-------
The United States Public Health Service has established a recommended
maximum concentration of 10 ppb and a maximum permissible concentration of 50 ppb
for arsenic in public drinking water; both of these limits are well below the
lowest reported concentration which resulted in chronic poisoning - 210 ppb.
Surveys of drinking water sources and supplies have been carried out over the
years in the United States. In 1943, 37 drinking water supplies were tested for
arsenic; the maximum concentration found was 8 ppb, and in 30 samples arsenic
was undetected (<2 ppb). In 1969, a survey of 969 water supplies found that 0.5
percent of them exceeded the 10 ppb Public Health Service recommended limit and
0.2 percent of them exceeded the 50 ppb upper limit. In two studies of fresh sur-
face waters in the United States in 1970 and 1971, arsenic was found in about 7
percent of 1500 samples from 150 rivers in one study, and in 21 percent of 727
samples from rivers and lakes in the other study. Although the limit of detection
in these studies was at the P.H.S. recommended limit for drinking water, 10 ppb,
most of the samples which had detectable arsenic were in the 10 to 20 ppb range.
According to Ferguson and Gavis, there have been many observations of high con-
centrations of arsenic in lakes and impoundments in the United States, and they
feel it is probable that arsenic concentrations in natural waters often approach
or exceed values thought to be safe for drinking water. ' A large portion of
arsenic in surface waters of the United States is probably from other than natural
sources; e.g., from arsenic in detergents, pesticidal runoff, and leachings from
excavations and mining operations.
Arsenic also occurs, along with other trace materials, in coal and petro-
leum as well as in mine tailings and in products made from phosphate rock, such
as fertilizers and detergents which are possible primary pathways of arsenic into
(7 8)
the Nation's fresh water supplies. ' Sullivan lists the arsenic content of coal
(2)
burned in the U.S. at 0.08 to 16 ppm. The National Inventory of Sources and
Emissions; Arsenic - 1968 gives a range of 1.18 to 9.95 ppm for domestic coal,
with an average of 5.44 ppm, on the order of the average crustal abundance.
According to Anderson , domestically-produced crude oil contains 0.007
to 0.61 ppm of arsenic, with an average concentration of 0.15 ppm; foreign crude
contains from 0.01 to 0.34 ppm with an average of 0.13 ppm; and residual oils (i.e.,
crude oils for electric power generators and for the heating of buildings) con-
-24-
-------
tains 0.1 to 0.2 ppm arsenic with average of 0.14 ppm. The National Inventory
of Sources and Emissions notes a group of 110 tests of domestic crude oil; in
97 of the tests arsenic was undetectable, but in 13 tests it ranged from 0.008
(9)
to 2.4 ppb, for an average concentration for all 110 samples of 0.042 ppm.
Oil from shale has been analyzed as containing 82 ppm of arsenic.
The arsenic concentration in phosphate rock mined in the United States
varies from values close to the average crustal abundance (about 5 ppm) up to
20 times this value. The arsenic content for commercial phosphate rocks has
been cited by various researchers:
Florida Florida Tennessee Vfestern
Reference Land Pebble Hard Rock Brown Rock Rock
(ppm arsenic)
(12) 3.5 - 22 1.5 - 11 5-56 4.5 - 105
(13) 7.5 - 37.5 3-9 15-30 7.5 - 112
Sauchelli reports the arsenic content of "a representative analysis
of 20-percent granulated superphosphate manufactured from Florida pebble rock
phosphate" as 14 ppm. The.Department of Agriculture reports the arsenic
content for 10 samples of industrial phosphoric acids as varying between 1.5 ppm
and 1200 ppm, with the majority being in the area of 25 ppnt, indicating that
phosphate processing does not tend to remove the arsenic carried in the ore.
Since arsenate is chemically similar to phosphate, it is not unreasonable
to think that arsenate might substitute for phosphate or at least to be fairly
concentrated in phosphate minerals. However, in Florida phosphate pebbles,
arsenic content is inversely proportional to phosphate content and directly
proportional to the iron content, indicating that the affinity of arsenic for
iron is tire predominating concentrating factor for arsenic in phosphate.
Domestic reserves of arsenic are estimated at 1.7 to 2.3 million kkg,
approximately 40 percent of the known world reserves. ' ' These values for
domestic reserves are principally a function of copper reserves. Since arsenic
is generally associated with magmatic deposits of complex base-metal ores, the
reserves are probably significantly greater than the amount available as a by-
product of copper production. '
-25-
-------
Chemistry of Arsenic
Of the toxic elements, arsenic is probably the most well known. Pure
elemental arsenic, however, is not very toxic, which is likely the result of its
being virtually insoluble in water or in the body fluids. In fact, elemental
arsenic is not readily attacked by water, alkaline solutions, or non-oxidizing
acids; hydrochloric acid will attack it only in the presence of an oxidizer.
Elemental arsenic is commonly referred to as a metal. Chemically, it is
a nonmetal or metalloid being classified in Group 5a of the periodic table, along
with nitrogen, phosphorus, antimony, and bismuth.
Properties of Arsenic
Atomic number 33
Atomic weight 74.9216
Melting point
at 1 atmos., sublimes at 613°C
at 28 atmos., melts at 817°C
Density at 20°C 5.72 g/cm3
Latent heat of fusion 88.5 cal/g
Latent heat of sublimation 102 cal/g
Specific heat at 20°C 0.082 cal/(g)(°C)
Lattice constants at 26°C a = 2.760A
b = 10.548A
Hardness (Mohs: scale) 3.5
There is only one stable arsenic isotope; therefore, the natural abundance
of As is 100 percent. The electron configuration is such that the five elec-
trons in the outer shell give rise to the three principal oxidation states which
are -3, +3, and +5.^
Elementary arsenic occurs in three allotropic modifications. They are
the yellow, the black, and the metallic or gray forms, the latter being the most
*
stable at room temperature. The electrical conductivity of the metallic form at
4
0°C is 2.56 x 10 mhos/cm, about half that of lead. The other allotropic modifi-
(23)
cations are listed as nonconductors.
-26-
-------
The yellow form of elemental arsenic can be produced by passing arsenic
vapor into cold carbon disulfide and cooling the solution to -70 °C. TMs yellow
form is an extremely volatile solid, subliming even from the heat of the hand.
Its density is 3.9 g/cm , and its molecular weight corresponds to that of tetra-
hedral As4 molecules. It is mstastable and transforms into the metallic form
even at low temperature; in sunlight, at room temperature, it changes virtually
(*\M OC\
instantaneously. '
The black modification of elemental arsenic is not as 'well characterized
as the other forms. It is obtained by the thermal decomposition of arsine, AsH.,.
3
The density of black arsenic is 4.7 g/cm . Its molecular configuration is not
definitely known, but it is probably tetrahedral.
Metallic arsenic forms hexagonal-rhombic crystals and cubic crystals. It
is stable in dry air, but exposure to humid air causes the surface to tarnish,
first to a bronze color then to black. The density of the :raetallic form is
the highest for the three allotropic modifications: 5.72 g/on . Metallic arsenic
is the common ccmmercially-available form, being the product of the reduction of
arsenic trioxide with coke according to
As,0,. + C"A*"As.. + 600
46 4
It can also be sublimed from arsenopyrite according to
When heated in air elemental arsenic sublimes and oxidizes to arsenic
trioxide. A garlic-like odor is produced during the oxidation process. At
about 200 °C it becomes phosphorescent. At about 400°C it burns with a bluish
flame and produces white smoke which is, of course, arsenic trioxide. 'In
the vapor state up to 800 °C, elemental arsenic consists of As. molecules having
a tetrahedral structure. Above 800°C, it begins to decompose to As_, and at
(24)
still higher temperatures, it becomes monatomic.
Though the common oxidation states are +3, +5, and -3, other oxidation
states are known. Examples are the polyarsenides Na-jAs.,, Na3As5'
-27-
-------
a series of naturally occurring copper minerals ranging in composition from
Cu-As to CIL.AS. Compounds or solutions containing the simple ions As
5+
and As do not exist because of the high energy requirements for acquiring
/2g\
three electrons 'or for ionization of five electrons.
o
In most compounds, arsenic exhibits a coordination number of 4, based on
tetrahedrally hybridized orbitals. Even the molecules AsH- and AsCl.,, where the
arsenic coordination number is 3, are assumed to be tetrahedral with a lone pair
/2g 2fi) "
of electrons in one of the hybMd orbitals.v ' '
Similarity to Phosphorus
As a member of Group 5a of the periodic table, the physiochemical pro-
perties of arsenic are closely related to those of phosphorus. Arsenates strongly
resemble the corresponding phosphates in solubility and crystal form, many phos-
phate-arsenate pairs being isomorphous. Arsenic also forms trihalides analogous
to those of phosphorus, and the arsenate ion reacts with ammonium molybdate in
nitric acid solution as does the phosphate ion. Generally, arsenates are much
more labile than corresponding phosphates, a fact important in the chemical and
(07)
biologic reactions within which both elements may participate.
Determination of Arsenic
The three most frequently used methods for the determination of arsenic
are:
1. Gravemetric determination as either As (+3) or As (+5) sulfide
which has been precipitated from an acidic solution by H2S.
2. Precipitation of silver arsenate with subsequent determination
of silver by Volnard's method.
3. lodcnetric titration of As (+3) in the presence of sodium bicarbonate.
To detect small quantities of arsenic, the Marsh test is used. The
arsenic-containing material is mixed with granulated zinc, and dilute sulfuric
acid is added. The zinc reacts with the acid to release hydrogen which reduces
the arsenic to gaseous arsine, AsH_. The arsine is then decomposed in a hot
glass tube giving a mirror of elemental arsenic. Arsine can also be detected in
a gas mixture by its reducing action on silver nitrate or mercury (+2) chloride.
-28-
-------
This is called the Gutzeit test, and the amount of silver nitrate or mercury
chloride reduced corresponds to the amount of arsenic present in the substance
are
(1)
(24)
being measured. Accuracies of these methods are 5 to 10 percent, and
limits of detection are on the order of 0.080 yg.
Highly accurate procedures (having limits of detection on the order of
0.001 yg) based on the determination of arsenic as arsine in an electric dis-
charge have been developed. These procedures permit the determination in
aqueous solution of arsenite ion and arsenate ion, as well as of the organic
species methylarsonic acid and dimethylarsinic acid (both of which are discussed
more fully below).(28)
Other highly accurate and precise procedures for measuring trace amounts,
though sometimes they are time consuming, include neutron activation analysis
1
(having a limit of detection near 0.001 yg), emission spectroscopy, and polaro-
(5 28)
graphic techniques. ' '
Inorganic Compounds
The most iitportant commercial arsenic compound is arsenic trioxide, also
known as arsenous oxide, "white arsenic", and (as a misnomer) arsenic. It occurs
as an octahedral crystal of As.O molecules. The dissociation to As40g can be
detected at temperatures of about 800°C. At a temperature of 1800°C, the mole-
(26)
cular weight is that of As2°3* Generally, however, the formula, As2°3' is
the one commonly applied, regardless of temperature.
Arsenic trioxide is a white solid (the commercial form is a white powder)
having a melting point of 275°C, though it begins to sublime at 135°C. It is
amphoteric and therefore soluble in both acids and bases, and is soluble in water
to the extent of 2 g/100ml water at 25°C and 11.5 g/lOOml at 100°C. Molecular
weight is 197.82 (76 percent As), and specific gravity is variable, 3.74 to
4.15.(29)
When arsenic trioxide is dissolved in water it forms arsenous 'acid, the
exact nature of which is not known; representative chemical formulas which have
been used include H_AS03, HAsO3, and As2°3 ^^ * ^ ^s a weak acid ha^^11? a
-29-
-------
sociation constant of 8 x 10 at 25 °C. HjAsO- is also thought to exist as tihe
hydroxide As (OH) which may explain the ability of arsenic trioxide to neutralize
both acids and bases:
As(OH)3 (s) + H+ -»• As (OH) 2 + H20
AS (OH) 3 : (s) + OH~ •*•
That only one dissociation constant is given for arsenous acid supports the
hypothesis that three hydroxyl groups are attached to the arsenic atom in the
free acid. The salts of arsenous acid are known as arsenites (As (+3) salts) .
The other commercially important oxide of arsenic is arsenic pentoxide,
(also referred to as arsenic oxide) . It is a white amorphous powder having a
(29)
molecular weight of 229.82 and a specific gravity of 4.086. ' Its chemical
structure is not known, though it is probably dimeric, As/).... The empirical
formula generally used is As-O-. The compound begins to decompose into a vapor
as AsJX and 0^ at a temperature of about 300 °C. It is very soluble in water,
though it dissolves slowly. Solubility is on the order of 2300 g/liter of water
at 20°C. (30)
In water, arsenic pentoxide forms arsenic acid (orthoarsenic acid) ,
HJVsO., a triprotic acid having three dissociation constants (as does phosphoric
acid). KL = 2.5 x 10~4, K2 = 5.6 x 10~8, and K3 = 3 x 10~13. The salts of
arsenic acid are known as arsenates (As (+5) salts) ; they are good oxidizing
agents. (30)
Arsenic pentoxide is commercially prepared by the dehydration of crystal-
line arsenic acid which is itself prepared by crystallization of a solution of
arsenic trioxide and concentrated nitric acid. The dehydration of the crystalline
arsenic acid takes place at about 200°C according to
Arsenic pentoxide cannot be prepared by the reaction of its constituent elements
or by the reaction of arsenic trioxide with oxygen. '
-30-
-------
The arsenates can be reduced by concentrated hydrochloric acid or sulfur
dioxide. Treatment of a solution of orthoarsenate with silver nitrate in neutral
solution results in formation of a dark-brown precipitate of silver orthoarsenate,
a method of distinguishing between arsenates and phosphates.
The most cannon arsenic hydride is arsine, AsH_, also known as hydrogen
arsenide and arsenic trihydride. It is a colorless gas, but it has a character-
istic garlic odor. Vapor density is 2.7 times that of air. (An atmospheric con-
centration of 1 mg of arsine per cubic meter of air corresponds to 0.313 ppm at
250fc and standard pressure). The melting point of arsine is -116.3°C and the
boiling point is -62.4°C. Its solubility in water is 200 ml/liter at room
temperature. Of all arsenic compounds, simple AsH_ and its methyl derivitives
are the most toxic.(26'30)
Arsine is the product of the reaction between atomic hydrogen and arsenic;
however, the reaction cannot.be carried out by the direct union of arsenic and
hydrogen because arsine is not stable and will decompose well below 300 °C. Arsine
is formed whenever any inorganic arsenic-containing material is reacted with zinc
and strong acids. Pure arsine can be condensed at low temperatures from a dried
gas stream produced by a reaction of arsenic pentoxide with hydrochloric acid
andztac.'26'30'
Exposure to arsine gas may result from the action of acids on metals con-
taining arsenic, from the use of impure sulfuric acid made from pyrites containing
arsenic, or from the use of hydrochloric acid made from impure sulfuric acid that
contains arsenic. Arsine poisoning has resulted from slushing out steel tanks that
had previously contained a commercial grade of sulfuric acid, the diluted acid
acting upon the metal tank to generate hydrogen, which combines with arsenic
impurities in the acid. Arsine may arise from the pickling of any metal con-
taining arsenic; it has been formed from the action of water on metallic arsenides
or hot dross containing arsenic and aluminum. Arsine may occur as an impurity in
acetylene and may present a hazard either in its manufacture or use. It may
occur in soldering, etching, lead plating, electrolysis of arsenious solutions,
by the action of moisture on ferrosilicon, or from the use of impure or inhibited
acids for scale removal. According to Patty, the faint garlic-like odor of
-31-
-------
araine cannot be considered a suitable warning property. The 1961 ACGIH thresh-
• • (29)
hold for arsine is 0.05 ppm.
Arsine is a good reducing agent, capable of reducing many substances.
It is not oxidized by air at room temperatures but may be ignited with the
formation of either arsenic trioxide or arsenic pentoxide, depending upon the
supply of air. Arsine reduces dilute silver nitrate solution with the forma-
»
tion of metallic silver; with concentrated silver nitrate solution, a complex,
Ag3As-3AgN03, is formed which yields metallic silver when diluted with water.
Mercury (+2) chloride is reduced stepwise forming initially the yellow compound,
AsH(HgCl)2, then the brown As(HgCl). and, finally, jblack AsJJg.. Chlorine re-
acts with arsine to produce hydrogen chloride and arsenic. However, at low 9
temperatures, the action of chlorine upon arsine prcxJuaes chloroarsines, AsH0Cl
/2g\ *
and AsHCl2/ both of which are relatively unstable yellow solids.v
Two other arsenic hydrides have been reported, but their exact chemical .
natures have not yet been determined. Reduction of trivalent arsenic compounds
by tin (+2) chloride in hydrochloric acid yields a brown amorphous powder
corresponding to the ocnposition As2H2 (or AsH). This material is soluble in
nitric acid but not in water, alkalies, or other acids. It reduces silver nitrate
and the salts of other heavy metals. Treatment with boiling water causes evolu-
tion of hydrogen and the formation of arsenic oxide. It is thermally unstable
and decomposes when heated in a vacuum to form metallic arsenic and sane arsine.
The other solid arsenic hydride is reported to have the formula As JH_ and is
(261
formed by oxidation of arsine with tin (IV) chloride.
The mono-, di-, and trimethylated forms of arsine are discussed below
in the Organic Compounds portion of this section.
The three arsenic sulfides are arsenic (+3) sulfide (arsenous sulfide,
arsenic sesquisulfide, arsenic red), arsenic sulfide (arsenic monosulfide, arsenic
disulfide), and arsenic (+5) sulfide (arsenic pentasulfide).
Arsenic (+3) sulfide (As.S&, As2Sj has a melting point of 320°C and a
boiling point of 707°C. Like many arsenic compounds, sublimation takes place
before melting. It is insoluble in acid and almost insoluble in water (0.52 mg/
-32-
-------
liter at 18°C) , but it dissolves readily in wary basic solutions. It will burn
in air, forming arsenic trioxide and sulfur dioxide.
Arsenic sulfide (As.S, /As2S2 , AsS) occurs naturally as realgar. It has
a melting point of 307 °C and a boiling point of 565 °C. Arsenic sulfide is
listed as insoluble in water and in hot concentrated hydrochloric acid, though
it is soluble in warm alkali hydroxide and sulfide solutions. The compound can
be oxidized by nitric acid and will react vigorously with chlorine.
Arsenic (45) sulfide (As0S,.) is a stable compound at room temperature,
£ O
but at temperatures above 95°C it dissociates into arsenic (+3) sulfide and
sulfur. It is soluble in water to the-j extent of 3 mg/liter, and in boiling water
it is hydrolyzed yielding sulfur and arsenous acid. It is soluble in basic
solutions and in nitric acid. It can be precipitated at lew temperatures from
strong acidic solutions which -contain arsenates by bubbling hydrogen sulfide
(26)
through the solution at a rapid rate.
Arsenic forms a complete series of trihalides, but arsenic (+5) fluoride
is the only simple pentahalide known. Whitacre and Pearse cite the reference of
Hbdgman, et al, to the possible existence of arsenic pentachloride and pentaiodide,
though such existence is believed unlikely.
Unlike phosphorus and antimony, arsenic forms no well-characterized oxy-
halides, but arsenyx chloride, AsOCl, and arsenyl bromide, AsQBr, are considered
likely to be present in the brownish material formed by treatment of arsenic
trioxide with the corresponding trivalent arsenic halide. All of the arsenic
are covalent compounds that hydrolyze in the presence of water. The
trihalides form pyramidal molecules similar to trivalent phosphorus analogs and
(26)
may be prepared by direct combination of the elements.
Arsenic fluoride (AsFO and arsenic chloride (AsCl_) are both colorless
liquids at 25°C, whereas arsenic bromide (AsBrJ is a yellow solid and arsenic
ilorL
: hal:
(25)
iodide (Aslj is a red solid. Arsenic (+5) fluoride (AsF_) is a colorless gas
3 (5Y
at 25°C, though it can be condensed to a yellow liquid.v ' Arsenic halides are
soluble in non-polar solvents such as benzene and carbon disulfide.
-33-
-------
Arsenic forms compounds with most metals, a number of which are naturally
occurring f such as safflorite (CoAsJ, niccolite (NiAs), rammelsbergite (NiAs2),
loellingite (FeAsJ , and sperrylite (PtAs,,). In addition, minerals containing
arsenic, sulfur, and one or more metals such as arsenopyrite (FeAsS), oobaltite
(OoAsS), glaucodot ((Co, Fe)AsS), and gersdorffite (NiAsS) are well known. Many
of the metallic arsenides may be prepared by tiie direct combination of the
elements. These compounds frequently resemble alloys and may consist of giant
molecular lattices. The apparent oxidation number of arsenic in many of these
compounds is frequently unusual. '
Arsenic generally behaves as an anion in the form of arsenites and
arsenates. There are no/arsenic carbonates, bicarbonates, or phosphates. The
only major inorganic compounds in which arsenic acts as a cation are the halides
and sulfides. There is an arsenic monophosphide (AsP) which dissociates in water,
and arsenic (+3) sulfate (As2SO.)_) which is formed by the reaction of arsenic
trioxide and SO, at a temperature of 100°C. Arsenic (+2) sulfate is soluble in
water.(30>
«
Organic Compounds
The largest class of arsenical compounds are the organic compounds. They
are seldom found in nature — most have been synthesized, largely in an effort to
find compounds having therapeutic value. Of the large number of organic arsenicals
the two most common are the arsonic acids and the arsinic acids. Their structural
formulas being:
As OH R As OH
OH R'
Arsonic Acid Arsinic Acid
-34-
-------
The R and the R1 refer to a variety of organic groups, and although there
are many derivatives of these two acids, only cacodylic acid (dimethylarsinic acid)
and methanearsonic acid (aLso referred to as methylarsonic acid) and its salts
(31)
are widely used, mostly as herbicides.
Methanearsonic acid and caoodylic acid are relatively strong acids,
capable of decomposing carbonates. Methanearsonic acid is a dibasic acid, forming
both a monsodium acid salt and a disodium salt with sodium hydroxide. Cacodylic
acid is normally monobasic, but in strong sodium hydroxide solution it forms the
disodium salt, (CH_)jAsO-HNa^, of a tribasic acid. Cacodylic acid is somewhat
amphoteric, forming a hydrorihloride, (CH-)-AsCLH'HCl, by direct reaction with
(31)
hydrogen chloride gas.
Both of these organic acids contain arsenic in the fully oxidized penta-
valent state, so only the methyl groups can be further oxidized. This requires
a strong oxidizing agent such as a nitric-sulfuric acid mixture. The end pro-
duct is orthoarsenic acid.
Methanearsonic acid and its salts can be reduced with mild reducing agents
such as nascent hydrogen and sulfur dioxide to form arsenosomethane, CH,AsO, a
trivalent organic arsenical. Cacodylic acid and its salts can. also be reduced
to form cacodyl oxide (CHO^AsQAs(CH_)0, also a trivalent organic arsenical,
(31)
although a much stronger reducing agent, such as phosphorous acid, is required.
Reduction of arsonic acids with mild reducing agents gives either the
arsonous acids, RAs(OH)0, or their anhydrides, (R&sO) , termed, arsenoso compounds.
At X
With aliphatic arsonic acids or with aromatic arsonic acids in which the ring is
unsubstituted or substituted with electron-repelling groups, the arsenoso com-
pounds are the reduction products. With aromatic arsonic acids containing electron-
attracting groups, the arsonous acids are usually obtained. The usual reducing
agent is sulfur dioxide and hydriodic acid. The actual reduction is accomplished
by the hydriodic acid, and the resulting iodine is reduced again to hydriodic
acid by the sulfur dioxide. Sulfur dioxide alone is used in some cases. The
reaction is usually carried out in hydrochloric acid solution in which case the
actual reduction product is tihe dichloroarsine, RAsCl2. Thase, however, are
readily hydrolyzed, either by alkali or by water alone, to the; arsonous acid
-35-
-------
or the arsonoso compound. The arsonous acids are woak acids. Both arsonous
acids and arsenoso carpounds dissolve in strongly alkaline solutions, but their
( 2fi}
salts have not been isolated.
The reduction of arsinic acids under the same conditions used for the
reduction of arsenic acids gives either the arsinous acids, lUAsQH, or their
anhydrides, (R-AsJ^O. These same compounds can also be prepared by the reaction
of arsenic trioxide with Grignard reagents. The anhydride of dimethylarsinous
acid, (CH3)2AsQAs(CH3)2, is cacodyl oxide, which is of historical interest since
it was the first organic arsenical ever synthesized.
Dihaloarsines, RAsX_, and monohaloarsines, K^AsX, may be prepared by a
wide variety of methods including reduction of arsonic acids in hydrohalic acid
solution with sulfur dioxide and hydroiodic acid. They may also be obtained from
arsenic trichloride and organonetallic compounds such as the organic mercurials
(26)
or organo-lead compounds.
When acetylene is passed into arsenic trichloride solution in the pre-
sence of a catalyst such as aluminum chloride or mercury dichloride, a mixture
of three products, ClCH=<2B\sCl2, (CH»=CH)2AsCl, and (ClOHSO-jAs, is obtained.
These are the "lewisites", after the American Chemist W.L. Lewis; the first of
the three is colorless or brown in the liquid state, and because of its powerful
(26)
vesicant qualities has been proposed as a war gas.
Diazomethane can be reacted with arsenic trichloride to form chloro-
methyl-dichloroarsine and bis(chloromethyl)chloroarsine. In addition to these
haloarsines, other compounds of the types RAsX_ and R-AsX, where X is a group
such as cyano, thio-cyano, or cyanato, can be formed. They can be formed by
(26)
metathesis between the halorarsine and a silver or sodium salt.
The reduction of arsonic acids with stronger reducing agents gives
arseno compounds having the empirical formula PAs. Appropriate reducing agents
of these compounds include sodium hydrosulfite and hypophosphorous acid. Electro-
lytic reduction has also been used for the preparation of arseno compounds. The
(26)
arseno compounds were at one time widely used in medicine, but not any longer.
-36-
-------
The reduction of arsinic acids and other arsenicals containing the
grouping gives secondary arsines, R_AsH. Dimethylarsine can also be prepared
from Ca(AsH2)2 and methyl chloride. Diphenylarsine can also te prepared by the
action of water on the Grignard reagent, (CgH5) -AsMgBr. Primary and secondary
arsines are readily oxidized in air and must be preserved in cin inert atmo-
(26)
sphere. They are very toxic.
(26)
According to Doak, the tertiary arsines, of the form R-As, are "more
important" than the primary or secondary arsines. Several methods of preparation
are given. Trimethylarsine is a byproduct of the action of certain molds growing
on a suitable substrate of arsenical compounds. Methylarsines, especially tri-
methylarsine , have been included in various natural cycles in soil and fresh
water. These cycles are discussed below.
Arsenic also forms a series of pentavalent chloro-comfounds of the form
RAsX4, R^X.,, and P.-AsX-. Compounds of the type RAsX., where: R may be either
aliphatic or aromatic, are not very stable and have not been thoroughly investi-
gated. Compounds of the type RJVsCl- are more stable than the tetrachlorides .
The reaction of tertiary arsines, both aliphatic and aromatic, with halogens to
give compounds of the type R-^AsX- has been studied extensively. When one of the
R groups is methyl, these compounds readily lose methyl chloride on heating to
(26)
give chloroarsines.
In addition to the dihalides, mixed compounds of the type R_AsXY are
known, in which X and Y are two different groups. Thus, the reaction of di-
methylphenylarsine with hydrochloric or nitric acid in the presence of hydrogen
peroxide gives the hydroxychloride , (CH.J2C,H_As(OH)Cl, or the hydroxynitrate,
(CH3) 2C6H^s (OH) N03 , respectively . (26 )
Oxidation of the tertiary arsines gives either the arsine oxides,
R_AsO, or the arsine dihydroxides. Arsine sulfides of the type R^AsS have also
(26)
been prepared.
The organic chemistry of arsenic is complex and involved, and the reader
interested in further information is referred to excellent summary by Doak, et
al, in the Encyclopedia of Chemical Technology; an extensive bibliography is
included.
-37-
-------
Various trj.va.lent organic prepara tions of or ionic have been usod, mostly
in the first half of this century, for the treatment of syphilis, trypanosoniasis
(sleeping sickness), spirochetal infections, amebic dysentery, psoriasis, and
even leukemia. Nowadays, however, arsenical compounds find little use in medicine
either because of the toxic hazards of the arsenicals or because more specific
medications having lesser side effects are available.
Arsenic Adsorption and Coprecipitation
Arsenic can be fairly easily separated from other elements, and can be
removed from solutions by adsorption and coprecipitation. Arsenic can be pre-
cipitated in the elemental state by reducing agents such as hypophosphite or
stannous chloride. Hypophosphite has been used to precipitate arsenic from
solutions of 1:1 hydrochloric acid, with a recovery of about 95 percent when
copper is present to catalyze the reduction.
Pentavalent arsenic, which includes arsenates, can be coprecipitated with
ferric hydroxide or magnesium ammonium phosphate. In the former case, it is
believed that the arsenates adsorb onto the surface of the hydrous iron oxides.
Ferguson and Gavis report that iron ores are enriched with arsenic because of the
high adsorptive capacity of the hydrous iron oxides. Iron oxide has a positive
surface charge and therefore adsorbs anions. Since arsenic exists primarily
as anionic arsenate and arsenite species in solution, it can be adsorbed on the
sur
(5)
surface charge and therefore adsorbs anions. Since arsenic exists primarily
is in
positively charged iron oxide surfaces. Arsenates can also be adsorbed by
aluminum hydroxides and clays.
Trivalent arsenic has a strong affinity for sulfur and will coprecipitate
with metal sulfides. Arsenic trisulfide, As_S_, is insoluble in hydrochloric
acid, and, hence, precipitation by hydrogen sulfide from a 25-percent solution
of HCl is used as a method of qualitative analysis for the presence of arsenic
in solution. The technique of adsorption of arsenites onto hydroxides and clays
is, according to Whitacre and Pearse, a premising candidate for arsenic water
pollution abatement;* ' Adsorption and coprecipitation processes are discussed
in the sections below dealing with water and soil chemistry.
-38-
-------
White Arsenic Refining
The precursor material for virtually all arsenical compounds is arsenic
trioxide, or white arsenic, as it is more commonly known in the trade. In the
United States the only producer of white arsenic is the American Smelting and
Refining Conpany, and their arsenic refining operations are carried out at Taooma,
Washington, where ASAROO has the facilities for the smelting of copper ores and
otiier base-metal ores containing large portions of sulfur and arsenic. The
arsenic refining portion of the plant is unique in comparison to other mined-
mineral production facilities in that the arsenic trioxide is recovered as a
flue-dust byproduct from other smelting operations; it is this relatively vola-
tile dust which must be purified. ASAROO processes its own flue dusts, which
contain as much as 30-percent arsenic plus other oxides of perhaps copper, lead,
zinc, and antimony. ASAROO also processes the flue dusts of other base-metal
producers both in this country and abroad. Carapella's description of white
arsenic refining, in the Encyclopedia of Chemical Technology, is the one most
commonly referred to in studies of arsenic:
Because arsenic trioxide is readily volatilized during the
smelting of copper and lead concentrates/ it is concentrated with
the flue dust. This crude flue dust is further upgraded by mixing
with a small amount of pyrite or galena concentrate and roasting.
The pyrite or galena is added to prevent arsenites from forming
during roasting and to obtain a clinkered residue which can be
returned for additional processing. The gases and vapors are
passed through a cooling flue which consists of a series of brick
chambers or rooms called kitchens. The temperature of the gas
and vapor is controlled so that they enter the first kitchen at
200 °C and by the time the gas and vapor reach the last kitchen
they are cooled to 100°C or less. The arsenic trioxide vapor which
condenses in these chambers is of varying purity. The condensed
product is obtained by resubliming the crude trioxide. The re-
subliming operation is normally carried out in a reverberatory
furnace. The vapors pass first through a settling chamber and
-39-
-------
then through approximately 39 kitchens that cover a length
of about 225 feet. The temperature of the settling chamber
is kept at approximately 295 °C, which is above the condensa-
tion temperature of the trioxide. A black, amorphous mass
containing about 95 percent As2^3 condenses in the kitchens
nearest the furnace and is reprocessed. Ihe bulk of the
trioxide is condensed in the kitchens with temperature ranges
of 180-120 °C. Ihe purity of the arsenic obtained from these
kitchens is from 99 to 99.9 percent. The dust which exits ,
from the kitchens as a temperature of 90-100 °C is caught in
the baghouse. It assays about 90 percent As?0_ and may be
sold as a crude arsenic or reprocessed.
The refined arsenic is analyzed for purity. It is also '
treated for "solubility"/ a term referring to its rate of re-
activity with nitric acid; this test is important if the
arsenic is used in the manufacture of insecticides and herfjicides.
The product is graded for marketing as white soluble (99 percent
min. As-CL) , white insoluble, or crude (95 percent min. As,,0.,) .
The diagram below is a schematic flow diagram of the operations in the
refining of arsenic trixoide.
SMELTER FLUE DUSTS FROM REVERBERATORY FURNACES
| (UP TO 30% As?0s)
ROASTING
COLLECTED FUME
90-95%
REPROCESS
; J 9
TO MARKET
REVERBERATORY
ciiDUArr .•
>-
REFINING
1
FUME
*
COOLING _
CHAMBERS —
OR
"KITCHENS"
, 1 ,
-JBAGHOUSE]
95% As20s
COLLECTED AT FURNACE END
^J
I
1 1
99.0—99.9% As-0,
TO MARKET
-40-
-------
Chemistry of Arsenic in Fresh Water
The chemistry of arsenic in aquatic systems is oonplex, involving oxidation-
reduction, microbial intervention, and adsorption and copreclpitation reactions,
among others, and not all of it is well understood. Ferguson and Gavis, in their
paper, A Review of the Arsenic Cycle in Natural Waters^ , have devised a diagram
showing the regions of stability of various inorganic arsenical species (e.g.,
arsenic acid in various states of dissociation) as a function of pH and oxidation
condition of water. With regard to the organic arsenicals, they state that
except under very reducing conditions in water, the organic ccmponent of the
arsenicals will undergo oxidation.
The equilibrium conditions of inorganic arsenic in solution are well
understood, but except for a few oxidation-reduction reactions as are used in
analytical chemistry, very little is known about the rates of arsenic reactions
in solution, and specific rate constants are unknown. For example, the rate of
oxidation of arsenite to arsenate with (X is reportedly very slow at neutral
pH values, but in strongly acid or alkaline solutions the reactions proceed
measureably in several days unless copper salts and carbon are: available in
the system to catalyzs the reaction. No quantitative information is available
about the rate of such reactions in aerobic waters, according to Ferguson and
Gavis.
Inorganic arsenic in water is commonly analyzed by means of colorimetric
methods based on colored complexes formed with diethyldithiocarbamide or molybdate.
Other analysis methods include neutron activation, atomic absorption and emission
spectroscopy, and polarographic methods. Colorimetric and polarographic methods
can also be used to determine oxidation states in inorganic arsenic.
A lack of suitable analytical chemical procedures has hampered studies
of arsenic in water/ especially the determination of the inorganic arsenic ions
and the methylarsinic acids at very low concentrations. Most methods used for
the determination of arsenic in low concentrations measure the total elemental
concentrations, and many depend on the reduction of inorganic arsenic ions to
arsine and subsequent colorimetric analysis. The lower limit of detection of
the silver diethyldithiocarbamate method is not lower than 0.2 ug/ and though
-41-
-------
neutron activation methods have a limit of detection of near 0.001 ug they are
relatively time consuming. The methods employed by Braman and Foreback have enabled
them to distinguish arsenite, arsenate, methylarsonic acid, and dimethylarsinic
acid to lower limits of detection of near 0.001 yg. Their procedures depend upon
pH selective reduction reactions of the various arsenic forms witii sodium boro-
hydride and a separation of the volatile arsines produced by selective volatiliz-
(28)
ation from a cold trap.
Arsenic forms stable bonds with sulfur and carbon in organic compounds.
As is discussed in the 'toxicology portion of this study, it is the affinity of
trivalent arsenic (arsenite) for sulfhydryl groups, most notably in the amino
acid cysteine in proteins, and the resultant enzyme .'nactivation, which accounts
for the primary mode of arsenic toxicity. Pentavalent arsenic (arsenate) does
not react with sulfhydryl groups, but reduction of arsenate to arsenite can take
place within organisms both large and small, and, in the case of certain water-
borne fungi, according to Challenger (as reported by Braman and Foreback), such
reduction processes in. natural waters could cause an increase in the ratio of
(28)
more harmful arsenite to less harmful arsenate.
The methylarsines are an important group of .arsenical conpounds within
natural systems. Mono-, di-, and trimethylarsines, .ind even simple arsine have
(5 8 28 37)
reportedly been synthesized by such organisms as yeast, fungi., and bacteria. ' ' '
The proposed metabolic processes producing these ars Lnes are based on both inorganic
and organic arsenical precursors, and have been stab id to occur in both aerobic
and anaerobic settings. Microbiological processes hive also been identified as
the sources of other methylated arsenicals, most nofcibly the methylarsonic and
dimethylarsinic acids, which themselves are included in biological cycles which
(8 28 33)
include the synthesis of methylarsines. ' '
Trimethylarsine has been identified as an imjortant reservoir of arsenic
(5)
in certain organisms. And although it is considered insoluble in water,
Ferguson and Gavis cite it as being sufficiently soluble to be of environmental
interest, especially sinoe it has caused human poisoning in its vapor phase in
air. It is more soluble in hydrocarbons than water, which may account for its
-42-
-------
accumulation in the fatty tissues of certain aquatic organisms. However, neither
its stability with respect to oxidation by oxygen in air or in water, nor its
adsorptive behavior appear to have been studied.
(28)
Braman and Foreback ' report that a large portion of arsenic that is
found in human urine (up to 90 percent) is methylated. They suggest this may be
the result of a biological defense mechanism against the much higher toxicities
of inorganic arsenates and especially arsenites which are 25 times more toxic
than dimethylarsinic: acid. The methylated types of arsenic in. urine are di-
methylarsinic acid (cacodylic acid) and methylarsonic acid (which is the same
as methanearsonic acid, the sodium salts of which are the herbicides DSMA. and
MSMA.).
Methylation of arsenic by bacteria has been studied by Wood and by McBride
(8 28 33)
and Wolfe.V0f Of J ' McBride and Wolfe have shown that anaerobic bacteria can
produce mono- and dimethylarsine from a variety of arsenic compounds, and they
have suggested a cycle in which methylcobalamin serves as the methyl donor in
the reaction system. (Methyl cobailamin is also cited by Braman and Foreback as
the methyl donor for the methylated arsenic found in human urine, the reaction
presumably taking place within the body.) Arsenate is first reduced to arsenite
which is then methylated to methylarsonic acid which is further methylated to
form dinethylarsinic acid which in turn finally becomes dimethylarsine. The
(8)
diagram below illustrates the process of methylation.
OH
llf\ — A n ._/~lU _
0
Arsenate
CH2-BI2 B
2*
1
0
Arsenite
CH3
0
Dimethylarsinic acid
CH3
> Bl2r ' 34
0
Methylarsonic
CH3
^- As-CH3
H
Dimethylarsine
OH
acid
CH3—B,2= Methyl cob(lll)alamin
B,2r= CobODalamin
-43-
-------
Braman and Foreback measured the relative amounts of dimethylarsiriic
acid and nethylarsdhie acid in various fresh and salt water systems in Florida.
Methylarsonic acid, though present, was generally in smaller concentrations than
dimethylarsinic acid, possibly because of the greater tendency of methylarsonic
acid to oxidize (whereas dimethylarsinic acid is very resistant to oxidation),
or possibly because, as shown in the methylatj.cn cycle suggested by Wood, *33^ the
oxidation product of the methylarsines is dimethylarsinic acid which finds its
way back into the water system. It appears possible that dimethyarsinic 'acid
and methylarsonic acid could accumulate (from both biological and pesticide runoff
sources) to an extent wnere methylarsine generation and its subsequent deposition
in marine organisms mioht become significant. As' stated earlier, very little is
known about the rates at which 'these reactions take place, and thus, the residence
time of the slow-to-oxidize dimethylarsinic acid could be appreciable, affording
possibly plenty.of time for further bacterial reduction to dimethylarsine and
subsequent accumulation in aquatic species harvested for_fopd. •
Air
CH
Water
\
CHs-A*tCH8 + H-
TrlmMhylartln* Olmothylartln
^
Moldt
HO-Air^OII-p» —A»5-OH f "-HO-As'-OH
0 iooUrlo 2 Oootirlo n BooUrla
ArMKtli AriiAlIt
Stdlmtnt
MtihylarMnlo
told
— HO— A»— C
Olmtlhylorilnlo
odd
A BIOLOGICAL CYCLE FOR ARSENIC
With regard to the biological methylation of metals, Ferguson and Gavis
report that the biological advantage, if any, is not known. Ferguson and Gavis
as well as Braman and Foreman suggest a possible detoxification advantage in
methylation since methylarsonic acid, dLmethylarsinic acid, and even the methyl-
arsines (so long as they are in solution) are less toxic generally than the tri-
and pentavalent inorganic precursors. Also, in anaercbic environments, the
methylation of metals by microorganisms may be more thermodyramically favorable
than the synthesis of methane, f
-44-
-------
Ferguson and Gavis state that only aerobic metabolism has been found
to yield nethylarsines, and that there is not a priori reason why anaerobic
synthesis of methylarsines could not also be possible. Wood, on the other
hand, referring to the work of McBride and Wolfe (which is also referenced by
Ferguson and Gavis), states the methylarsines are produced by anaerobes.
Whether aerobically or anaerobically synthesized, Ferguson and Gavis state that
msthylation is not thermodynamically favored in water and can occur only in the
presence of organisms.
(Of all sources referenced here with regard to the methylation of
(5 8 28 33)
arsenic, ' ' ' all express concern about the extreme toxicity of the methyl-
arsines. However, there is evidence, as cited in the lexicological Assessment
portion of tiiis study, that methylarsines, while in solution or otherwise con-
tained within aquatic organisms, may be of extremely low toxicity, especially
in comparison to their gaseous state).
Arsenic is reiroved from the solution phase by such reactions as adsorp-
tion onto clays and coprecipitation into metal ion precipitation. Arsenate,
because it is the fully oxidized form of arsenic, is the stable form in aerobic
waters, but it may be removed by several mechanisms. For example, that fact
that iron oxide has a positive surface charge in most geologic environments has
been cited as a reason for the high arsenate adsorption (arsenate is anionic)
onto hydrous iron oxides. Arsenate species ooprecipitate with or absorb onto
hydrous iron oxides. In addition, ferric arsenate is very insoluble.
Arsenite species (trivalent) may be present in surface waters under
sufficiently reducing conditions, or if the oxidation to arsenate (pentavalent)
is not complete. Arsenous acid species will adsorb onto and/or ooprecipitate
with iron oxide in a manner similar to that of arsenic (As (+5)) acid.
Aluminum hydroxides and clays also adsorb arsenate species; however,
bauxite and silicates are usually only moderately enriched in arsenic. The
affinity of arsenite, on the other hand, for clays, and hydroxides other than
iron has not been investigated. However, because of the strong affinity of
arsenite for sulfur, metal sulfides readily adsorb and ooprecipitate arsenite.
Goldschmidt and Peters measured up to 3000 mg AsAg of sedimentary pyrite, FeS .
-45-
-------
Chemistry of Arsenic in Soil
Arsenical ccnpounds arrive in the soil in tho form of pesticides, as
fallout from smelting operations, from the burning of coal and cotton wastes,
and from runoff from mining operations. Arsenic can accumulate in soil to levels
that are phytotoxic. Treated soils in North America may contain between 1.8
and 830 ppm As, while untreated soils range from 0.5 to 14.0 ppm As.^ '
When arsenic reaches the soil, it reacts with tihe soil and soil solution
(34)
to form compounds of various solubilities. Among the cations that react with
arsenic are iron, aluminum, calcium, and magnesium. It also reacts with the hydrous
iron and aluminum oxides that cover clay particles in soil. During the reaction
process, the chemical equilibrium of arsenic is changed. The amount of arsenic
in solution decreases in accordance with such factors as soil pH, available
(35)
cations, and the amount of organic matter present. Nutrients in the soil,
especially phosphorus because of its chemical similarity to arsenic, also affect
the rate and degree of arsenic fixation. Phosphorus competes with arsenic both
for fixation sites on clay particles and for uptake by plant roots. The degree
of phytotoxicity due to arsenic is a function of the total amount of soluble
arsenic in the soil.(35'36)
Of the sources of arsenic reaching the soil, arsenical pesticides are
the most widely distributed. Arsenic acid (H~AsO.) is applied to cotton for leaf
desiccation or to vegetation as a general weed killer. The organic arsenicals,
methanearsonates and cacodylic acid (dimethylarsinic acid), are selective and
general postemergence herbicides, respectively. Other forms in which arsenic
may reach the soil are as trivalent salts, pentavalent salts, and, in the case
of smelter fallout, simply as arsenic trioxide, As^0^* But regardless of the
form in which the arsenicals arrive, they are eventually oxidized and/or metabo-
lized to arsenates.(35'37)
The amount of time for an equilibrium condition between soluble and in-
soluble arsenical species to be reached can be anywhere from several days to many
months, depending upon the initial amounts of arsenic introduced to the soil and
upon the soil variables listed above (available catiohs, pH, etc.). Insoluble
-46-
-------
arsenical species predominate in soils rich in iron, calcium, and aluminum, which
means that such soils would tend to exhibit rapid initial reduction of arsenical
phytotoxicity, and orce equilibrium is reached, arsenical phybotoxicity would be
low even though the total amount of arsenic in the soil might be appreciable.
When the initial application of soluble arsenicals is large, the rate of
conversion to insoluble forms (on a percentage basis) is slower than when small
amounts are applied. But with either large or small initial amounts, the soil
decrease of solubles, and the corresponding increase in insoluble salts, typically
varies as shown.
insol'jble Fe arsenical
insoluble Al
arsenical
/"water sol. As
insoluble Ca arsenical
TJiat iron-arsenical is shown as being the predominant insoluble compound
is purely arbitrary, simply for the sake of illustration. Low volumes of initially
water soluble arsenical, in a given soil type, generally decrease more rapidly
(e.g., 90 percent conversion to insoluble form within one wee]*:) than large volumes
(e.g., 50 percent in 24 weeks). In other words, the initial slope of the water
soluble curve decreases as the initial (applied) amount of water soluble arsenicals
increases; chemical equilibrium is reached more rapidly witih lower initial levels
of arsenicals.
-47-
-------
Arsenic Removal Fran Soils
Arsenic is renewed fran soils by three mechanisms: Leaching and runoff;
plant uptake; and biological transformation.
Leaching and Runoff
Soluble arsenical compounds can be carried in solution from soil. In
the case of leaching, the soluble forms can be carried deeper into the soil pro-
file where they combine with available fixation sites. Leaching of arsenicals
to sufficient depths can effectively remove arsenic from the part of the soil
where crop roots are likely to absorb it. In the case of runoff, soluble arsenicals
are carried away from the soil, and eventually, find their way into ground water
(38)
or streams and rivers. '
Plant Uptake
Plants concentrate arsenic, and with sufficient concentration they die.
Plant roots concentrate arsenic at a rate of 10 to 100 times higher than plant
tops. Phytotoxicity results from "root pruning"; i.e., arsenic accumulation in
root tissues slows or halts root growth while the still-grcwiiig plant tops eventu-
ally become starved because of insufficient root size. The harvesting of crops
and especially the removal of whole plants - crop, stalk, roots, and all - is
(38)
a mechanism of arsenic removal from soils. Arsenic concentration in plants
and its effect on plant growth are discussed in the lexicological Assessment
portion of this report.
Biological Transformation
The biological transformations of arsenicals in soil are similar in many
respects to those taking place in water, especially with regard to the formation
of highly volatile arsenicals. Soil microorganisms both aerobic and anaerobic,
can mediate the transport of arsenic through soil; arsenic removal by volatization
results fron bacterial formation of arsine (AsH.J, methylarsines (mono-, di-,
(37)
and trimethylarsine), and other volatile organoarsenic compounds.
In no discussion of the soil chemistry of arsenic surveyed for this study
is the possibility mentioned of micrdbially-nediated reduction of arsenates to
-48-
-------
arsenites, as has been shown in water systems. Soluble arsenates in soil nay
undergo chemical reduction to arsenites, but the literature indicates that
oxidation to arsenates is more likely in soils, and that the product arsenates
are either washed from the soil or locked into insoluble complexes, which, as
described above, effectively controls arsenical phytotoxiciby. Microbiological
metabolic processes therefore act chiefly on organic arsenicals, mainly methane-
arsonic acids (which include the herbicides DSMA and MSMA) and dinethylarsinic
acid (the herbicide, cacodylic acid).
In aerobic soils, the organic arsenicals will, in time, be oxidized,
either chemically or as a result of biological processes, to carbon dioxide and
arsenate. It is also possible in aerobic soils for organic arsenicals to be
reduced to volatile organo-arsenical compounds in the same manner as described
in the section of this study dealing with the chemistry of arsenic in water?
namely, organo-arsenioals are reduced and methylated to mono-, di-, and trimethyl-
arsines, as well as to inorganic arsine and to other volatile organo-arsenicals«,
As would be expected, it is under the anaerobic conditions where the
largest portion of nonvolatile organo-arsenicals (specifically, caoodylic acid)
are converted to volatile forms instead of being oxidized. In a study by Wbolson
(37)
and Kearney of the degradation of cacodylic acid in three types of soils, an
average of 61 percent of applied cacodylic acid was converted under anaerobic con-
ditions to a volatile organo-arsenical within a 24-week period, whereas under
aerobic conditions, 35 percent was made volatile and 41 percent oxidized into
002 and arsenate in the same period. (Under the anaerobic conditions, none of
the cacodylic acid was oxidized to OX and arsenate).
The reactions of tiie methanearsonic acids in soil are similar to those
of cacodylic acid - metabolism to volatile compounds in both aerobic and anaerobic
(38)
soils, and oxidation (and off-gasing of (XL) in aerobic soils.
Dimethylarsine is a common volatile arsenical, and according to WoolsoW
(37)
and Kearney, it is so unstable that it may be oxidized back into cacodylic
acid by its contact with air and return to the soil to either repeat the cycle
or to by oxidized and finally fixed into the soil. The ultimate environmental
fate of arsenic in soil appears to be the formation of inorganic arsenate which
becomes bound as insoluble compounds in the soil.
-49-
-------
Oxidation of cacodylic acid and methanearsonicVacids can, however, be
part of biologically-mediated metabolic processes — at least in aerobic soils.
Waplson and Kearney found such evidence in measuring the evolution of CCL from
f*
soil which had on two occasions received cacodylic acid; on the second occasion
the adapted microbiological population metabolized the cacodylic acid much more
readily than in fresh soil: 13-percent release of the initial carbon ( C) after
98 days versus only 2 percent after the same period in fresh soil. Ohe diagram
below illustrates the difference.
14
12
10
8
C
4
2
(37)
20
40 60
DAYS
80
100
EVOLUTION OF l4COg FROM AN ADAPTED (a) AND A NON-
ADAPTED (b) SOIL TREATED WITH CACODYLIC ACID
Effects of Phosphorus (36)
Increasing phosphorus levels in nutrient solutions containing sufficient
arsenic to reduce growth has been shown to cause less arsenic to accumulate in
plants and to improve plant growth where it would otherwise be slowed by the
presence of arsenic. This affect, however, does not always hold true. In one
study where soil levels of Al and Fe were low, phosphorus seemed to magnify the
phytotoxicity of arsenic, possibly because the phosphorus combined with the few
fixation sites available so that arsenic did not form insoluble compounds which
would have taken it out of solution.
Since phosphorus is an important ingredient in fertilizers, it could play
a part in the phytotoxicity of the total arsenic in a given soil situation. If
the soil has a high potential for fixation of these two chemically similar elements,
the available phosphorus in solution will be preferentially absorbed by plants,
and the arsenic will not be as harmful. In soils with a low potential for fixation,
especially with respect to Fe and Al, phosphorus will be predominatly fixed while
arsenic will remain available for plant uptake, and, hence, phytotoxicity.
-50-
-------
In general, plant content of arsenic arid phosphorus appears to be a
function of their soil availability. Soils vihich are high in accumulated arsenic
(highly insoluble forms associated with Fe and Al, mainly in clay particles) are
affected by the addition of phosphorus. Phosphorus can increase the amount of
soluble arsenate in soil and, thus, hasten the leaching of arsenic from the top
soil. Thus, in soils containing initially high levels of insoluble arsenic,
high phosphate fertilizers may provide a mechanism for moving some of the toxic,
more soluble arsenic deeper into the soil profile.
-51-
-------
SECTION IV
ARSENIC PRODUCTION AND USES
Table 1 summarizes arsenic supply and use volumes for those years be-
tween 1968 and 1974 for which data were available. Though there is only one
domestic supplier of arsenic trioxide, production information is unavailable.
The supply information in the Table shows considerable variation betsaeen data
sources. With regard to exports, approximately 25 percent of the domestic pro-
duction was exported in 1974, according to a spokesman for ASARCO, the single
(39)
arsenic producer. '
The major uses of arsenic are:
Pesticides
Insecticides
Herbicides
Fungicides
Wood Preservatives
Feed Additives
Glass Manufacture
Nonferrous Alloying
Data Collection and Use Trends
Pesticides
Arsenical pesticides account for less than 3 percent of the total pesti-
cide market, and their share is decreasing as a result of cancellation of pesti-
cide registrations and because of the increasing concern of the Occupational
Safety and Health Administration for the health of arsenic workers - ocnpliance
with exposure standards, both current and proposed, has been named as a factor
in the decreased use of inorganic arsenicals.
Lead and calcium arsenate and Paris green, account for virtually all of the
arsenical insecticides currently used. The arsenical herbicides (including de-
foliants and desiccants) are the methanearsenates, cacodylic acid, and arsenic
-52-
-------
TRHLE I
ARSBJIC SUPPLY AND USE
1968 - 1974
(kkg As)
White Arsenic Supply - Total
Dcnestic Product] on
Imports (Metallic end Oxide)
Exports
Pesticides - Total
Lead Arsenate (n)
Calcium Arsonatoe m>
Jtethanear sonic Acid Salts
Cacodylic Acid IP)
Arsenic Acid
Wood Preservatives - Total
Chronated Copper Arsenate (CCA) (k)
Fluor Chrone Arsenate Phenol (FCAP) 00
Peed Additives (1)
Glass Manufacture
Ncnferrous Alloying
Xj figgl T grwiaiQ
1968
25, 460 (a)
21,600(f)
7,260(a)
5,540(f)
18,200'a)
13,500(e)
j.7,90(Xf)
NA*
17,70(Xa)
16,250(f)
465
340
NA
NA
724[i)
363[a)
637
346
291
349
3,725(f)
5,OOOCa)
2,80CKb)
363(a)
63«f)
2,00«a)
1969
NA
. 12,700 (e)
NA
750
364
NA
NA
96',-fi)
836
504
332
370
3,500 fa)
363 (n)
1,650(0)
1370
9,550(b)
13,200 (e)
NA
575
496
4,820(h)
NA
l,212(i)
847
650
197
397
3,160(m)
372 (n)
1,700(0)
1971
NA
12,250(6)
NA
>
406
421
3,900
-------
(d) Based on information supplied by the American Smelting and Refining Company
to the Occupational Safety and Health Administration with regard to the
proposed standard for exposure to inorganic arsenic.39
.(e) Fran the Arsenic section of the Connndity Data Stannaries 197518
(and personal comunication with Gertrude Greenspoon, BCM)
(f) Paone, Janes. Arsenic. In: Mineral Facts and Problems, 1970 edition.
U.S. Department of the Interior, Bureau of Mines, pp. 479-487. l7
(g) Personal onmrTinir^tiT? with a spokesman for the Ansul Conpany.
(h) Based on data from the Pesticide Review 1973, "3 '
and fron Fanners1 Use of Pesticides in 1971 . . . Quantities.1"
(i) Based on assumed linear growth between 1966 and 1971, two years for which
data were available - Fanners1 Use of Pesticides in 1971 . . . Quantities."9
( j) Projection of 1975 demand supplied by arsenic acid producers - assumed
linear relation 1971 through 1975.
(k) Data supplied by the American Wood-Preservers' Association.19
(1) Feed additive use in 1973 was inferred from information supplied by the
National Agricultural Chemical Association - years 1968 through 1972 are
based on broiler production for those years, adjusted to 1973 broiler
production.
(m) Based on assumed linear decrease between average of figures for 1968 and Asarco
data for 1974. J8
(n) Based on 90 percent of metallic arsenic imports as reported by Bureau of Mines.
(No metallic arsenic produced in the U.S. between 1968 and 1974, at whicn time
Asarco started producing it. ) ' 8
(o) Includes such uses as animal dips and paint pigments and additives which no
longer contain arsenic;63'65 from 1969 to 1971, basis is 10 percent of
imports of white arsenic; 1972 through 1974 based on 5 percent of white arsenic
(p) Caoodylic acid is a major arsenical herbicide, but no production or use data
ere available.
-------
acid; other arsenical herbicides such as sodium arsenite and mixtures containing
arsenic'trioxide were in use in the last decade, but if they sire still in use, no
producers or production information were uncovered in this study. The EPA
Compendium of Registered Pesticides lists several dozen arsenical herbicides and
«
insecticides, but the ones listed above are the only ones finding any use today.
Similarly for arsenical fungicides, several are listed but only one - 10,10'-
oxybispheroxarsine - is in use (as an additive to flexible vinyl plastic formu-
lations) , and production information is not available from the single producer.
Most of the pesticide data in the Table is from Department of Agricul-
ture sources, and part of it has been supplied by producers. Total arsenical
pesticide production is decreasing, but the high figure of 17,700 kkg of ele-
mental arsenic used ii pesticides in 1968 probably was a result of military use
of cacodylic acid as a defoliant (agent BLUE) in Vietnam. 1/42' use of caco-
dylic acid in Vietnam probably reached a peak around 1970, but no data are
available to substantiate this. Production of calcium and lead arsenate de-
creased between 1968 and 1972, but "domestic disappearance" data for the same
period (which takes account of imports, exports, and producer ;year-end inven-
tories) shows no trend. The lead arsenate and calcium arsenate data in the Table
is based on domestic disappearance since it better reflects demand than does the
production data.
The use of arsenic acid has increased dramatically during the last
decade. Arsenic acid is used almost exclusively as a cotton harvest aid (speci-
fically, as a desiccant) in Texas and Oklahoma where the so-called "dry-land"
cotton is grown. Cacodylic acid is also used as a cotton harvest aid, largely
in the nine other cotton-producing states.
Wood Preservatives
The two main arsenical wood preservatives are chromated copper arsenate
(CCA) and f luor chrcite arsenate phenol (PCAP). Small amounts of ammoniacal cop-
per arsenate (ACA) are also being used by a wood-treating plant: on the west coast.
CCA presently accounts for about 90 percent of the arsenical wood preservative
(44)
use, since once it becomes bound to the wood fibers it is impervious to
-55-
-------
leaching, as opposed to the other water-borne wood preservatives. Data were
supplied by the American Vfeod-Preservers' Association and by the producers of
arsenical wood preservatives. Use of CCA is increasing; PCAP is decreasing.
Feed Additives
Arsenical feed additives are used in the feed of poultry and swine to
increase growth rate and feed efficiency, and, in some instances, to control
poultry disease. Data were available only for the years 1973 and 1974; data for
&
previous years was derived on the basis of broiler production for the years 1968
through 1973, assuming a constant ratio of arsenic to broilers for each year.
t AC\ 6
According to an FDA source, arsenical feed additives are proportional in
total volume to broiler production, and will grow or decline accordingly.
e
Glass Manufacture
Data on the use of arsenic in the manufacture of glass is conflicting.
Of the glass manufacturing specialists contacted, one stated emphatically that
arsenic is no longer used in glass because of the handling hazards and because
of problems in disposing of arsenic containers. Another specialist said arsenic
was still used, but only in specialty and "art" glass. The use of arsenic in
glass is definitely decreasing.
Nonferrous Alloying
Arsenic is used in lead, brass, and copper alloys to improve certain
metallurgical properties. The apparent increase in this use shown in the Table
is based on increased imports of metallic arsenic as reported by the Bureau of
Mines. One of the major alloying uses is in lead shot where arsenic increases
the hardness and sphericity of the shot. However, the use of lead shot may be
curtailed by the Bureau of Sport Fisheries and Wildlife because of evidence that
ducks and other birds are being lead-poisoned from eating the shot; an iron-
tA->\
based shot will likely be the alternative.v '
Miscellaneous
The miscellaneous uses of arsenic include, or have included, the fol-
lowing during the last decade:
-56-
-------
- anijtal dips, paint pigments and additives, and leather tanning
chemicals - all of which no longer use arsenic
- Pharmaceuticals - human use of arsenical pharmaceutijcals has
effectively ceased, but they are still used in verterinary
application other than as feed additives
- electronics - semiconductor uses in diodes, transistors, lasers
and infrared devices, plus increasing use in light-emitting
diodes for digital readout.
The remainder of this section is a detailed discussion of the uses of
arsenic, the enviornmevital emissions resulting from these uses,, and alternative
materials for these uses.
Pesticides
Herbicides and insecticides are the two main pesticides. Other pesti-
cides include fungicides, rodenticides, miticides, acarcides, and nematocides.
At the present time, more that 80 percent of pesticides are organic chemicals,
a substantial change from the 1940's when pesticide chemicals were almost en-
tirely inorganic, with insecticides, the largest part of the pre-war pesticide
market, consisting largely of lead arsenate and calcium arsenate.
Insecticides
Volume III of the EPA Compendium of Registered Pesticides lists the
following arsenical insecticides:
Arsenic Pentoxide Lead Arsenate
Arsenic Trioxide Paris Green (copper acetoarsenite)
Basic Copper Arsenate Potassium Arsenite
Cacodylic Acid Sodium Arsenate
Calcium Arsenate Sodium Arsenite
Copper Arsenate Sodium Pyroarsenate
Two of these compounds, sodium arsenite and potassium arsenite, are
listed in the Compendium as acarcides (tick killers) used in animal dips. As
such, they are discussed separately in the Animal Dip section of this report.
-57-
-------
Inorganic
Tfifvvf* S f* i H£>***
(kkg)
•total
Insecticides
(kkg)
t Inorganic
of Tbtal
1
°
31
33,348
0.094
I
^
44
1,883
2.3
1
ft
8
2,724
0.28
B
8,
0
1
i
~
0.9
1,039
0.088
'1
1
1
1.4
1,313
1.0
s
1
I
I
"
90
3,861
2.3
H
°
154
19,605
0.78
•
1
<
842
4,782
17.6
t
1
1
S
•— I CO
i j i
s g
220
8x417
2.6
fi
I
'tj CO
fl fjl
(D p
II
g 35
0.5
300
0.15
.|
B
1,383*
97,717
1.41
*Figure includes cryolite (115 kkg)
arsenical.12
and sodium flue-silicate (7.3 kkg); remainder is
(That the total inorganic insecticide use on crops, 1,383 kkg, is less
than half of the reported production of lead arsenate and calcium arsenate for
the same year is accounted for by the uses of lead arsenate and calcium arsenate
in non-crop uses by government, industry, and homeowners.)
The only organic arsenical insecticides listed in the EPA Compendium
are cacodylic acid and paris green. Cacodylic acid is not used on crops and
paris green is used exclusively as a mosquito larvacide. Thus, all arsenical
insecticides used on crops are included under the heading "Inorganic Insecti-
cides in the above table. The largest use of inorganics is on apples (17.6% of
all insecticides used on apples) followed by "All other fruits and nuts"* (2.6%).
Thus, arsenicals - except in the case of apples - account for only a
small part of the total insecticides used on crops in 1971 (1.4%), and the down-
ward trend noted between 1966 and 1971 in inorganic seems to be continuing.
(50)
*grapes, avocados, figs, blackberries, blueberries, boysenberries, currants,
gooseberries, loganberries, raspberries, strawberries, almonds, filberts,
pecans, walnuts, olives, tung nuts.
-58-
-------
Arsenic pentoxide is used to protect wood against termites; it is discussed in
the Wood Preservative section.
Prior to WW II, lead and calcium arsenate were the "backbone of the
cide
/yioi
1940.
pesticide industry". These taro compounds significantly decreased in use since
(48)
1940 I960 1965 1967 1968
(1000 kkg)
lead arsenate 34.1 4.5 3.2 2.7 4.1
calcium arsenate 22.7 3.2 1.8 0.9 1.4
The compounds are no longer the backbone of the pesticide industry, but
they are the backbone of the arsenical insecticide industry, with lead arsenate
carrying the major portion of the burden.
1967 1968 1969 1970 1971 1972
(kkg)
lead arsenate 2,700 4,100 4,170 1,880 2,800 2,530
calcium arsenate 930 1,540 527 522 427 (w)
(w) E data withheld to avoid disclosure
0
Agricultural Economic Report No. 252, Farmers' Use of Pesticides in
fAQ)
1971 ... Quantities, " states that inorganic insecticide use in 1966 and 1971
was "relatively insignificant"; it dropped from 2,630 metric tons in 1966 to
1,450 metric tons in 1971, "down from 4 percent to less than 2 percent of all
Insecticides used".
The same publication lists the crops and amounts of insecticides used
in 1971. Under the heading "inorganic insecticides" (unspecified), ten crops
are listed.
-59-
-------
Herbicides
Arsenical herbicides used today are largely of the organic variety.
Inorganic arsenicals are rarely used any more/ with the notable exception of
arsenic acid used as a cotton harvest aid. In the past, arsenical herbicides
were usually of the trivalent form, which is usually more water soluble and
thus more easily absorbed by plant tissues, either through the roots or directly
through, the leaves.
Sodium arsenite was the standard weed killer for most of this century
until about 1960; that is, when more effective and more highly selective or-
ganic herbicides became available.
Arsenic trioxide, which is a relatively insoluble trivalent arsenic
compound, is used in soil sterilization. Its disadvantages include high dosage
rates (400 to"800 pounds per acre are required) and soil residues, which remain
for many years even though actual soil sterilization may be effective for only
A
a year or so. Non-arsenical herbicides are effective at dosages on the order
(31)
of only several pounds per acre.
Arsenicals kill plants via inhibition of enzymes containing sulfhydryl
groups. Protein precipitation within plant cells is a consequence of high ar-
senical concentrations. Arsenicals generally are not specific in their herbi-
cidal action.*31'
The organic arsenicals are classed as either arsonic or arsinic acids.
The basic structural formulas are:
0 O
II II
R — AS — OH R — AS — OH
I I
OH R'
Arsonic Acid Arsinic Acid
where R and R1 correspond to a variety of organic groups, cacodylic acid, which
is dimethylarsinic acid, is based on arsinic acid with methyl groups in place
of R and R1. The salts of arsonic acid are disodium methanearsonic acid (DSMA),
-60-
-------
inonosodium acid methanearsonate (MSMA), amine methanearsonate (AMA), and cal-
cium acid nethanearsonate (CMA).
Cacodylic acid and the arsenic acids are pentavalent arsenicals. They
are generally less toxic to animals than organic trivalent arsenicals, and they
are considerably less toxic than inorganic arsenicals. (See Table 11 which shows
the relative toxicities of various arsenical corrpounds.) The methanearsonates
and cacodylates are classed as contact herbicides, which means they don't have
to be absorbed through the roots to be effective.
The arsenical herbicides listed in Volume I of the EPA Compendium of
Registered Pesticides are:
Arsenic Acid (orthoarsenic acid)
Arsenic Trioxide
Arsenous Oxide
Basic Copper Arsenate
Cacodylic Acid
Calcium Acid Methanearsonate
Calcium Arsenate and Tricalcium Arsenate
Calcium Propanearsonate
Diammoniun Methanearsonate
Discdium Methanearsonate and Methanearsonic Acid
Lead Arsenate and Standard Lead Arsenate
Monoscdium Acid Methanearsonate and Monammoniun Methanearsonate
Dcdecylamtonium Methanearsonate and Octylamtnonium Methanearsonate
Sodium Arsenite
The underlined names are the ones for which information has been in-
cluded in the Compendium - the other items are merely listed in the index and
further data will be incorporated at some future time. The existing write-ups
include registered uses, tolerances (in soil and on agricultural products), and
limitations (eg., State prohibitions, part of plant life-cycle when most
effective).
-61-
-------
The Criteria and Evaluation Branch of the Office of Pesticide Programs
has compiled a list of arsenical herbicides and a preliminary list of non-
arsenical alternatives registered for the same uses as the arsenicals. This
compilation and search for non-arsenical alternatives is part of an on-going study
at EPA. The list of alternatives will be trimmed as EPA gathers information
on the economics and characteristics of each alternative, such as cost and
availability of the alternative, efficiency within different climates and on
various soil types, and information on the methods of application (which could
entail a large capital outlay for new equipment, if the alternative material
must be applied differently from the arsenical).
The 1975 Weed Control Manual, included in the February 1975 edition of
Agri-Fieldman, lists the currently available herbicide products and their
manufacturers. The arsenical compounds used in these products are:
Monosodium Acid Methanearsonate (MSMA)
Disodium Msthanearsonate (DSMA)
Cacodylic Acid
Amine Methanearsonate (AMA)
Calcium Acid Methanearsonate
Dodecylammaiium Methanearsonate
Qxtylamnonium Methanearsonate
All of these compounds are listed in the EPA Compendium; only the first
three are listed in the table compiled by the Criteria and Evaluation Branch of
the Office of Pesticide. Programs. Cacodylic acid, DSMA, and MSMA constitute
virtually the: entire organic arsenical herbicide market.
Arsenic acid, because of its high water solubility, is a very potent
herbicide. During the past decade, it has increased dramatically in use as a
cotton plant desiccant; i.e., a harvest aid which is applied prior to machine
picking.
(52)
The Farm Chemicals Handbook ' lists four producers of arsenic acod
used in cotton desiccation. In checking with these producers, it was found that
one no longer produces arsenic acid, and that too produce it for in-house uses
-62-
-------
only, such as the production of lead and calcium arsenates. Two producers,
both located in Texas, supply virtually all of the arsenic acid used in cotton
production. They are the Bryan, Texas, Division of the Pennwalt Company, and
Voluntary Purchasing Group (VPG) of Bonham, Texas.(53)
In the Texas-Oklahoma area, 85 percent of the cotton is "dry land"
cotton which yields 1/4 to 1/2 bale per acre, and 15 percent is spindle cotton
which yields 1 to 2 bales per acre. (Spindle cotton is more expensive to raise
since it requires irrigation and many doses of insecticides.) The dry land
cotton is machine picked ("stripped") at a cost on the order of 10 percent of
the cost of manual picking. Industry spokesmen claim that many cotton growers
would have to go out of business if arsenic acid becomes unavailable for desic-
cation purposes. The non-arsenical alternative is paraquat, but this is
allegedly not as effective as arsenic acid, and it is more expensive per appli-
cation. (53'54>
Demand for arsenic acid is increasing. In 1966, the amount of arsenic
acid used in cotton desiccation was 443 metric tons, while by 1971 it had in-
creased to 2,750 mstric tons/ 9) in 1975, arsenic acid consumption is estimated
to be on the order of 7,460 metric tons.^53^
The following table shows the use-distribution of arsenical herbicides
on crops (1971data).:
.(49)
Organic
Arsenical
(kkq)
Total
Herbicides
(kkq)
As % of
Total
8
3,440
51,200
6.7
•
I
84
5,250
1.6
i
1.4
3,620
i
0.038
1
I
22
16,600
;
i
0.13
|
1
10
0.45
18,600
0.0024
s
3.2
30,500
0.01
a
$
a
10
9,900
0.1
CM
1
0.45
715
0.063
•
1
3,550
167,000
2.1
Note that arsenical herbicides used account for only about 2 percent of total herbicide use
on crops, and that cotton is the roost arsenically-dependent crop.
-62-
-------
Fungicides
The bulk of arsenical fungicides are used as wood preservatives. There
is one arsenical fungicide, however, which is used to control fungus attack of
vinyl plastics. It is lO^O'-oxybisphenoxarsine, marketed as Vinyzene (in about
five different formulations) by the Ventron Corporation of Beverly, Massachusetts.
The Vinyzene formulations contain either 1 or 2 percent of the active
fungicide lO^O'-oxybisphenoxarsine (10,10'-CBPA). The empirical formula for
10,10'-CBPA is C^aPiepSjOy it is 30 percent As and 40 percent As2°3*
Fungicides such as Vonyzene are used in plasticized polyvinylchloride
products which are exposed in use to humid environments. The PVC resin itself is
not normally attacked by fungi; it is the plasticizers which need protection from
fungal attack. Rigid vinyl products such as plumbing pipe do not require fungicide
additives. Typical calendered and flexible vinyl products in which fungicides
might be used include shower curtains, hospital sheeting, upholstery, electrical
cable jackets, refrigerator gaskets, wall coverings, automobile landau tops,
other auto parts, boat covers, awnings and tarpaulins, pond and swimming pool
liners, and plastic, pants for babies.
In the case of the arsenical Vinyzene, the cannon formulations contain
1 percent 10,10'-OBPA, with 2 percent formulations used for very humid outdoor
applications. The active ingredient is normally dispersed in epoxidized soybean
oil, a plasticizer, for use by plastic molders at about 3 parts per hundred of
resin (phr). Dispersions in other plasticizers such as dioctylphthalate or
diisodecylphthalate, and formulations using solvents such as methyl ethyl ketone
or mineral spirits, are produced to a lesser extent.
The use of Vinyzene has grown very rapidly over the past ten years, but
it still accounts for only a very tiny fraction (0.02 percent) of all the white
arsenic used in the U.S.:
-63-
-------
Year
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
Vinyzene Formulations,
kkg/year
30
40
100
180
340
380
600
850
1,030
1,390
10,10'K)BPA,
kkg/jfear
0.3
0.4
1.0
1.8
3.4
3.8
6.0
8.5
10.3
13.9
Arsenic as As203,
kkg/year
0.1
0.2
0.4
0.7
1.3
1.5
2.4
3.3
4.1
5.5
Although Vinyzene has more than 50 percent of the total fungicide (for
plastics) market, the overall market penetration of all fungicides is only about
10 percent. Hence, even if the entire potential market were captured by Vinyzene,
the quantity of 10,10'-ORPA would be only 280 metric tons per year, with an
As20_ equivalent of only 110 metric tons per year.
At the standard use rate of 3 phr for a 1 or 2 percent formulation, the
active ingredient 10,10'-QBPA is in the vinyl product at a concentration of 300
or 600 ppm. The tight structure of the molecule makes it extremely stable, and
it has no disoemable vapor pressure. Skin irritation and sensitivity tests
gave acceptably low results. The active ingredient was not extracted from PVC
film by water, perspiration, or skin oil; and none was volatilized up to 250°C.
Since much of the PVC products in which Vinyzene is an ingredient enters the
municipal solid waste stream, a significant portion is incinerated, whereupon
As203 is emitted to the atmosphere. The quantity involved, however, is extremely
small conpared to other As203 atmospheric emissions.
The large market share of all vinyl fungicides enjoyed by Vinyzene is
in part due to its 10 to 15 percent lower cost than competitive materials. The
-64-
-------
substitutes include barium metaborate and the following organics:
N- (trichloronethylthio) phthalimide
2-n-octyl-4-isothiazolin-3-one
N- (trichlcarcrnethylthio) -4-cyclohexene-l, 2-dicaxboxintide
Triphenylin nonylphenoxide
diphenylstibine 2-ethylhexoate
Wood Preservatives
The American Wood Preservers' Association lists three arsenical wood
preservatives:
Chromated Copper Arsenate (OCA) (Types A, B, and C)
Fluor Chrome Arsenate Phenol (FCAP)
Aitnoniacal Copper Arsenite (ACA)
These compounds are classed as water-borne preservatives. The standards
established by the .fiWPA call for the following compositions (tolerances not given
here):(56>
Chromated Copper Arsenate
Type A Type B Type C
Hexavalent chromium as Cr03 65.6% 35.3% 47.5%
Copper As CuO 18.1% 19.6% 18.5%
Arsenic as As205 16.4% 45.1% 34.0%
-65-
-------
Fluor Chrome Arsenate Phenol
Type B
Fluoride as F 22%
Hexavalent chrcmium as CrO3 37%
Arsenic as As^O.. 25%
-------
Fence posts
Piling
Switchties
Crossarms
Plywood
Arsenical
4%
42%
The arsenicals used in these applications are mainly chromated copper
arsenate and fluor chrome arsenate phenol; anrnoniacal copper arsenite accounts
only for 9 percent of the arsenical usage in plywood preservation, 5 percent of
the lumber and timber, and no other significant usage. Thus, CCA and FCAP are
effectively the only arsenical preservatives listed by the American Wood Pre-
servers' Association, and these two compounds are used only in poles (~2%),
lumber and timbers (45%), fence posts (4%), and plywood (~38%).
Consumption of solid preservatives (as opposed to liquid preservatives
creosote and petroleum - which are the mainstays of the wood preservative indus-
try) has followed a pattern where fluor chrome arsenate phenol is gradually
phasing out, while chromated copper arsenate is growing in use.
(57)
Solid Preservatives
(Wag)
(45)
1965
1966
1967
1968
1969
1970
1971
1972
1973
Chromated
Copper
Arsenate
-
-
1,060
1,460
2,120
2,740
3,960
4,430
5,320
Fluor Chrome
Arsenate Phenol
2,610
3,140
2,430
1,800
2.060
1,220
987
870
767
Pentachlorophenol
Total
Preservatives
9,200
11,800
11,300
12,000
11,600
12,900
14,600
16,600
17,700
14,300
17,600
16,350
17,900
18,050
17,800
20,500
23,200
25,300
-67-
-------
Thus, in 1973, chroraated copper arsenate accounted for 21 percent of
the total solid preservatives consumed and fluor chrome arsenate phenol accounted
for 3 percent.
Looking at the total preservative picture (liquids and solids),
254,443,000 cubic feet of wood ware preserved in 1973, of which chronated copper
arsenate accounted for 12 percent (29,414,000 ft ), fluor chrome arsenate phenol
accounted for 1.4 percent (3,604,000 ft ), and ammoniacal copper arsenite and all
other arsenical wood preservatives (see below) accounted for even less.
The wood preservatives discussed above are used to protect wood against
fungus attack and nticrobially-mBdiated rot. Insects, especially termites, are
also a consideration in wood preservation. The EPA Office of Pesticides Programs
has compiled a list of arsenical insecticides used as wood preservatives. They
are:
Ammonium arsenite
Arsenic acid
Arsenic pentoxide
Arsenic trioxide
Sodium pyroarsenite
Wolman salts (fluor chrome arsenate phenol)
The reason for the absence of CCA (chromated copper arsenate) from this
list is not known. The registered alternatives for these ccmpounds are given as
creosote and pentachlorophenol.
EPA has also conpiled a list of arsenical fungicides used as "industrial
wood preservatives". They are:
Arsenic acid
Arsenic trioxide mixtures
Arsenic pentoxide mixtures
Disodium arsenate mixture
Sodium arsenate mixture
Ammonium arsenite mixtures
-68-
-------
The use of these compounds is extremely limited, if indeed they are
used at all any more.
Feed Additives
The arsenical feed additives are:
Arsanilic acid
Roxarsone (3-nitro-4-hydroxyphenylarsonic acid)
Cafbarsone (p-ureidobenzenearsonic acid)
Nitarsone (4-nitrophenylarsonic acid)
The purposes of arsenical feed additives are disease prevention and
control, and to improve feed efficiency and weight gain. Arsenicals are re-
stricted almost entirely to poultry, though some are used for swine.
Carbonsone is used in turkey feed only; its function is to prevent or
control histomoniasis (blackhead), a protozoan parasite disease of turkeys. It
is sold in combination with colloidal aluminum silicates containing 37.5 percent
Carbarsone (Carb-O-Sep), or in combination with antibiotics. Carbarsone is also
used to prevent and control coccidiosis, a common poultry disease. In combina-
(CQ\
tion with Bacitracin, Carbarsone also increases weight gain in turkeys.
Nitarsone is used to prevent and control blackhead in chicken and tur-
keys. In combination with various antibiotics, Nitarsone will also stimulate
(KO\
growth and improve feed efficiency. ;
RDxarsone promotes growth and improves feed efficiency and pigmentation
in chickens and turkeys. It also increases egg production in laying chickens.
In combination with antibiotics, Roxarsone prevents and controls various chicken
/cq\
and turkey diseases, and promotes growth and improves feed efficiency.
Arsanilic acid (or sodium arsanilate, the water-soluble salt of arsanilic
acid) increases weight gain and improves feed efficiency in chickens and turkeys;
it also increases egg production and feed efficiency in laying chickens and pre-
vents coccidiosis in both layers and nonlayers. In combination with various
antibiotics, arsanilic acid prevents and controls certain diseases, as well as
-69-
-------
/CQ\
improves weight gain and feed efficiency. v
In swine, arsanilic acid (often in its water soluble form - sodium
arsanilate) increases weight gain and feed efficiency. In ccrribination with
antibiotics is prevents, treats, or controls various internal diseases such as
salmonella, and dysentary.
All commercial combinations of antibiotics with arsenical feed additives
must be approved by the FDA. Federal law requires a 5-day withdrawal period be-
fore slaughter for poultry. ' Arsenicals tend to accumulate in poultry
livers, but the 5-day withdrawal period is sufficient for liver levels to return
(59)
to normal, non-arsenical levels.
Names of producers of arsenical feed additives are listed in the Feed
Mditive Compendium. (58)
Nonferrous Arsenical Alloys
Arsenic in small amounts can influence the mechanical and chemical prop-
erties of copper, lead, and brass.
Copper
Arsenic in copper increases corrosion and erosion resistance, raises
annealing temperature, and possibly serves as a deoxidizer. Arsenical copper,
as it is called, contains up to 0.5-percent arsenic. The higher annealing tem-
perature of arsenical copper allows the material to retain its strength after
soldering; thus, automobile radiators and other such copper parts fabricated by
soldering are likely to contain arsenic. ( '
Arsenical copper has also been used in the manufacture of boiler tubes
used in power plants in the central U.S. where water conditions are relatively
mild. This use has, however, diminished in recent years because of high cost
compared with Muntz metal and inhibited admiralty. ( '
The most widely known use of arsenic in lead is probably in the manu-
facture of lead shot. Arsenic in amounts of up to 1 percent alters the charac-
teristics of molten lead so that the shot produced is of a more spherical shape.
-70-
-------
Lead shot is likely to be banned on flyways because of the poisoning of
ducks who eat the shot (lead poisoningl, Iron-lead alloy will likely replace
the lead-arsenic alloy now used. The Bureau of Sport Fisheries and Wildlife
will be the banning agency. An estimated 1/145 kkg of arsenic is used annually
in the production of lead shot. t39'47)
, In engine bearings of the type used in automobiles and trucks, additions
of up to 3 percent arsenic to the usual lead-tin-antimony babbitt improves bear-
ing life significantly; it produces an increase in both strength and endurance
limit, especially at; high engine temperatures, and it probably also inhibits the
bearing corrosion which is common in engines under adverse service conditions.
In the electrolytic deposition of copper from solutions containing chlo-
rides and nitrates, the anodes have been made from lead-antimony alloys contain-
ing from 0.6 to 6.8 percent arsenic. The arsenic reduces the anode solubility.
Lead-acid storage batteries typically contain antimony in the lead as a
hardener, especially for the posts and plates. Arsenic is used in amounts up to
0.5 percent/ also for hardening, and to otherwise extend battery life. Arsenical
lead for batteries is purchased as such from lead suppliers.*• '
Cable sheathing must be strong and corrosion resistant. Chemical lead,
1 percent antimonial lead, and arsenical lead alloys have been used in cable
sheathing made of lead.
Brass
Arsenic in brass inhibits dezincif ication and the resultant season
cracking, corrosion processes whereby zinc dissolves out of brass thus making it
brittle and spongy.
Arsenic usage in nonferrous alloys was estimated to have been 360 kkg
/q\ (18)
in 1968. ' Bureau of Mines data ' show a general increase in metallic ar-
senic imports between 1970 and 1974, and since the main use of metallic arsenic
is in nonferrous alloying, then such consumption could possibly be increasing.
-71-
-------
1970 1971 1972 1973 1974
Metallic As Imports 415 486 600 583 590
(Hog)
No metallic arsenic was produced in the U.S. during this period.
Glass
Arsenic is not widely used in glass any more, and when it is, it is
classed as a "minor ingredient" - i.e., measured on the order of ounces per ton.
Arsenic in glass serves two manufacturing purposes: (1) as an oxidizing
agent, and (2) as a fining agent. As an oxidizing agent, arsenic in the form of
sodium arsenate oxidizes iron (FeO oxidized to Fe-OJ so that it will not discolor
the glass. This oxidizing operation can also be performed using non-arsenical
sulfate or sodium nitrate (niter) - in fact, in most instances these days, the
non-arsenical alternatives are used, especially since they are efficient in the
oxidation of other possible impurities such as carbon (which can be oxidized by
niter).'62'
Fining is the removal of bubbles. As a fining agent, arsenic trioxide
and niter are mixed into the glass. The arsenic trioxide is oxidized to arsenic
pentoxide which, by thermal decomposition, releases oxygen bubbles which rise to
the surface carrying with them bubbles of other gases in the glass. Non-arsenical
i(ij\
sulf ates are mostly used for fining these days.
Sulfates are used for oxidizing and fining to such an extent these days
that arsenic use has dwindled to almost nothing. The types of glass where arsenic
would most likely be used (for fining and oxidizing) are flat glass, container
glass, and "art glass". In virtually all forms of these glasses, however, non-
arsenical sulfates are likely used nowadays, and in the cases where arsenic is
(62)
still used, a possible alternative (which has been used) is antimony oxide.v '
-72-
-------
The oxidation of Fe in glass is one step in the "decolorizing" process.
The steps are:
(1) Magnetic removal of iron impurities from the components to
be used in the glass
(2) Oxidation of the remaining iron (generally in the form of
FeO, which causes a bluish color) to ferric oxide (which
produces a less objectionable tan color)
(3) Masking of remaining color by the addition of cobalt or
seleniun which complements the objectionable colors
The second step (oxidation) is where arsenic has been used and is still
f 62}
used to a limited extent.
Special glasses having high infrared transmissibility sometimes contain
arsenic as a component; such glasses are used in infrared cameras and in night-
sighting reconnaissance systems. Infrared spectrometers used in such applica-
/OQ\
tions as nondestructive testing of plastics contain arsenic trisulfide.v ' In-
(42)
frared lasers also use arsenic trisulfide in their glass components.v '
Gallium arsenide has been used as window film and has also been con-
sidered for use in bulk form in windows for high-powered lasers; boron arsenide
has found a possible application in the same area, but difficulty has been en-
countered in growing crystals of sufficient size. Combination of arsenic with
tellurium, germanium, iodine, selenium, thallium, and sulfur has been used in
(42)
specialty glasses having low melting-point properties.v '
A note of interest is that the government ordered that arsenic not be
used in f luorosilicate glass (of which Pyrex is the most well-known example)
during World War II because the arsenic was needed elsewhere in the war effort.
The producers of f luorosilicate glass didn't think it would be possible to comply,
(62)
but they did, using the alternative oxidizers and fining agents discussed above.
Electronics
Gallium arsenide was once considered as a potential replacement for
silicon semiconductors, but silicon devices are currently favored because they
-73-
-------
are easier to fabricate and there is less hazard in materials handling. Arsenic
is also used as a dopant for silicon materials, but only for certain special
semiconductor properties; boron and phosphorus are tlie dopants of choice. Gallium
arsenide semiconductors are preferred to silicon types in high-temperature
(A?)
conditions.v '
Light Emitting Diodes (LEDs) rely heavily on arsenic intermetallic com-
pounds. LEDs are used commonly in the latest generation of calculators; the
diodes are arranged for digital readout/ and their low power requirements make
them ideal for battery-powered applications. Gallium arsenide, GaAs, and gallium
arsenide phosphide, GaAs P , are the most commonly used, though indium arsenide
x v (42)
and indium arsenide phosphide are used in some devices.
Animal Dips
Sodium arsenite and potassium arsenite were the arsenicals used in dips
for cattle. These compounds are now available only in laboratory-sized lots -
no production quantities are available from American chemical manufacturers.
There are no arsenical dips used any longer. Chemical suppliers and dip manu-
facturers have stopped production because of the risks and problems in the
handling of inorganic arsenic.
Arsenical dips have been replaced by formulations of Ooumaphos or
Ttaxaphene.
Arsenic in Paint
Arsenic compounds have been used in paints both a pigments and as anti-
fbuling agents (marine uses). These uses, however, are in rapid decline - if,
indeed, they exist at all any more.
According to the National Paint and Coatings Association - which repre-
sents 70 to 75 percent of the paint manufacturers and 90 to 95 percent of the
paint sales - arsenic is no longer used in paint, either as a pigment or as an
antifouling agent. The Marine Coatings Committee of NPCA says that arsenic is
definitely not used as an antifouling agent in paints at this time.
-74-
-------
Note on Inorganic Arsenic Production
Many of the large suppliers of inorganic arsenicals - specifically
potassium and sodium arsenite - have ceased this facet of their operations over
the past several years. In telephone conversations 'with such chemical suppliers
as Allied Chemical and Chipman Division (Rhodia), it was learned that EPA and
OSHA rulings on registration and testing have made it unprofitable to handle
arsenical compounds, especially since the demand for most inorganic arsenicals
has been continuously decreasing over the past taro decades.
Environmental Emissions Resulting from Arsenic Uses
Ferguson and Gavis estimate the average annual arsenic contribution by
man to the environment (worldwide) is about 100,000 kkg/yr; this includes the
amount which results from increased erosion processes resulting from excavation
and mining operations. The total cultural contribution is believed to be on the
order of 3 times the natural arsenic flow due to natural erosion processes. A
significant portion of this - between 15 and 20 percent - results from the uses
of arsenic.
Pesticides
Pesticides (.insecticides, herbicides, and fungicides) are the largest
single use of arsenic, and because pesticides are deliberately introduced into
the environment, they account for about 80 to 90 percent of all arsenic emissions
(onto land and into water and air) resulting from the uses of arsenic. In 1974,
about 13,000 kkg of arsenic was consumed by pesticide manufacturers in the United
(39)
States.v ' This is therefore the limiting amount of arsenic in pesticides that
could reach the environment in 1974 (assuming no decrease in pesticide stockpiles
for that year). Emissions not only result from the uses of pesticides, but from
the manufacture of pesticides and from the disposal of such commodities as cotton
gin trash and other agricultural wastes containing arsenical pesticide residues.
The SPA publication, Emission Factors for Trace Substances, lists
the emission factor for the production of pesticides as 10 kg As/kkg of arsenic
processed. This emission results from handling losses of arsenic trioxide as
-75-
-------
it is transferred into reactor vessels. The 13,000 kkg of arsenic shipped to
pesticide producers in 1974 corresponds to a total emission of 130 kkg.
Pesticidal emissions occur during the application of sprays and dusts,
during the incineration of pesticide containers and agricultural waste, and as
a result of evaporation processes. The emission factor given in the National
(9)
Inventory of Sources and Emissions is 168 kg As/kkg of arsenic applied as
pesticide. (The remainder of the applied pesticidal arsenic is assumed to be-
come firmly bound into the soil matrix.) Pesticidal emissions to the atmosphere
resulting from the actual use of pesticides was on the order of 2184 kkg As in
1974.
t
Thus, in 1974, the latest year for which information is available, the
total atmospheric emissions due to pesticide use was about 2300 kkg. The re-
mainder of the 13,000 kkg used in pesticides in 1974 became either locked into
insoluble solid systems (and is effectively removed from the environment) or
found its way into natural water systems; there is no information available,
however, upon which to base estimates of the portion of pesticidal arsenic which
moves from the land into water systems; the range is probably on the order of
2 to 20 percent of the volume of pesticide used, i.e., between 260 and 2600
kkg.
Glass Manufacture
The EPA source gives the emission factor for arsenic in glass pro-
duction as 0.08 kg/kkg of glass produced; however, this is based on 1968 data,
and since then the amount of arsenic used in glass production has decreased.
The factor cited in the National Inventory is given in terms of the amount of
arsenic used - 116 kg/kkg of arsenic used. The amount of arsenic used in glass
(39)
in 1974 was 1805 kkg. Thus, total atmospheric emissions in 1974 were on the
order of 210 kkg. It is unlikely that arsenic in glass would find its way into
the environment since it would be firmly fixed into the glass, except possibly
as a consequence of recycling operations, but data on the amounts of arsenical
glass recycled are not available.
-76-
-------
Wood Preservatives
Atmospheric emissions resulting from the manufacture of wood preserva-
tives is considered negligible. With regard to the uses of arsenically-preserved
wood, however, it is reasonable to assume that after a sufficient period of tiitie
(decades and, in some applications, centuries) deterioration of the wood would
release the arsenic to the environment. Such release would be very slow since
the preservative compounds bind tightly to the wood fibers. The amount of
arsenic moving into the environment by this method is too slow to pose a pollution
hazard to air, water, or soil, and at the expected slow rate of release, con-
centration of soluble arsenic in adjacent soil and water would be low enough for
the arsenic to become readily bound into insoluble species in soils and sediments.
Feed Additives
The pollution potential of arsenical feed additives is-similar to that
of pesticides in that arsenic is lost to the environment during both manufacture
and as a result of use; the excreta of arsenically-fed animals is used as ferti-
lizer, and since the arsenic in the feed additives passes through the animals in
virtually the same amounts in which it is ingested, the arsenic eventually finds
its way to the land where it undergoes the same processes which act upon the
arsenical pesticides. However, whereas with pesticides a large portion of at-
mospheric emissions during application results from dusting, misdirected spray,
volatilization, and so on, animal wastes are not subject to these mechanisms, and
the arsenic contained in the excreta finds its way to the soil where it becomes
bound into either insoluble soil complexes or, if in a soluble form, is carried
into surface and ground water supplies. Animal excreta as a source of arsenic
pollution is negligible, however, as has been shown in one study where no in-
crease was found "in soil, water or forage after poultry litter containing from
15 to 20 ppm arsenic had been applied to land at a rate of 4 to 6 tons per acre
(59)
per year for 20 years".
With regard to the manufacture of feed additives, no emissions data are
available. Using half the emission factor for the manufacture of pesticides —
i.e., 5 kg As/kkg of arsenic used — the 409 kkg of arsenic used in feed additives
in 1974 would have resulted in an atmospheric release of 2.04 kkg.
-77-
-------
^tor^fgyTOUfl Alloying
The National Inventory of Sources and Emissions places the total atmos-
pheric emissions of arsenic due to nonferrous alloying in 1968 at about 1/4 kkg.
This is a negligible amount. In 1974 the amount of arsenic used in nonferrous
alloying was on the order of three times the amount used in 1968, thus, the total
emission would be on the order of 3/4 kkg, still a negligible amount.
A potential hazard might exist, however, in work environments where
arsenical metals are melted or joined by fusion, as in the so-called "burning
stations" in battery factories. Lead parts of batteries are fused together using
a natural-gas flame for heating (a hand-performed process), and overheating could
lead to the production of arsenic fumes. Specific data is not available.
Miscellaneous
The emissions factor for the miscellaneous uses of arsenic is given
collectively as 2 kg/kkg of arsenic processed. ' This factor is based on data
for 1968, and different minor uses of arsenic prevail today. For example,
arsenical animal dips and paint pigments and additives are no longer used, and
the amount of inorganic arsenic used in such applications as leather tanning and
non-feed-additive Pharmaceuticals is on the decline - if still used at all.
Electronics is probably the largest consumer of arsenic in the miscellaneous
category, but emissions data for this use is lacking. Assuming that the major
atmospheric emissions occur during handling and that the above emissions factor
applies, then the 317 to 330 kkg of arsenic in miscellaneous uses in 1974 would
result in an emission of about 0.65 kkg.
Non-Arsenical Alternatives
Pesticides
All pesticides are registered with EPA for use against specific pests
in specific situations (e.g., with specific crops or industrial uses). EPA's
Office of Pesticide Programs currently has a project underway to find alternatives
to arsenical pesticides. The problem is a difficult one since the factors to be
considered include cost of alternatives, method and cost of application, soil pH,
-78-
-------
regional climate, status of alternative (is it about to be banned or restricted?);
availability of alternative, and possible side effects of alternatives on other-
wise beneficial species.
An example of the magnitude of this problem of specifying alternatives
is the case of lead arsenate as applied to apples. There are 15 apple pests
which lead arsenate is registered to control. Of these, the number of registered
alternatives varies from none (applethorn skeletonizer, case bearers, and others)
to sixteen alternatives in the case of the codling moth. In addition to apples,
ten other fruit crops are registered, along with associated pests and the regis-
tered alternative for each pest. Vegetables and non-crop uses must also be
considered.
With regard to the vinyl fungicide 10,10'-OBPA, nonarsenical alternatives
are available, but there is some question as to their relative effectiveness.
Wood Preservatives
The main alternatives to arsenical wood preservatives are creosote and
pentachlorophenol. These alternatives, being oils, are not adequate substitutes
in applications where aesthetics, discoloration, odor, or suitability for painting
are important. This limitation dictated by the intended use is the major reason
for the extensive use of arsenicals in lumber, timbers, and plywood; as opposed
to the use of the preservative oils for poles, crossties, piling, etc. There
are non-arsenical and non-oil-base alternatives: ACC (acid copper chromate) and
CZC (chrcmated zinc chloride). However, CZC is not recommended for use where
soil or water contact is encountered; and the health hazards from these alterna-
tives may be equal to or worse than from the arsenical wood preservatives.
The Office of Pesticide Programs has also compiled a tentative list of
registered alternatives to the arsenical fungicides used as "industrial wood
preservatives". The list is tentative because (1) some of tlie substitutes are of
limited availability or are only in limited use, (2) economic factors have not
been taken into account, and (3) the alternatives might be more hazardous than
the arsenicals.
-79-
-------
Glass
Cerium oxide can be used in place of arsenic trioxide as a glass de-
colorizer and fining agent. Sulfates and nitrates nay be used (and are being
used) in place of arsenicals in fining, though for some glasses (as in table-
(42)
ware and TV tubes) the alternatives do not give as satisfactory a result.
With respect to laser windows, alternatives are available, e.g.,
Ge2gSb,2Sego, Ge28^D12T360' anc* t^ie alkali halide materials (e.g., KC1) can
replace Ge_3As,2Se5c- However, there are no alternatives for the unique prop-
erties afforded by arsenic trisulfide in certain glasses used as infrared lenses
(42)
and windows.v '
Feed Additives
Alternatives to feed additives used to improve weight gain and feed
efficiency are hormones and antibiotics. Hormones, however, are restricted
mainly to cattle because the withdrawal period prior to slaughter is too long
to make the use of honrones profitable in poultry or swine.
Low level use of antibiotics in poultry feed will improve weight gain
(2- to 3t-percent dinprovement) and feed efficiency - though generally not .to the
sane extent as will antibiotics in combination with arsenicals. Antibiotics
will also prevent and control the same diseases prevented and controlled by
arsenicals. The controversy and possible hazard of long-term use of antibiotic
feed additives centers on the potential development of resistant strains of patho-
gens. It is possible for resistance factors to be transmitted between bacterial
species; it is even considered possible that a nonpathogenic bacteria could pass
a resistant gene to a pathogenic strain of the same bacterial type - the resultant
pathogenic and resistant strain, if affecting man,,would then not be amenable
to treatment using the antibiotic to which the resistance had been developed.
OSiis problem is being studied by the FDA, and there is a movement to restrict
the.use of antibiotics in feed additives to those antibiotics which would not
normally be used to treat diseases in people.' ' '
Since hormones are not and cannot be economically used in poultry (the
withdrawal period for hormones is too long to be of value in chickens — they go
-80-
-------
to market only 7 weeks after hatching), and since antibiotics have potential
drawbacks, alternatives to arsenical feed additives are not readily available.
In discussing the disadvantages of arsenical feed additives with a
feed-additive specialist at FDA, the chief disadvantage cited was the 5-day
withdrawal period required before slaughter - which required the producer to
take the "positive action" of formulating a different feed mixture for those
animals about to be slaughtered, as opposed to those animals for which slaughter
is weeks away; arsenical feed additives are, for all practical purposes, not
stored in muscle tissue.
An article in the British Medical Journal cites arsenical feed additives
as an alternative to antibiotics in the control of piglet scour and turkey poult
morbidity. Antibiotics are said to have "practical difficulties, apart from
the risk of drug resistance". The use of arsenical feed additives is seen as
an impetus to new studies of arsenic pharmacology. Arsenic in the livers of
arsenically-fed pigs is supposedly so low that one would have to eat 110 Ibs.
of pork liver per day to consume a dangerous level of arsenic, though such a diet
would present more than just a potential, arsenic problem.^ '
Nonferrous Alloying
For most applications of arsenic in lead, copper, and brass, similar
properties could be supplied by other materials, though increased cost would
likely be a concern. In the case of lead-acid batteries, an industry spokesman
stated that there are no known alternatives to arsenical lead in batteries at
this time.(61)
Electronics
Light Emitting Diodes (LEDs) are based on gallium arsenide compounds
for which there are no alternatives. However, LEDs themselves could be replaced
by gas-discharge displays, incandenscent bulbs, and (soon to be available) liquid
crystal displays - none of which will supply the complete complement of advantages
of LEDs such as ruggedness, low power requirement, high visibility, and com-
(42)
patability with semiconductor circuitry.
-81-
-------
Other arsenical semiconductor uses are based upon the specific electri-
cal and chemical properties of arsenic, and to the extent that these properties
(42)
can be conpromised, alternatives are available.v
-82-
-------
SECTION V
INDUSTRIAL SOURCES OF ARSENIC MOBILIZATION
The natural occurrence of arsenic in the earth's crust is 2 to 5 ppm, but
higher-than-average concentrations occur in metallic ores of two types: in sulfide
deposits (associated with copper, lead, zinc, and other ores); and in sedimentary
deposits such as iron ore, phosphate rock, borax ore, manganese ore, and fossil
fuels.
The very high temperatures associated with smelting of metallic ores
generally result in the release of a large portion of the naturally-occurring
arsenic to the atmosphere. Both elemental arsenic and its common oxide, AsJX,
are extremely volatile materials at common smelting temperatures: arsenic
sublimes at 613°C; As203 sublimes appreciably at 135 °C and fully at 315 °C.
Three other factors, aside from the inherent volatility of As2°3' con~
tribute to the generally high losses of this material to the atmosphere:
1. The As2^3 is slow to condense as higher-temperature flue gases
are cooled; a very long time is required for nucleation and
growth of the particles. This phenomenon is the reason why
the commercial process for As203 manufacture includes several
high-volune condensation chambers (called kitchens) arranged in
series at successively lower temperatures from 220°C in the
first to 100°C or less in the last. <2'22'24/37'38) T^ 33^
technology is used in the commercial manufacture of P~0_,
(77)
only the "kitchens" are called "barns". ' Hence, As2°3 may
very well pass through baghouses and electrostatic precipitators
as a supersaturated vapor even if the temperature is below the
equilibrium sublimation temperature.
2. Dust collection devices such as electrostatic precipitators
and baghouses are routinely operated at elevated temperatures
so as to stay well above the dew point of the flue gas. For
many metallurgical operations such as roasting and sintering,
-83-
-------
the flue gases have a very high moisture content, necessitating
high dust collection temperatures.
3. It is common practice in the nonferrous metals industry to re-
cycle collected flue dusts to the process until concentrations
of valuable metals build up to an ecanaiacally-processable level.
At each stage of recycle, the very volatile As-jCL has another
opportunity to escape collection.
Each segment of the nonferrous metals industry will be examined to
determine the quantities of arsenic involved in the industry and the fate of
this arsenic. The quantities of arsenic in comrnercially-developed sedimentary
deposits will also be investigated.
The Primary Zinc Industry
Table 2 lists the arsenic content of representative zinc concentrates
(comparable data for domestic concentrates were not available). The average
ratio of arsenic to zinc in these concentrates is about 1,050 ppm.
The quantity of zinc produced from concentrates in the U.S. for the past
several years is listed in Table 3. The domestic slab zinc production has been
decreasing in a rather dramatic fashion, primarily because foreign consumption
of zinc has grown rapidly. Since the U.S. always had to airport ore concentrates
the domestic competitive position deteriorated as its share of demand decreased
and as foreign metal production capacity increased. Other major factors contri-
buted to the decline in U.S. zinc production. Older pyrcrnetallurgical plants,
especially horizontal retort plants, are closing because they are labor-intensive,
because they have severe air pollution problems, and because they cannot manu-
facture the high grades of zinc. Only two small U.S. horizontal retort plants
are still in operation, and account for only 13 percent of the total U.S. pro-
duction capacity of 689,000 kkg/fyear. The two large pyrometallurgical plants
(one vertical retort plant and one electrothermal plant) account for 48 percent
of the U.S. capacity, and the three electrolytic plants account for the remaining
39 percent.
-84-
-------
Table 2
Arsenic Content of Zinc Concentrates
Source of
Concentrate
Broken Hill, Australia
Broken Hill, Australia
Broken Hill, Australia
Broken Hill, Australia
Valleyfield, Quebec
Cartagena, Spain
Cerro de Pasca, Peru
Mt. Isa, Australia
Arsenic Content,
ppn
500
700
1,170
610
350
200
640
350
Zinc Content,
Percent
51.0
53.7
52.1
52.9
52.9
49.3
59.2
50.4
As/Zn,
ppn
980
1,300
2,240
1,150
660
410
1,080
700
Reference
67c
67c
67c
67c
67b
67a
74a
74a
-85-
-------
Table 3
Primary Zinc Refined in the U.S., Metric Tons/Year
Sources: Bureau of Mines
Year
1968
1969
1970
1971
1972
1973
1974
From Domestic
Concentrates
452,000
416,000
366,000
366,000
363,000
311,000
281,000
From Imported
Concentrates
473,000
527,000
429,000
329,000
211,000
180,000
209,000
Total
925,000
943,000
795,000
695,000
574,000
491,000
490,000
-86-
-------
Ttoo new electrolytic plants are either planned or being cons true ted, and
the two remaining horizontal retort plants are being phased out. It appears that
the decline in domestic zinc production has been halted and possibly reversed.
For the purposes of this study, the 1973-74 production level of 490,000 kkg/year
will be used, with 290,000 kkg/year pyrometallurgical and 200,000 kkg/year
electrolytic. Based upon this level, the zinc concentrates processed annually
in the U.S. contain 520 kkg of arsenic. Refined zinc of all commercial grades,
and commercial zinc oxide (either French or American process) , contain no
appreciable (greater than 10 ppm) arsenic, ''e' equivalent to less than 5
metric tons per year.
In the primary zinc process, the arsenic in the zinc concentrate is largely
retained through the roasting process, as indicated by data from both a multiple-
hearth furnace and a fluid-bed furnace. In pyrometallurgical zinc smelters,
all of the arsenic remaining in the calcine (frcm the roaster) is volatilized in
the sintering operation, with large losses to the atmosphere. At one sintering
plant, the dusts collected in a baghouse contain 15.0 percent As2°3* Ihese
dusts are then processed for cadmium recovery; one route involving burning of the
dusts which volatilizes more of the arsenic. Other routes to cadmium recovery
from flue dusts involve oxidative leaching, in which ferrous sulfate is added to
precipitate the arsenic as ferric arsenate. This residue, normally disposed of.
on land, amounts to 1.8 kg per metric ton of pyrometallurgical zinc produced, or
520 kkg/year.(75)
Davis reported on air emission factor from pyrometallurgical zinc smelters
(9)
of 0.65 kg of arsenic per kkg of zinc produced. ' Based upon a pyrometallurgical
zinc production level of 290,000 kkg/year, the arsenic air emissions amount to
190 kkg/year. Since the zinc concentrates processed pyrometcJ.lurgically originally
contained 310 kkg/year of arsenic, the difference of 120 kkg/year may be assumed to
be in the residues sent to disposal. Although other solid wastes from the primary
zinc industry amount to 1.50 kkg per kfcg of zinc produced, there should be no
appreciable arsenic in either the acid plant sludge (which arises from the roasting
operation, upstream of where the arsenic is volatilized) or in residues from re-
torting and ZnO production (downstream of where the arsenic is volatilized).
-87-
-------
In electrolytic zinc refineries, the arsenic in the calcine (the roasted
zinc concentrate) ends up in the residue from the acid leaching operation (which
dissolves the zinc). This residue amounts to 360 kg per metric ton of zinc pro-
duced; it contains significant quantities of lead, copper and cadmium and is
shipped to a lead smelter. <68'69a'75) This residue, amounting to 72,000 kkg/year
(dry basis), contains virtually all of the 210 kkg/year of arsenic originally in
the zinc concentrates which are refined electrolytically, plus 8,100 kkg/year of
lead (based upon a ratio of lead to zinc of 0.0165 in zinc concentrates).
The arsenic found in wastewaters from primary zinc refining are summarized
in Table 4. The resulting recommended effluent limations (30-day averages)
-4
were 8.0 x 10 kg of arsenic per metric ton of zinc produced (1977) and
—4
5.4 x 10 kg of arsenic per metric ton of zinc produced (1983). These values are
equivalent to 0.4 kkg/year (1977) and 0.3 kkg/year (1983) of arsenic in waste-
water effluents. No special control and treatment is required for arsenic,
over and above standard water use minimization and segregation and lime-and-
settle treatments; and no control and treatment costs are directly attributable
to arsenic removal.
In summary, the distribution of the arsenic originally in the zinc con-
centrates is as follows:
Loss to atmosphere, 190 kkg/year
Retained in zinc products, 5 kkg/year
In land-destined wastes, 120 kkg/year
In wastewater effluents, 0.4 kkg/year
In residues shipped to lead smelters, 210 kkg/year
Total 525 kkg/year
The Primary Lead Industry
As Table 5 indicates, the arsenic content of representative lead con-
centrates varies from about 600 ppm to 1,500 ppm (comparable data for domestic
-88-
-------
Table 4
Arsenic in Wastewaters from Zinc Smelting
Source: EPA
Plant
B
D
F
H
B
B
B
B
Contributing Operations
Roasting and Electrolysis
Roasting, Leaching, Electrolysis,
Casting
Pyrolytic Smelting
Horizontal Retort
Acid Plant
Metal Casting Cooling
Auxiliary Metal Reclamation
Auxiliary Metal Reclamation
Arsenic,
kg/kkg Zinc
< 0.0001
0.01
0.0002
0.000004
0.003
<0. 00008
0.000017
0.011
-89-
-------
Table 5
Arsenic Content of Lead Concentrates
O
Source of
Concentrate
Broken Hill, Australia
Broken Hill, Australia
Broken Hill, Australia
Broken Hill, Australia
Cerro de Pasca, Peru
Casapalca, Peru
Boliden, Sweden
.
Arsenic Content,
ppn
1,530
1,200
1,111
570
600
800
800
Lead Content,
Percent
70.1
74.0
' 75.9
75.8
43.7
61.7
74.9
As/Pb,
ppn
2,180
1,620
1,470
750
1,370
1,300
1,070
Reference
67c
67c j
67c
67c
7-to
74b
74c
-------
concentrates were not available). The average ratio of arsenic to lead in these
concentrates approximately 1,400 ppm, will be used to estimate the overall
quantities of arsenic contained in lead concentrates.
Table 6 lists the quantity of lead produced from concentrates in the United
States for the past several years. The primary lead produced since 1970 has
been relatively stable at about 610,000 kkg/year, implying that the quantity of
arsenic in these concentrates is 850 metric tons per year. An additional 210
metric tons of arsenic per year enters the primary lead industry via residues
from the electrolytic zinc industry, so that the estimated total quantity of
arsenic entering the lead industry is 1,060 metric tons per year.
The refined lead product has specifications (ASTM B29-55) which limit the
total of arsenic, antimony, and tin to 20 ppm for undesilverized lead; and to
50 ppm for desilverized lead. Hence, the final refined lead contains no more
than about 20 kkg/year of arsenic. Hence, virtually all of the arsenic is re-
moved in the smelting and refining process, in one or more of the following forms:
1. A constituent in slags or sludges (as an arsenate)
2. A constituent in collected dusts and fumes (as As2O.J
3. An air emission pollutant (as As20_)
In the smelting of lead concentrates, some arsenic is volatilized in a
sintering operation, and some is removed via the slag from the lead blast furnace;
but much of the arsenic remains with the lead in the base bullion product from
the blast furnace.(68)
The base bullion passes through a drossing operation for copper removal.
A subsequent oxidation process with a fluxing agent (called a "softening" opera-
tion) removes the arsenic as well as antimony, tin, and residual copper from the
bullion as a calcium or sodium arsenate in a slag layer. The blast furnace and
lead refinery slags are sent to a zinc fuming furnace, but the stable arsenates
remain with iiie slag. Small quantities of arsenic remaining in the softened
lead are removed either via fire-refining (as a fume or a slag) or via electro-
lytic refining (as a sludge).
-91-
-------
Table 6
Primary Lead Refined in the U.S.^ Metric Tons/year
Sources: Bureau of Mines '
Year
1968
1969
1970
1971
1972
1973
1974
From Domestic
Concentrates
316,000
465,000
479,000
519,000
537,000
532,000
526,000
From Imported
Concentrates
107,000
113,000
126,000
70,000
93,000
92,000
82,000
Total
423,000
578,000
605,000
589,000
630,000
624,000
608,000
-92-
-------
Davis and Anderson , based upon material balance data, reported on
emission factor for lead smelters of 0.4 kg per metric ton of lead. Based
upon an annual lead production of 610,000 metric tons, the arsenic lost to the
atmosphere is 240 kkg/year.
Arsenic was not found in any appreciable quantity in the wastewaters from
primary lead smelters; the slag granulation operation has a closed water loop,
and sludges from wet scrubbers are lime-treated (precipitating the arsenic) and
settled prior to discharge.(69b'75)
The solid wastes per metric ton of lead product amount to 410 kg of slag
plus 40 kg of settled sludges (dry basis). Based upon a production level of
610,000 kkg/yr; the solid wastes amount to 274,000 kkg/year. The remainder of
the arsenic entering the lead industry, less the losses to the air and the quantity
retained in lead products, amounts to 800 metric tons per year and reports in the
solid wastes from the lead industry. An average concentration of arsenic in these
wastes of 0.29 percent is 'iinplied from this analysis; it compares favorably with
two separate values, both 0.2 percent, for the arsenic content of lead blast
furnace slag.(74b'c>
In summary, the distribution of the arsenic originally in the lead con-
centrates (850 kkg/year) and in residues from the zinc industry (210 kkg/year)
is as follows:
Loss to atmosphere, 240 kkg/year
Retained in refined lead, 20 kkg/year
In land-destined wastes, 800 kkg/year
Total 1,060 kkg/year
The Primary Copper Industry
Table 7 lists the quantities of domestic copper ore, and the copper in ore
concentrates processed in the United States during the past several years; the
primary domestic copper production has averaged about 1.60 million metric tons
per year. Of interest is the average copper content of ores; first, the con-
centration is low compared to most other metallic minerals; and second, the
concentration is decreasing with time (i.e., poorer ores are being mined as time
-93-
-------
Table 7
Primary Copper in the U.S.
CT8.70)
Sources: Bureau of Mines
Year
1968
1969
1970
1971
1972
1973
1974
Domestic Ores Mined
Ore, kkg/yr
154,200,000
202,900,000
233,800,000
220,100,000
242,000,000
263,100,000
-
% Cu in Ore
0.60
0.60
0.59
0.55
0.55
0.53
-
Copper from Ore Concentrates, kkg/yr
From Domestic
Concentrates
1,054,000
1,331,000
1,380,000
1,280,000
1,524,000
1,559,000
1,440,000
From Foreign
Concentrates
251,000
248,000
221,000
164,000
175,000
135,000
30,000
Total
1,305,000
1,579,000
1,601,000
1,444,000
1,699,000
1,694,000
1,470,000
-94-
-------
progresses). This latter trend has been continuing for quite some time; ores
mined in 1900 averaged 4 percent copper. The first U.S. porphyry ores mined in
the 1905-1915 period had 2 percent copper. The average copper content of domestic
ores in 1950 was 1 percent, and it is projected that the grade will decline to
/7g\
0.25 percent by the year 2,000. '
The arsenic content of copper ores and concentrates is highly variable.
Ores from New Mexico and Arizona have much lower arsenic concentrations than ores
from Montana. Data is extremely sparse, especially in recent years. A circa 1913
copper ore from Butte, Montana, contained 3.25 percent copper and 0.37 percent
/2i \
arsenic (a ratio of As/Cu of 0.114).v ' More recently, Butte ores have contained
0.6 percent copper and 0.1 percent arsenic (As/Cu = 0.17);*30'79' and Butte ore
(79) (20)
concentrates have contained 26 percent copper and 1.6 percent arsenic
for a ratio of As/Cu = 0.062. In 1963, a Colorado copper ore had a ratio of
(2)
As/Cu of 0.0028. Also, in 1963, a copper concentrate from Highland Valley
British Columbia, assayed 41.54 percent copper and 0.012 percent arsenic (As/Cu =
3ni«
(2)
(2)
As/Cu of 0.0028. Also, in 1963, a copper concentrate from Highland Valley,
Coin
0.0029). In northern Chile, copper ore assayed 0.054 percent arsenic, and
the copper concentrats contained 1.64 percent arsenic (As/Cu~= 0.06).
The U.S. Bureau of Mines bases its estimate of domestic arsenic reserves,
1.72 million metric tons, upon its estimate of domestic copper reserves, 77.6
million metric tons. ' An inferred ratio of arsenic to copper in copper ores
and concentrates is therefore 0.022. While the data for specific ores, quoted in
the previous paragraph, have As/Cu ratios highly divergent from 0.022, this value
will be used for the purposes of this study. Based upon a primary copper pro-
duction level of 1.60 million metric tons per year, it is estimated that 35,000
metric tons per year of arsenic accompany the copper concentrates to the smelters.
Refined copper is manufactured to very stringent purity specifications,
since small quantities of impurities adversely affect its electrical and mechanical
properties. Electrolytic copper has a specification (ASTM B224) for 0.01 percent
maximum impurities other than oxygen? the specifications for deoxidized copper
and oxygen-free copper are equally demanding. Normally, electrolytically-refined
copper contains arsenic at levels reported as 1 to 10 ppm, and 4 to 11 ppm.' '
-95-
-------
Of the total copper production of 1.60 million kkg/year, approximately 1.46
million kkg/fyear is electrolytic; at an average arsenic concentration of
5 ppm, the quantity of arsenic in the product copper is about 7 kkg/year.
Another 80,000 kkg/year is fire-refined casting copper, with a specifica-
tion of 75 ppm maximum arsenic content, implying a maximum quantity of 6
kkg/year of arsenic. The remaining 60,000 kkg/year of copper is lake (elemental)
copper, not derived from concentrates. There are three grades of lake
copper: Prime, which contains 25 ppm arsenic; Natural, which contains 200 to 600
ppm arsenic; and Arsenical, with 600 to 5,000 ppm arsenic (primarily used in
(83)
making arsenical copper alloys). At an average of 500 ppm, the total quantity
of arsenic in lake copper would be 30 kkg/year.
Of the total arsenic in copper concentrates, 35,000 kkg/yr, only about 13
kkg/yr remains in the refined copper, the remainder being removed in the smelting
and refining operations upon copper concentrates (which contain between 15 and 35
percent copper). Roasting of copper concentrates is an optional first step.
Older plants built in the 1930's incorporated roasting since ore concentrators
at that time were unable to reach a sufficiently low level of iron sulfide; recent
advances in separation technology have made the overall sulfur content of the
concentrate low enough to bypass the roasting operation. An additional important
factor is the arsenic and antimony content of the concentrate; roasting is often
required for their partial removal prior to smelting, ' °' '
Roasting is either accomplished in the older multiple-hearth units or in
(68)
fluidized beds, at temperatures approaching 1,000°C. A significant portion of
the arsenic is driven off; in a 1913 Anaconda roaster flue gas where the S02
(plus S0_) concentration was 2.82 percent, the As-0- concentration was 0.0073
. <87)
Either roasted or unroasted concentrates are smelted in reverberatory
furnaces or blast furnaces at 1,100 to 1,650°C. ' The products are matte
(a copper and iron sulfide material, containing approximately 30 percent copper),
slag (oxides of iron, silicon, calcium, and aluminum), and S02-bearing flue gas.
Of the total sulfur content of the concentrate (nominally 31.5 percent S and
27.5 percent Cu), up to 20 percent is liberated during smelting in a reverbera-
-96-
-------
tory furnace. the 1913 Anaconda data for the reverberatory flue gas was 0.427
(87)
percent S02 (plus S03) and 0.0156 percent As2°3* ^ one o°PPer blast fur-
nace operation, the dusts collected by a cyclone and then by a baghouse (down-
stream of the cyclone) are:
Cyclone Dusts
Baghouse Dusts
Percent of
Tnt-^l Charge
1.75
1.26
Cu Content,
. Percent
26.35
1.35
As Content,
. . .Percent
1.09
7.77
These data strongly indicate that much of the arsenic is vaporized during smelting.
The next step is converting the matte to blister copper by blowing with
air or oxygen in the presence of a silica flux. The iron is converted to an iron
silicate slag, the remaining sulfur is oxidized to S0_, and volatile impurities
such as arsenic and lead are largely released. Then, the blister copper is fire-
refined to "anode" copper by further blowing with air; more SCL is driven off, and
more iron, zinc, and tin are removed via a silicate slag. Finally, the anode copper
is electrolytically refined, removing almost all of the residual impurities.
Table 8 lists representative arsenic levels in the copper as it pro-
gresses in refining from the concentrate (which nominally has a ratio of As/Cu of
22,000 ppm), to copper matte, to blister copper, to anode copper, and finally to
the cathode copper product (which nominally has a ratio of As/Cu of 5 ppm). These
data show that the roasting and smelting operations remove approximately 70 percent
of the arsenic from the copper, reducing the arsenic content from 22,000 ppm (2.2
percent) to 6,400 ppm (0.64 percent). The data of Table 8 indicate that the
arsenic contents of blister copper and anode copper are approximately the same;
if a value of 900 ppm (0.09 percent) is taken, then 25 percent of the arsenic
originally in the copper concentrate is removed in the converter. The remaining
5 percent of the original arsenic is virtually all removed by electrolytic re-
fining.
Slag from smelting in reverberatory furnaces amounts to 3 metric tons per
metric ton of copper produced.
(75)
A copper blast furnace slag assayed about
-97-
-------
Table 8
Arsenic Content of Copper in Various Stages of Refining
Copper Matte
Ref.
74d
74d
74d
74d
74d
82
Avg.
As/Cu, ppm
6,600
6,900
6,800
6,100
6,700
5,400
6,400
. . Blister Copper
Ref.
68,69c
68,69c
68,69c
76
76
76
82
82,86
82,86
82,86
82,86
82,86
82,86
82,86
Avg.
As/Cu, ppm
200
1,000
350
10
370
70
80
2,300
100
100
200
1,000
1,500
4,000
800
Anode Copper
Ref.
82,86
82,86
82,86
82,86
82,86
82,86
82,86
82,86
82,86
82,86
Avg.
As/Cu, ppm
200
100
40
60
1,500
1,600
3,200
1,900
500
1,000
1,000
-98-
-------
0.04 peroent arsenic, which amounted to 5 percent of the arsenic originally in
(74d)
the concentrate; assuming there were 0.022 kkg of arsenic per kkg of copper
in the concentrate, a slag quantity of 2.75 kkg per kkg copper produced is implied,,
At another plant, 2.56 kkg of reverberatory slag plus 1.77 kkg of converter slag
(88)
are produced per metric ton of copper product. Hence, for each metric ton of
copper produced, about 3 metric tons of slag are produced which contain 1..2 kilo-
grams of arsenic.
Of the 0.9 kg of arsenic per metric ton of copper product which remain
in the anode copper, 0.22 kg are found in the slimes from the electrolysis cells,
(82)
and 0.68 kg are found in the electrolyte. The slimes are filtered and the
electrolyte bleed is evaporated yielding a sludge; both streams together amount
(75)
to 3.0 kg (dry basis) per kkg of copper product and contain the 0.9 kg of
arsenic, implying an ersenic concentration of 30 percent. These wastes are further
processed for precious metals recovery, and the arsenic eventually is disposed of
(68)
on land as a slag resulting from smelting with a basic flux.
The raw wastewaters from the primary copper industry are from four main
sources, and the quantities of arsenic are as follows:
Slag Granulation Water, 0.19 kg As/kkg Cu Product
Acid Plant Slowdown, 0.06
Contact Cooling Water, 0.00
Electrolytic Refining, 0.03
Total 0.28 kg As/kkg Cu Product
Control and treatment technology emphasizes recycle and reuse of these acidic
wastewaters, plus lime treatment (with ferric chloride flocculant) and sedimenta-
tion. One new treatment facility will reduce the arsenic concentration of the
wastewaters from 9.4 mg/1 to 1.2 mg/1; an existing facility shows no reduction
from about 10 mg/1, while a third shows a reduction from 0.85 mg/1 to 0.73 mg/1.
The recommended 1977 effluent limitation guidelines (30-day averages) are based
upon a concentration of 10 mg/1 arsenic in the effluent, equivalent to 0.02 kg of
arsenic per metric ton of copper product. Based upon a 90 percent reduction in
the quantity of wastewater, the recommended 1983 effluent limitations guideline
- in-,
-99-
-------
calls for 0.002 kg of arsenic per metric ton of product copper. The estimated
costs for compliance (not all attributable to arsenic control and treatment, of
course) are:
Capital Costs
Annual Operating Costs
1977
$334,000
$118,000
1983
$1,581,000
$ 805,000
Thus far, of the original 22 kg of arsenic per metric ton of copper, only
2.4 kg have been accounted for: 1.2 kg in slag, 0.9 kg in slimes and sludges, and
0.3 kg in raw wastewaters. Hence, almost 20 kg of arsenic per kkg of copper are
in the flue gases from roasting and smelting and from converting. Since 70 percent
of the original arsenic, or 15.4 kg/kkg, is lost in roasting and smelting, and
since 1.2 kg/kkg reports in the slag, the quantity of arsenic in the roasting and
smelting flue gases amounts to 14.2 kg/kkg. Similarly, the 25 percent of the
original arsenic lost in converting, 5.5 kg/kkg, must be in the converter flue
gases.
It is important to note that the predominant loss of sulfur is opposite
to the loss of arsenic. The S/Cu ratio is about 1.15 in the concentrate, about
0.80 in matte (after roasting and smelting), and about 0.15 in blister (after
(68)
converting). While two-'thirds of the arsenic is lost in roasting and smelting,
only one-third of the sulfur is lost in these steps. In the past, it was general
practice for byproduct sulfuric acid to be made from converter gases, while roast-
ing and reverberatory gases are released to the atmosphere after particulate con-
trol. The S02 concentration in roasting and reverberatory gases is generally
too low for economical SO- recovery. Recent air pollution regulations calling for
an overall 90 percent recovery of sulfur oxides would require an additional capital
investment by the copper industry estimated to be in excess of $250
million.(79/84>89/90'91>92) -n^ impact of additional sulfur oxides control has
been to force process changes whereby roasting and smelting gases as well as con-
verter gases, are used for acid manufacture.
-100-
-------
Since converter flue gases are predominantly used for acid-making, the
5.5 kkg of arsenic in these gaaes should appear in the acid wasted from cold-gas
cleaning of the SO2 prior to catalytic conversion to SO-,. Hot gas cyclones and
hot electrostatic precipitators upstream of the cold-gas cleaning operation should
not remove appreciable As2°3 lbecaase of its reluctance to condense. The cold-gas
venturi scrubbers and packed towers, however, should be extremely; effective in
f91)
removing the acid-soluble As^.,. The commercial sulfuric acid byproduct from
copper smelters contains no more than 0.5 ppm arsenic, verifying the effective
removal of arsenic. The arsenic-containing scrubber liquor may be treated
for removal of the arsenic (and other contaminants) with, subsequent land disposal,
or it may be used as a waste acid in the copper mining operation as a leach liquor.
This latter route is thought to be more common; in this case, the arsenic is
eventually bound to the ore residues (as ferric arsenate) . For the purpose of
this analysis, this 8,800 kkg/year of arsenic will be thought of as arsenic dissi-
pated to land.
The flue gases from roasting and smelting are generally passed through
cyclones, "balloon flues", electrostatic precipitators, and baghouses, for partic-
ulate control. As was discussed previously, these techniques have limited success
in capturing As2°3* At one c°PPer blast furnace, the dust collection system con-
sisted of bag filters downstream of cyclones. The cyclones captured 8 percent of
the arsenic in the flue gas, the bags captured 41 percent, while 51 percent escaped
collection. (74d)
The emission factor reported by Davis ( ' and by Anderson1 is 3 kg of
arsenic per metric ton of copper product. Since the arsenic in the roasting and
smelting flue gases amounts to 14.2 kg per kkg, it is implied that 11.2 kg per
kkg are collected and that the collection efficiency is 79 percent.
These collected dusts are the source of ccmnercial white arsenic in the
United States, produced solely be ASAPCO at Tacoma, Washington. If all the dusts
collected at all of the copper smelters were shipped to ASARDO/Tacoma, the 11.2
kg of arsenic per metric ton of copper product multiplied by a copper production
level of 1.60 million metric tons per year would be equivalent to 17,900 metric
tons per year of arsenic or to 23,600 metric tons per year of As2°3* Howeverf
-101-
-------
the ASARCO production level of white arsenic has been reported to be about 7,300
(42)
kkg/Vr according to one source and about 33 kkg/day (or 11,000 kkg AsJD-Xyr)
according to another source. The plant capacity for producing white arsenic
(42)
at Tacoma is on the order of 11,000 kkg/year. The difference between the
estimated flue dusts collected and the commercially-produced white arsenic is 12,600
Kkg/year of As2°3' or 9'600 Kkg/year of elemental arsenic, or 6.0 kg As/kkg of
copper produced. An explanation is that not all flue dusts collected at copper
smelters are shipped to ASARCO; this is verified by the EPA estimate that within
the copper industry flue dusts are deposited on land at a rate of 17 kg flue dust
per metric ton of copper produced. ' An ASABOO spokesman^ ' has stated that
in 1974 the amount of white arsenic shipped from the Tacoma plant was about 16,400
kkg. Since the capacity of the Tacoma plant is 11,000 kkg As2CL/year, the excess
5,400 kkg must have cone from ASAROO stockpiles.
In summary, the distribution of the arsenic originally in the copper concen-
trates .(or native copper) is as follows:
In lake copper product 30 kkg/year
In fire-refined copper product 6 kkg/year
In electrolytic copper product 7 kkg/year
In slags to land disposal 1,900 kkg/year
In sludges to land disposal 1,500 kkg/year
In flue dusts to land disposal 9,600 kkg/year
In leach residues dLssipated to land 8,800 kkg/year
In treated wastewaters 32 kkg/year
In air emissions 4,800 kkg/year
In commercial white arsenic 8,300 kkg/year
Total 35,000 kkg/year
Other Primary Nonferrous Metals
Arsenic, at concentrations significantly greater than the average crustal
concentration of 2 to 5 ppm, occurs in sulfide ores of nonferrous metals other
than zinc, lead, and copper. Among these are ores of gold, silver, mercury,
uranium, vanadium, and antimony. Table 9 lists the U.S. production levels of these
metal ores.
-102-
-------
Very little data is available on the arsenic content of these ores, or on
the fate of the arsenic during the mining, milling, snclting, and refining opera-
tions. Arsenic occurs at 3 percent of antimony in antimony ores; it is recovered
(and sold) via a sulfide precipitation step in the hydrometallurgical Sb_0_ pro-
cess. Arsenic did show up at appreciable concentrations in the raw waterborne
wastes from raining and milling operations:
(75)
Source
Gold Mines
Gold Mills
Silver Mills
Mercury Mills
Uranium Mines
Uranium Mills
Vanadium Mills
Antimony Mills
Arsenic Concentration in Raw Waste, mg/1
0.03 - 0.08
0.05 - 3.5
0.07 - 3.5
0.02 - 0.38
0.01 - 0.03
0.1 - 1.5
0.35
0.23
For the purposes of this study, a very rough estimate of the quantity
and fate of arsenic is based in part upon the similarity of these minor nonferrous
metal ores and recovery processes to the lead-zinc industry, and in part to the
data for antimony. It is assumed that the arsenic in ore concentrates is one
percent of the quantity of each of the metals in Table 9; and that one-third is
recovered as sulfides and sold as pigments, that one-third is lost to the atmosphere
and that one-third is in land-destined wastes.
Based upon a total production level of 15,000 metric tons per year for
all of the metals in Table 9, the arsenic involved is 150 metric tons per year,
of which 50 kkg/year is recovered for commercial purposes.
Arsenic in Nonferrous Metal Products
The quantity of metallic arsenic used in 1974 for non-ferrous alloying
is between 540 ' and 1,240 metric tons. In addition, it was previously
estimated that the quantities of new arsenic retained in refined primary metals
are:
-103-
-------
Table 9
Primary Ncal-Farirous Metal Ores Mined in the U.S.
Metric Tons/fear of Mstal ^Content
Sources* Bureau of Mines(18/70)
Gold
Silver
Mercury
Uranium
Vanadium
Antimony
1970
54.1
1,400
941
9,360
4,830
1,025
1971
46.6
1,295
616
9,430
4,760
930
1972
45.1
1,155
253
9,900
4,420
443
1973
36.7
1,175
75
9,920
3,970
494
1974
34.9
1,050
59
8,910
—
544
-104-
-------
In Zinc, 5 kkg/year
In Lead, 20 kkg/year
In Lake Copper, 30 kkg/year
In Fire-Refined Copper, 6 kkg/year
In Electrolytic Copper, 7 kkg/year
Total, o 68 kkg/year
(42)
Arsenic at one percent concentration in lead shot amounts to 60 metric
(47)
tons per year. Uiis arsenic, used to enhance the sphericity of the lead shot,
should be all new arsenic since no lead shot is recycled.
Arsenic is also used at about 0.6 percent in lead-tin bearing metals
(babbitts). (42'6°) in 1971 and 1972, the quantity of lead consumed for bearing
metals was 14,600 metric tons per year; since the lead content of babbitt
metal is 83 percent, the arsenic quantity is approximately 175 metric tons per
year. However, babbitt metal is extensively recycled, with 12,500 kkg of lead
recovered in 1972 from babbitt metal scrap. By difference, only 2,100 kkg/year
of new lead is consumed in bearing metals. Since the melting point of lead
(326°C) and of babbitt metals (260-270°C) is low compared to the vaporization
temperature of arsenic (613°C), little arsenic is lost in secondary lead kettle
refining. Hence, about 150 kkg/year of the arsenic is recycled, while 25 kkg/year
is new arsenic.
Antimonial lead (hard lead) is used primarily for the posts and grids of
lead-acid storage batteries, and for lead cable sheathing. The arsenic con-
centration of such alloys ranges from 0.15 percent for arsenical lead, to no more
than 0.5 percent for antimonial lead. 'An arsenic content of 0.25 percent
will be used for the purposes of this study. In the 1971-1972 time period, the
lead consumption for these purposes was:
Battery posts and grids 303,000 kkg/year containing 760 kkg As/yr
Cable covering 45,000 kkg/year containing 110 kkg As/yr
Total 348,000 kkg/year containing 870 kkg As/yr
It is therefore inferred that 870 metric tons per year of arsenic is contained in
lead alloys for batteries and for cable covering.
-105-
-------
Like babbitt metals, antimonial lead is extensively recycled. In 1972,
322,000 metric tons of lead was recovered fron old antimonial lead scrap and old
cable convering scrap, which inust nave contained 805 kkg of arsenic (at 0.25
percent). Conversely, only 6,800 metric tons of primary antimonial lead was
manufactured in 1972. Antimonial lead is recovered in lead blast furnaces, ,
(2\
and much of the arsenic accompanies the lead through this process. ' Using the
air emission factor for lead smelters, 0.4 kg of arsenic per metric ton of lead; *9'
(74b 74c)
the concentration of arsenic in lead blast furnace slag, 0.2 perpent; ' ' and
the quantity of slag from secondary lead blast furnaces, 148,300 kkg/year; the
805 kg/year of arsenic is distributed as follows:
Loss to atmosphere 130 kkg/year
In slag to land disposal 300 kkg/year
Retained in secondary lead (by difference) 375 kkg/year
The difference, then, between the 870 kkg/fyear of arsenic in lead alloys produced,
and the 375 kkg/year of arsenic retained in recycled lead, is 495 kkg/year of new
arsenic which must be added.
Arsenic, at a concentration of 0.03 percent, is used in Admiralty brass
for condenser and heat-exchanger tubing. ' ' This alloy contains 71 percent
copper. In 1968, the copper demand for all industrial non-electrical machinery
(17)
was 250,000 metric tons; ' if 10 percent of this is taken as an extreme estimate
of the Admiralty brass production, the quantity of arsenic involved would be 7.5
metric tons per year. A much more important use for arsenic-containing copper
is for automotive radiators, where a nominal 0.3 percent of arsenic is used. A
typical auto radiator weighing 6.6 kilograms contains 5.9 kilograms of copper;
at a production level of 10 million autos (and other vehicles) per year, 59,000
kkg/year of copper coitaining 175 kkg/year of arsenic are consumed. Because of
the high value for scrap copper, virtually all auto radiators are recycled. In
1972, the consumption of old unsweated auto radiators amounted to 67,000 metric
tons.'70'
Auto radiator scrap is therefore the predominant form of arsenic-bearing
scrap copper. This category of scrap is normally not processed in blast or cupola
-106-
-------
melting furnaces (which are used for lower-grade scrap, slags, and drosses), but
(122)
is fire-refined. It was earlier shown in the discussion of the primary copper
industry that fire-refining removes little if any arsenic from elemental copper;
this is substantiated upon theoretical grounds which show that arsenic in elemental
(85)
copper is most difficult to either oxidize or volatilize. Hence, it may be
concluded that the arsenic in secondary copper remains with the product. Some
of this secondary refined product is later electrolytically refined (removing arsenic
and other impurities), and some is directly used. Scrap segregation practices
(121 122)
are ccnnon in the secondary copper industry ' to meet product purity re-
quirements by careful blending of available scrap; it appears that some of the
arsenical copper scrap would be used in manufacturing arsenical copper for new
auto radiators. A gross estimate is that of the 175 kkg/year of arsenic in
copper scrap, 75 kkg/year reports in new radiators (along with 100 kkg/year of
new replacement arsenic), while 100 kkg/year is dissipated in other copper alloy
or is removed via electrolytic refining.
In summary, the flow of arsenic in the nonferrous metals industry is
estimated as follows:
Additions of New Arsenic
Retained in Primary Zinc
Retained in Primary Lead
Added to Lead for Lead Shot
Added to Lead for Bearing Metals
Added to Lead for Batteries, Cables
Retained in Primary Copper
Added to Copper for Admiralty Brass
Added to Copper for Auto Radiators
Total New Arsenic in Nonferrous Metals
5 kkg/year
20 kkg/year
60 kkg/year
25 kkg/year
495 kkg/year
43 kkg/year
7 kkg/year
100 kkg/year
700 kkg/year
-107-
-------
Old Arsenic Recycled via Secondary Metals
In Bearing Metals
In Battery and Cable Lead
In Auto Radiators
Total Old Arsenic in Nonferrous Metals
150 kkg/year
375 kkg/year
75 kkg/fyear
600 kkg/year
Arsenic Losses in Nonferrous Metal Processing
Air Emissions, Secondary Lead Blast Furnaces 130 kkg/year
300 kkg/year
430 kkg/Vear
Land-Destined Slag, Secondary Lead Blast
Furnaces
Total Arsenic Losses
Arsenic; Dissipated in Nonferrous Metals
By Diffierence, New Arsenic Less losses
325 kkg/year
An alternate irethod of accounting is by individual end items of alloys
containing arsenic (kkg/yr):
End Items
lead Shot
Lead Bearings
Lead Batteries
Lead Cables
Total Lead Items
Copper Radiators
Heat Exchangers
Total Copper Items
Total Pb & Cu Items
Arsenic In
End Items
60
175
760
110
1,105
175
7
182
1,287
Arsenic
Reclaimed
0
150
700
105
955
75
0
75
1,030
New Arsenic
Added
60
25
430
65
580
100
7
107
687
Arsenic Lost
in Processing
0
0
370
60
430
0
0
0
430
-108-
-------
Phosphate Rock
Arsenic is a common trace constituent of phosphate rock, and occurs as
adsorbed in ions on colloidal iron oxide rather than as a substitute for phosphorus
in the fluorapatite. A statistical analysis of 51 commercial Florida pebble
phosphates indicated a direct linear correlation between arsenic and iron in the
rock, with arsenic varying from 3 to 15 ppm at a ratio of As/Fe of 800 ppm.
Reported values for the phosphorus, arsenic, and iron content of commercial
phosphate rocks are tabulated in Table 10.
These data may be summarized by the following ratios:
(124)
Florida Pebble, Rock
Tennessee Rock
Wsstern Rock
As/P205,
ppm
45
82
230
As/P, ppm
100
190
520
As/Fe, ppm
1,370
1,370
7,900
The quantities of marketable phosphate rock in 1972, the corresponding
quantities of arsenic in the rock and the breakdown of phosphate and contained
arsenic by consumption patterns are shown in Table 11.
The arsenic in phosphate rock follows the phosphorus quantitatively,
whether the wet process for phosphoric acid (i.e., acidulation of the rock) or
the furnace process (reduction to elemental phosphorus) is followed. In Table 11,
the "nan-agricultural11 uses are those derived from the furnace process. Arsenic
is intentionally removed from food-grade phosphoric acid by precipitation with
Na2S or NaHS followed by filtration; and arsenic is removed in the manufacturing
processes for phosphorus pentasulfide phosphorus trichloride, and phosphorus
oxychloride. ' ' It is estimated that the arsenic removed (and disposed of
on land) amounts to all of the 60 kkg/year associated with food-grade phosphoric
acid plus half of the 60 kkg/year associated with miscellaneous uses. Conversely,
all of the 293 kkg/year of arsenic associated with fertilizer, the 32 kkg/year
-109-
-------
Table 10
Arsenic in Catmercial Phosphate Rock
Florida land pebble
Otennessee brown rock
Western rock
Ref.
13
12
124
124
13
12
13
12
124
P2°5'
V
30-36
30-36
27-36
Fe2°3'
V
0.7-2.6
2.2-3.4
0.5-2,1
as2o3,
10-50
5-30
22
9
20-40
7-75
10-150
6-140'
63-200
-110-
-------
Table 11
Production, conversion, and consumption of Phosghates
1
Florida Rock Produced
Used as Domestic Fertilizer
Used for Animal Feed
Exported
Tennessee Rock Produced
For Nan-Agricultural Uses
Western Rock Produced
Used as Domestic Fertilizer
Used for Animal Feed
Exported
For Non-Agricultural Uses
Total Rock Produced
Total Used as Domestic Fertilizer
Total Used for Animal Feed
Total Exported
Total for Non-Agricultural Uses
Used for Detergents
Used for Food Products
For Miscellaneous Uses
PjOg Quantities,
Metric T^ns/V^ar
9,960,000
5,450,000
350,000
4,160,000
510,000
510,000
1,170,000
210,000
70,000
70,000
820,000
V
11,640,000
5,660,000
420,000
4,230,000
1,330,000
630,000
350,000
350,000
Arsenic Quantities,
Metric Tons/Year
»
448
245
16
187
42
42
268
48
16
16
188
758
293
32
203
230
no
60
60
-111-
-------
associated with animal feeds, the 110 kkg/year associated with detergents, and
the remaining 30 kk'g/year associated with miscellaneous uses, remains with the
phosphate products and is dissipated with these products.
The consumption of fertilizers is expanding at a 5 to 7 percent growth
rate in North America: ^125*
1965 3.6 million metric "tons P_0,_/year
1970 • 5.0
1975 6.3
1980 8.0
Hence, the arsenic associated with phosphate fertili.ifers is expected to grow to
410 metric tons per year by 1980.
Arsenic in household detergents and presoaks was measured at concentrations
ranging from 2 to 59 ppm. The production of so; lium tripolyphosphate for the
detergent industry has been cut back over the past soveral years because of the
environmental concern over phosphorus in wastewaters.
Sludges from Mffiicipal Sewage Treatment and Municipal Water Treatment
Some of the arsenic in domestic sewage concentrates in the treatment plant
sludge, in a similar fashion as other metals. At one secondary treatment plant,
the arsenic in the thickened waste sludge was at a concentration of 61.4 yg/1; at
an assumed 8 percent solids content, the sludge solids would have contained 0.75
(127)
ppm arsenic. If the per capita dry sludge solids quantity is 0.091 kilograms
per day, and if 120 million people are served by municipal sewage treatment plants,
then the quantity of arsenic in sewage sludge is 3.0 metric tons per year, con-
tained in a dry sludge quantity of 4 million kkg/year.
The arsenic emission factor for sludge incineration is reported as 0.01
kg per kkg of "sewage and sludge". Assuming a solids concentration of 20
percent in dewatered sludge (feed to an incinerator), the emission.factor is
equivalent to an arsenic concentration in dry sludge of 2 ppm, and it implies
that all of the arsenic is volatilized. Since about one-third of all sludge is
incinerated, one kkg/year of arsenic is emitted to the air and 2 kkg/year is
applied to land.
-112-
-------
It was also determined that municipal water treatment plants remove arsenic „
Cold-lime softening removed 85 percent of the arsenic in raw water, from a raw
concentration of 3.1 ppb. If the per capita water use is 200,000 liters per
12
year, or 24 x 10 liters per year for 120 million people; then at a removal rate
of 2.7 yg/liter the quantity of arsenic in water treatment sludge is 65 metric
tons per year.
Sulfur Deposits
(128)
Based upon one reported value of less than 10 ppb of arsenic, it is
apparent that Erasch process sulfur does not contain appreciable arsenic.
Borax and Boric Acid
While boron is not an extremely rare element, few commercially attractive
deposits of boron irinerals are known. It is estimated that about half of tiie
cormercial world boron reserves, estimated at about 72 million tons of boron,
are in southern California as bedded deposits of borax (sodium borate) and col©-
manite (calcium borate), or occur as solutions of boron minerals in Searles Lake
brines. The United States is the largest producer of boron, supplying 71 percent
of the free world demand, and also the largest consumer, requiring about 36 per-
cent of the vprld output. The U.S. production of boron minerals and compounds has
averaged 1.07 million metric tons per year in the 1972-1974 time period; the corres-
ponding quantity of B2
-------
dissolve the borax. The concentrated borax liquor goes to a series of thickeners,
is filtered and pumped to vacuum crystallizers. One of .the crystallizers pro-
duces borax pentahydrate, and the other produces borax decahydrate. The penta-
hydrate is used for boric acid manufacture.
d
Arsenic is present as a sulfide (Realgar) in the mine run ore and
associated shales. The occurrence is intermittent, and a given ore horizon can
vary from 0 to over 1,000 ppm of arsenic. The residue from the digested ore
amounts to 800 kg per metric ton of borax products, and contains approximately 45
ppm of arsenic. On the basis of 860,000 metric tons per year of borax derived
from ore, the quantity of wastes is 690,000 kkg/year, and it contains 31 metric
tons per year of arsenic. These1'wastes are deposited in ponds, and covered with
water to prevent blowing dust. Since there is no ground water in the remote desert
area, there is no likelihood of contamination derived from percolation. Process
wastewaters are evaporated in ponds.
Sodium berates are also extracted from Searles Lake brines by Kerr-MoGee
Corporation whose primary products are soda ash, salt cake, and potash. Searles
Lake is a dry lake covering about 34 square miles in San Bernardino County,
California. Brines pumped from beneath, the crystallized surface of the lake are
processed by carbonation, evaporation, and crystallizatioji procedures, producing
an array of products including boron oonpounds.
The salt body is actually two deposits separated by a layer of muds, and
each deposit contains brines of different compositions. Hawever, both the upper
structure brine and the lower structure brine contain Q.05 percent Na-AsO.,
n_29i
equivalent to 180 mg/1 elemental arsenic.v ' The total brine processed is about
9 (129)
12 x 10 liters per year, ' so that the contained arsenic is 2,160 metric tons
per year. The depleted brines, plus added process waters, are returned to the
lake; almost all of the arsenic in the brine extracted from the lake is directly
returned to the lake in the depleted brine. The only arsenic extracted from
the brines is that unintentionally carried as an impurity- in the products of the
operation.
-114-
-------
Boric acid is made by acidulation of borax pentahydrates
From, the acidulator, thesboric acid solution is fed to a vacuum crystallizer,
where boric acid crystals are formed, and then to a filter. One sodium sulfate is
remowsd in the filtrate, and the technical grade boric acid is dried and packaged,
The technical grade product can also be diverted upstream of the final drying step,,
redissolved, crystallized, filtered and dried to produce a higher purity product.
Sodium sulfate is a co-product and most of the wastes are waterborne. The com-
bined waste liquors from several filtration and centrifugation steps amount to
2,800 liters per metric ton of boric acid product, and contain 36 grams of
arsenic. The quantity of borax used as a raw material is 1.72 metric tons per
metric ton of boric acid product. Since the production volume of boric
acid is 110,200 metric ton of boric acid product, and contain 36 grams of
borax is used for this purpose, and the raw wastewaters contain 3.9 metric tons
per year of arsenic. At present, the arsenic-containing wastewaters are dis-
charged, but the iiipact of effluent discharge limitations should cause arsenic
wastes to be divarted to land disposal by 1977.(123)
Furthermore, if it is assumed that the arsenic in boric acid wastewaters
represents all of the arsenic in the borax raw material, then the concentration of
arsenic in the borax is 21 ppm. Since the residue from borax manufacture amounts
to 800 kg per kkg of borax and contains 45 ppm of arsenic, then the material
balance of arsenic is as follows:
Borax Ore Mined
Borax Product from Ore
Residue from Ore
Borax Product from Brines
Total Borax Product
Borax Consumed for H^BO^
Other Borax Products
Total
Quantity,
kkgArear
1,550,000
860,000
690,000
210,000
1,070,000
190,000
880,000
Arsenic
Concentration,
pcro
32
21
45
21
21
21
21
Arsenic
Quantity,
kkg/year
49
18
31
4
22
4*
18
*This arsenic is subsequently a waterborne residual from H_B03 manufacture.
-115-
-------
Iron Ore
Sedimentary iron ore has. been reported by U,S,G,S. to contain 400 ppro
arsenic. Although, little substantiating data has been found, this, value is
consistent with, the pattern of coprecipitation of arsenic with, hyjarous iron oxides,
For example, the As/Fe ratio in Florida phosphate rock, deposits was previously
shown to be 1,370 ppm; applying this ratio to crude iron ore with an iron content
of 37 percent results in an extrapolated arsenic concentration of 500 ppm in the
crude ore.
On the other hand, the very lack of substantiating data for the concen-
tration of arsenic in iron ore leads one to question the validity of this one re-
ported value of 400 ppm. In comparison, the phosphorus content is universally
reported; it has averaged 400 ppm for Lake Superior ores in the 1970-1972 period.
wastewaters from iron mines and from iron ore processing have, been characterized
f73)
in terms of almost 20 constituents, without mention of arsenic.
Arsenic was discussed as a minor constituent of iron ore in a United
Nations survey: "Arsenic in excess of 0.1 percent is uncommon in iron ores;
when present, it is usually found in brown hematites as arsenopyrites (FeAsS),
loellingate (FeAs2) and scorodite.(FeAs04•4H20)." Based upon the U.S.G.S. and
the U.N. references, the arsenic content of iron ore will be assumed to be 400
ppm for the purposes of this study, although more effort should be expended in
verifying this concentration level.
For the past five years (1970 through 1974), the average usable iron
ore statistics have been as follows:(18'70)
Production 84.6 million metric tons/year
Imports for Consumption 43.0
Exports 3.1
Consumption, Total 134.5
Based upon the above level of consumption of iron ore in blast furnaces
and upon an arsenic concentration of 400 ppn, the quantity of arsenic entering
the U.S. blast furnaces is 54,000 metric tons per year.
-116-
-------
Arsenic acts very much like phosphorus in the blast furnace; it is com-
pletely reduced forming non-volatile iron arsenide (FeAs) and iron diarsenide
(PeAs2), and reports in the pig iron. (131'132) in 1970 to 1974f the average U.S.
pig iron production was 83.0 million, metric tons per year, ' ' so that the
54,000 metric tons per year of arsenic would result in a concentration of 650 ppm
CO.065 percent) in pig iron.
Of this pig iron, 78.0 million metric tons per year was consumed in steel-
making, while 5.0 million kkg/year was consumed for cast iron products (2.4 million
in cupolas and 2.6 million in direct castings). The arsenic retained in cast iron
would be 3,300 kkg/year. In the basic steelmaking processes using pig iron (basic
oxygen and basic open hearth), most of the arsenic as well as the chemically-
similar phosphorus is removed by the lime flux, and reports in the slag as calcium
arsenate. The phosphorus content of pig iron is in the range of 0.15 percent, while
the corresponding content in steel is 0.035 percent. By analogy, it is assumed
that the arsenic content of pig iron, 0.065 percent, is reduced to 0.015 percent in
basic steelmaking.
Some of the arsenic lost in steelmaking would be in the steelmaking dusts
(as a consequence of entrainment of solids rather than as a result of volatility).
In 1972, the basic oxygen process consumed about 56 million kkg of pig iron while
the open hearth process consumed about 22 million kkg. The uncontrolled dust
emission factors are 25.5 kg/kkg steel produced for the basic oxygen furnace and
4.15 kg/kkg steel produced for the open hearth.(133) Moreover, the 1972 steel pro-
duction quantities were 67.6 million kkg for the basic oxygen and 31.7 million
kkg for the open hearth.(70)
If it is assumed that the arsenic in the uncontrolled dust emissions is
at the same concentration level as it is in the steelmaking charge, the following
may be derived for the steelmaking processes:
-117-
-------
Arsenic In Pig- Iron, kkg/yr
Dust/Pig Iron, kkg/kkg
Arsenic In Dusts, kkg/yr
Arsenic In Slag, kkg/yr
Arsenic In Steel, kkg/yr
Total Dusts, kkg/yr
Basic Oxygen
36,400
0.0308
1,100
25,200
10,100
1,720,000
Open .Hearth .
14,300
0.00599
100
9,400
4,800
132,000
. Total
50,700
—
1,200
34,600
14,900
1,852,000
If a 99 percent dust collection efficiency is assumed, then the collected dusts
would contain about 1,200 kkg/year of arsenic while the air emissions would con-
tain 12 kkg/year of arsenic.
In past years, the collected dusts from steetaaking furnaces (which
contain iron oxide) were sent to the sintering plants along with ore fines, coke
breeze, limestone and recycled material from various mill processes. The purpose
of the sintering process is to form larger agglomerates from the fines for
recycle to the blast furnace. However, the sintering operation has been under
recent attack because of its poor record of air pollution, and the recent trend
has been to dispose of furnace dusts as landfill rather than to recover the iron
values by sintering and recycling. little is presently known of the environmental
hazards of land-destined dusts containing arsenic, which of course involve much
more arsenic than the arsenic emitted to the atmosphere.
The 2.4 million metric tons per year of pig iron which is used for cast
iron production via cupola and similar furnaces is augmented by 14.8 million metric
tons per year of scrap feed, for a production level of 17.2 million metric tons
per year. EPA reports an arsenic uncontrolled emission factor for cast iron
production of 0.007 kg per metric ton of metal charged, which implies a total
arsenic emission of 120 metric tons per year from these sources. This amounts to
one percent of the arsenic in the metal charged (at 650 ppm), 11,200 kkg/year.
The emitted arsenic may be partially due to dust entrainment, and it may also be
due to volatization of As2°3 from ^•ron axsen^-^Q (th6 intermediate stage of re-
duction between iron arsenate and iron arsenide). The arsenate and the arsenide
are both non-volatile, but the arsenite is volatile. The high level of
arsenic in tiie cupola dusts, 0.7 percent, suggests that volatility plays a
significant role.
-118-
-------
Of the uncontrolled emissions of 120 kkg/year of arsenic from cast iron
furnaces, an estimated 20 kkg/year is released to the atmosphere, with the re-
maining 100 kkg/year collected and disposed of on land. The relatively low
collection efficiency is based upon the implied volatility of tihe arsenic emissions
The slag from steel-making furnaces is widely used, as the following 1972
data indicate:(70)
Use
Railroad Ballast
Highway Base or Shoulders
Paved-area Base
Misc. Base or Fill
Bituminous Mixes
Agricultural
Other Uses
Metric Tons/year
1,200,000
3,240,000
1,610,000
1,750,000
510,000
100,000
800,000
Total 9,210.000
It should be emphasized that the estimate of the quantity of arsenic in
steel slag (34,600 metric tons per year) is a very rough one, indeed. In addition,
no information was obtained on the potential for arsenic leaching from slag in the
uses typified by the above data.
Manganese Ores
Analyses of three typical manganese ores are as follows:
. (134)
Mn, %
Fe, %
P, %
AS, %
Brazil
50
4.1
0.07
0.18
Brazil
48
5.2
0.09
0.15
Mexico
47
1.8
0.01
0.25
An average value of 0.20 percent arsenic will be used in this analysis.
-119
-------
All of the manganese ore (with 35 percent or more Mn) consumed in the
U.S. is imported, principally from Africa and Brazil. The U.S. government stock-
piles manganese ore, and in recent years has released significant quantities to
industry. There is a sizable domestic production of manganiferous ore (5 to 35
percent fin.). The quantities involved are shown on Table 12. At an average level
of domestic industrial consumption of 2.0 million metric tons per year, the
quantity of arsenic involved is 4,000 metric tons per year.
The smelting of manganese ore to produce manganese ferroalloys (ferro-
manganese, silicomanganese, and spiegeleisen) is generally accomplished in blast
furnaces or electric furnaces, with technology very similar to iron and steel
manufacture. ' ' Although little data is available on the fate of the arsenic
in the smelting of manganese ores, an analogy may be drawn to the transport of
the chemically-similar phosphorus: 60 percent of the phosphorus in the ore passes
into the ferroalloy, 30 percent passes into the slag and 10 percent escapes with
furnace gases. Since about 90 percent of the manganese ore is consumed in
ferroalloy production, the fate of the arsenic is estimated (by analogy with
phosphorus) as follows:
Retained in ferroalloys, ,
consumed in iron and steel 2,160 kkg/year
In slag from ferroalloy furnaces 1,080 kkg/year
In collected dusts from furnaces 350 kkg/year
Air Emissions from furnaces 10 kkg/year
Total 3,600 kkg/year
The remaining manganese ore is used for making carbon-zinc and alkaline
manganese dioxide dry cell primary batteries, and for use in the chemicals and
glass industries. In 1972, 208,900,000 alkali batteries were produced by seven
plants, with a total battery weight of 14,087 metric tons (an average of 67.3
grams per battery). Of the total battery weight, 27.4 percent is manganese
-120-
-------
Table 12
Manganese Ore Statistics
-
Year
1968
1969
1970
1971
1972
1973
1974
Manganese Ore, kkg/year
Imported
1,660,000
1,780,000
1,570,000
1,740,000
1,470,000
1,370,000
1,090,000
Govt Stockpile
Releases
140,000
110,000
200,000
170,000
910,000
Consunption
2,020,000
1,980,000
2,140,000
1,950,000
2,110,000
1,940,000
1,630,000
Manganiferous
Ore Produced,
kkg/yr
220,000
390,000
330,000
180,000
130,000
-121-
-------
dioxide, iirplying that 3,870 metric tons of manganese dioxide were con-
sumed in 1972. At an arsenic concentration of 0.2 percent, the quantity con-
tained is 7.75 kkg/year. These alkaline dry cells have found wide usage in
flashlights, camera equipment, battery-powered toys, radios, tape recorders, etc.;
since the alkaline cell yields an improved performance (at higher cost) over
carbon-zinc cells, particularly for heavy or continuous current drains.
The carbon-zinc batteries produced in 1972 amounted to 95,920 metric tons.
Manganese dioxide amounts to 61.5 percent of the battery weight, implying a consump-
tion of 58,900 metric tans per year of Mn02.' ' The MnQ2 is used as a depolarizer
in conjunction with annoniurn chloride, zinc chloride, and starch to form the
electrolyte. The oarbon-zinc batteries are used for similar purposes as the
alkaline battery, although larger industrial carbon-zinc batteries are also
used. At an arsenic concentration of 0.2 percent, the quantity contained is
118 metric tons per year.
Hence, the total arsenic dissipated in primary batteries is 126 metric
tons per year. The remainder, approximately 274 metric tons per year, is involved
with chemical-grade manganese ore, and is dissipated in products such as hydro-
quinone or potassium permanganate. It appears that virtually all 400 kkg/year
of arsenic in non-ferroalloy manganese ore is dissipated in end products.
In addition to manganese ore reserves, the potential for large-scale re-
covery of manganese nodules on the deep ocean floors has attracted intense U.S.
and foreign attention. Ferromanganese nodules in the mouths of rivers and in
bays in Lake Michigan contain 200 to 500 ppm arsenic.
Fossil Fuels
Tfie average arsenic content of domestic coal has been reported to be
(9)
5.44 pom. Eastern coals contain 10 ppm arsenic and western coals 1 ppm. The
(138)
arsenic content of coal increases with increasing sulfur and iron pyrite content;
this observation is consistent with the sulfur contents of eastern (3 percent)
and western (0.7 percent) coal.
-122-
-------
The U.S. bituminous ooal statistics, including a projection for 1980,
(18)
in millions of metric tons per year, are: ,
. Year . .
1970
1971
1972
1973
1974
1980
Production .
547
501
540
537
535
812
. .Exports .'.
64
51
51
48
55
(54)
U.S. Consumption
(total)
468
449
469
505
490
(758)
U.S. Consumption
(electric power)
290
299
317
351
355
580
Based upon an annual consumption of 450 million metric tons and upon an
average arsenic content of 5.44 ppm, the arsenic associated with coal is 2,450
metric tons per year. In a study of coal-fired power plants, 73 percent of the
arsenic in the coal reported in the bottom ash and in the collected fly ash,
while 27 percent (1.46 grams arsenic per metric ton of coal burned) was emitted
(9\
to the air after dust collection. '
The data above also show that in 1974, 72 percent of the total coal
consumed was for electric power generation. Of the remainder, 17 percent was
consumed by coke plants, 11 percent by other manufacturing and mining industries,
(18)
and only 1 percent was delivered by retail dealers. The proportion for
electric utilities is expected to increase by 1980. The coal consumed by coke
plants is selectively the lew-sulfur coal, so by inference the arsenic quantities
should be small. Applying the above emission factor to the total coal consumption
should therefore be a reasonable procedure. The arsenic emitted to the atmosphere
is estimated to be 650 metric tons per year; while the arsenic in bottom ash and in
collected fly ash, destined for land disposal, is estimated to be 1,800 metric
tons per year. The arsenic in the ash is, in general, partially mobilized into
the environment via dusting and via leaching.
-123-
-------
Of major importance in this estimate of arsenic emissions from coal is
the projected increases in coal utilization due to the energy situation. The
foregoing table lists the rather stable coal statistics for the past five years,
but the 1980 projection reflects an annual growth rate of over 6 percent. The
impact is that the domestic consumption in 1980 is expected to be around 760
million metric tons. Hence, the arsenic quantity could be increased to about
4,100 metric tons per year (1,100 kkg/year in air emissions and 3,000 kkg/year
in land-destined wastes).
The growth of coal consumption is expected to continue well past 1980;
the U.S. recoverable reserves are estimated to be 394 billion metric tons.
Much research is currently underway in developing coal conversion pro-
cesses (synthetic oil and synthetic low-and-high-Btu gas). The EPA is actively
investigating the fate of the heavy metals in these conversion processes. In one
preliminary study of a high-Btu gasification process, starting with Pittsburgh
No. 8 coal containing 9.6 ppm of arsenic, 22 percent of the arsenic was volatilized
in the first stage (430°C and 1 atmosphere), an additional 25 percent in the second
stage (650 °C and 74 atmospheres), and an additional 18 percent in the tMrd stage
(1000 °C and 74 atmospheres), leaving 35 percent of the original arsenic in the
residue. As expected, the more volatile trace elements (Od, Hg, Pb, As, Se) wound
up primarily in the product gas, while most of the less volatile trace elements
(Cr, Ni, and V) remained primarily in the residues.
A projected implementation of coal gasification is that by 1990 the U.S.
will have the capacity to process 220 million metric tons of coal per year.
The above preliminary data indicating that two-thirds of the arsenic is volatilized
(and therefore vrould become air emissions upon combustion of the synthetic gas)
is the incentive for research to remove this arsenic.
The average arsenic content of foreign and domestic crude oils and of
(9)
residual oil was 0.14 ppm. At an average specific gravity for crude oil of
-124-
-------
0.85 kgAiter, the volunetric arsenic concentration is 0.12 mg/literi The total
domestic demand for petroleum products is as follows: f7^
1970 0.853 x 1012 liters/year
1971 0.882
1972 0.951
1973 1.000
1974 0.982
Hence, the arsenic in consumed petroleum amounts to 120 metric tons per
year. In 1972, the oonsurrption pattern was as follows:
(70)
Gasoline
Jet Fuel
Other light Fuels
Distillate Fuel Oil
Residual Fuel Oil
Total Fuels
Chemical Feedstocks
Asphalt, Road Oil
Misc. Products
Total Non-Fuel
39.2%
6.4
11.2
17.8
15.5
90.1%
3.8
1.9
4.2
9.9%
For the 90 percent of the total petroleum that is burned, all of the
arsenic is in the form of air emissions; this amounts to 108 kkg/year. The
remaining 12 kkg/year of arsenic may be assumed to be dissipated in end products,
Oil shale is projected to fill a small but significant fraction of the
U.S. energy demand:
.(140)
Year
1975
1980
1985
1990
1995
2000
Total U.S. Energy
Demand, 10 le joules/yr
83
98
120
140
170
200
Oil Shale Production/Yr
10 9 Liters Oil
0
16
52
70
87
105
10 18 joules
0
0.6
2.0
2.7
3.4
4.0
Percent of Derrend .
filled by 'Oil Shale I
0 1
0.6 !
1.7 ;
1.9
2.0 i
2.0 !
-125-
-------
This projection is highly dependent, of course, upon the relative economics of
oil shale vs. petroleum; water availability is a serious constraint. The arsenic
in oil shale has been reported to be at a level of 82 ppm. Since the expected
(140)
oil recovery is about 140 liters per metric ton of oil shale, the quantities
of rained oil shale and of arsenic corresponding to the above projections of oil
production are:
Year
1975
1980
1985
1990
1995
2000
Oil Shale Mined,
Million kkg/year
0
115
370
500
620
750
Arsenic In Oil Shale,
kkg/year
0
9,000
30,000
41,000
51,000
62,000
Ihe oil shale will be mined with underground mining methods, since the
amount of overburden is prohibitive for surface mining. It is anticipated that
the spent shale residue will be disposed of on land in 80-neter-deep piles. Once
shale has been retorted, the organic binding is destroyed and the rock loses its
strength and is easily crushed, thereby exposing soluble minerals to leaching
actions.(140)
It appears likely that while some of the arsenic would be in the re-
covered oil, process wastewaters, or process gases, most will probably be retained
in the spent shale residue as non-volatile arsenates. Hence, the primary concern
over arsenic may be the possibility for erosion, leaching and runoff. If slurry
transport of processed shale is employed, the mobilization of arsenic would be
accelerated. In order to protect surface and ground waters, control measures
such as impermeable basin liners and surface revegetation would likely be employed.
It is also possible that any organic arsenic originally in the oil shale
would be volatilized in reducing atmospheres in the retorts to arsine or to methyl
arsines.
-126-
-------
Geothermal Energy
Geothermal waters, such as the waters of hot springs, contain much more
arsenic than the averaga of one ppb of normal fresh water. Extreme concentrations
up to 13.7 ppm have been reported for hot springs,
(130)
and it has been considered
that hot springs and volcanic exhalations contributed much of the arsenic now pre-
(22)
sent in the sediments and sedimentary rocks of the earth's crust. A report
of the composition of geothermal fluids from three locations makes no mention
of arsenic, although 20 other components were reported at concentrations in the
(140)
100 ppb range. For the purposes of this study, arsenic concentrations of
10 ppb and of 1 ppm will be investigated.
Qeothermal energy (like oil shale) is projected to fill a small but
significant fraction of the U.S. energy demand:
Year
1975
1985
2000
Geothermal Energy
Produced,
Billion KWH/yr
4
50
400
Liquid Brought to
Surface,
Million kkg/yr
8
900
14,000
Arsenic In Liquid, kkg/yr
At 10 ppb
0.08
9
140
At 1 ppm
8
900
14,000
In the above tabulation, the factors used were 40 kilograms of liquid
per KWH for wet geothermal processes and 2 kg/KWH for dry processes.
Reinjection of the fluids into the subsurface geottiermal reservoir,
after the heat energy has been extracted, is the likely course that will be
followed. In this event, none of the arsenic in the fluid (regardless of its
concentration) should be mobilized into the environment.
-127-
-------
SECTION VI
ARSENIC TOXICOLOGY
The medicinal potential of arsenic has been acclaimed for nearly 2500
years. Hippocrates (460 to 377 B.C.) is said to have treated ulcers and other
(27)
disorders with realgar (As2S2, arsenic sulfide). ' The toxic properties of
arsenic have supposedly been known for at least 2000 years, and for the past
300 years, arsenic has found use as a poison for virtually all living things
including animals, plants, humans, intestinal parasites, and the bacteria as-
sociated with such diseases as syphilis and sleeping sickness.* ' Arsenical
compounds have, through the last few centuries, acquired reputations as stimu-
lants and tonics; they have been considered at times to be specific remedies
for anorexia, neuralgia, rheumatism, arthritis, asthma, chorea, malaria, tuber-
culosis, diabetes, and skin diseases. As recently as 1937, arsenical nvedicinals
accounted for about two-thirds of the 12-thousand organo-metal medicinals used
at that time.(27)
Pure metallic arsenic and arsenous sulfide have practically no toxic
effect on plants or animals, probably because of their extremely low solubility
in both water and body fluids. No toxic effects have been reported from the
handling of elemental arsenic.^ The most toxic of the arsenical compounds is
arsine (AsH3, hydrogen arsenite) and its methyl derivatives, mono-, di-, and
trimethyl arsine, all of which are gases having a characteristic garlic odor.
The toxicities of all other arsenical compounds fall between these extremes.
From the standpoint of chemistry and toxicology, the important compounds of
arsenic fall into three major categories:
1. Inorganic arsenicals - white arsenic (As^O-), arsenate
(As +5) salts, and arsenite (As +3) salts.
2. Organic arsenicals - the trivalent (As +3) arsenicals
generally have the greatest physiologic significance;
they may be mono-, di-, or trisubstituted; biological
action is a function of molecular structure.
3. Gaseous arsenic - arsine and the methyl derivatives of
arsine.
-128-
-------
The National Institute for Occupational Safety and Health estimates
1.5-million American workers are potentially exposed to arsenic. This number
includes people working in arsenic and nonferrous metals (especially copper)
production as well as agricultural personnel exposed to arsenical agricultural
products (including insecticides, herbicides, fungicides, and feed additives).
Other industries having exposure potential are glass manufacture, lead-acid
battery manufacture, wood preservative production, and nonferrous alloying.
Exposure Standards
The current Occupational Safety and Health Administration standard for
atmospheric exposure to inorganic arsenic (defined by OSHA as arsenic and its
inorganic compounds, except arsine) is 0.5 mg/fo , averaged over an 8-hour period.
The OSHA standards for lead and calcium arsenates are listed separately and are
3 3
0.15 and 1.0 mg/m , respectively. The current standard for arsine is 0.2 mg/m .
These standards (except for arsine) are based on the 1968 ACGIH (American Con-
ference of Governmental Industrial Hygienists) list of Threshold Limit Values
(97)
for Chemical Substances and Physical Agents in the Workroom Environment.
The ACGIH standards were based on the controversial study by Dr. Sherman Pinto -
then medical director of the American Smelting and Refining (ASABCO) plant in
Tacoma, Washington - where he concluded that no conclusive correlation exists
(98 99)
between arsenic exposure and respiratory cancer. ' Pinto*s study is dis-
cussed below.
OSHA has recently proposed new guidelines for workplace exposure to in-
organic arsenic; the maximum exposure would be 0.004 mg/m and an "action level"
would be 0.002 mg/m . Workers must be provided with protective equipment at
levels above the lower limit. Exposure limit for a 15-minute period would be
0.01
Standards for exposure to airborne inorganic arsenic compounds have
varied considerably over the past three decades. In 1943 the American Standards
3
Association reconnended a level of 0.015 mg/m . After WWII the War Standard of
0.15 mg/m was used. In 1947 ACGIH adopted a maximum acceptable concentration
of 0.1 mg/m , but in 1948 this was raised to 0.5 mg/m which is the value now
-129-
-------
prevailing. In 1974, NIOSH proposed a new standard of 0.05 mg/m , but after
Dow and Allied acknowledged studies indicating a possible link between exposure
to arsenic and higher than normal cancer rates, the standard was reduced to the
currently proposed levels stated above. ' ' The proposed new standards have
not been met with any enthusiasm by producers and users of inorganic arsenic. A
spokesman for ASARCO has pointed out that the proposed new limit is 650 times
lower than for vinyl chloride (on a mg/m basis) , and that though carcinogenicity
has been proven for vinyl chloride, it has not been proven for arsenic.
(NIOSH, on the other hand, is convinced arsenic is carcinogenic; this is dis-
cussed further below) .
The United States Public Health Service has established a recommended
maximum concentration of 10 ppb (0.010 mg/1) and a maximum permissible concentra-
tion of 50 ppb (0.050 mg/1) for arsenic in public drinking water; both of these
limits are well below the lowest reported concentration known to have resulted
in chronic poisoning - 0.21 mg/1.
Acute and Chronic Effects
Arsenic absorbed into mammalian bodies is excreted in the urine, feoes,
skin, hair, and nails, and possibly trace amounts are released through the lungs.
Arsenic, even in low dosages, tends to bind to keratin in skin, hair, and nails;
keratin is a class of fibrous proteins characterized by, among other qualities,
a high content of sulfur-containing amino acids. Arsenic bound to keratin is a
slow route of arsenic elimination - i.e. , via release of the metabolically dead
tissues; hair, skin, nails. Table 13 lists the "normal" arsenic content for
various tissues and fluids of the human body.
The major route of arsenic elimination is urine. Arsenic can be de-
tected in the urine of people with no known exposure to arsenic, apparently in-
gested in food (especially seafood) or through other low-level environmental
sources. The urine of workers exposed to arsenic may contain, and usually does
contain, much higher levels of arsenic, even though no other symptoms of exposure
may be apparent. Vallee, et al, cites the "normal" urine level of arsenic as
(27)
0.002 to 0.150 ppm. v ' In the NIOSH document, Criteria for a Recommended
Standard . . . Occupational Exposure to Inorganic Arsenic, reference is made
-130-
-------
TABLF 13
NORMAL AFSENIC CONTENT CF HUMAN TISSUES AND FLUIDS
(ppm, unless otherwise specified)
Whole Body
Urine
Blood
Nails
Hair
Reference
(27).
0.2 - 0.3
0.003 - 0.150
0.1 - 0.64
0.087 - 4.0
0.036 - 0.88
(1Q71
4-210 raog/
24 hrs
0.03 - 0.13
(29)
0.015 - 0.06
C4)
0.2 - 1.0
(96)
_
0.02 - 0.13
'
(111)
0 - 0.1
0.1
1.0
-------
to a study of 756 urine specimens from 29 people having no known exposure;
average level of concentration was 0.08 mg/1 with 79 percent of the samples
being below 0$. mg/1. The highest levels were 2.0, 1.1, and 0.42 mg/1, attri-
buted to seafood consumption. In another study of 26 adults and 17 children,
the average arsenic content of the urine was 0.014 mg As/1.* '
Seafood is generally considered the main source of arsenic for
"unexposed" people. In one test to establish the relation of seafood to urine
arsenic levels, three subjects with pretest levels of 0.01, 0.05, and 0.3 mg
As/1 were given lobster tail for lunch. Pour hours later urine levels were
b
1.68, 1.40, and 0.78 mg As/1, respectively. Ten hours after eating, urinary
levels were 1.02, 1.32, and 1.19 mg As/1, and after 24 hours the values were
0.39, 0.39, and 0.44 mg As/1. After 48 hours, the values were approaching the
pretest levels. ' ' Table 14 lists the arsenic content of various foods.
The excretion of inhaled arsenic has been studied experimentally.
Eleven terminal lung cancer patients inhaled the radioactive isotope As-74.
Uptake and distribution were measured with, a radiation counter. Within four
days, the lung level of arsenic had decreased to only 20 to 30 percent of the
initial level, and thereafter the rate of disappearance tapered off slowly.
1 t
About 28 percent of the inhaled arsenic was released in the urine in the first
day. By the end of 10 days, urinary and fecal excretion of arsenic was approach-
ing zero, with 45 percent having been excreted in the urine and 2.5 percent in
the faces. The remainder was assumed to have been deposited in the body, exhaled,
or eliminated over a longer period.
Interpretation of urine arsenic levels with regard'to previous exposure
or to individual tolerance for arsenic is difficult. Urinary arsenic levels of
exposed workers vary widely and levels above 4.0 mg As/1 have been reported
without apparent adverse effects; however, signs of mild systemic poisoning
have been reported in a worker excreting only 0.76 mg As/1. It has been con-
cluded that, while no relationship can be shown betatfeen urinary arsenic levels
and evidence of poisoning, urinary arsenic levels may well be used as a check
on the efficiency of control measures of arsenic in worker environments. '
-132-
-------
TabtelA
Arsenic Content of Various Foods
Pood
Fish
Haddock
Kingfish
Molluscs
Clams
Oysters
Smoked Oysters
Crustacea
Crabs
Lobsters
Shrirtp
Shrimp Shells
Pork Loins
Pork Kidney
Pork Liver
Stewing Beef
Chicken Breast
Milk, evaporated
Tea
Rhubarb
Corn
Corn Oil
Coffee
Wine Yeasts
Baker's Yeast
Egg Lecithin
Puffed Rice
Table Salt
Butter
Sugar
Lettuce
Oranges
Lemons
Rice and Wheat
Flour
Apples
Pears
Grapes
(118)*
0.1 - 15
2.17
8.86
1-68
0.0 - 2.94
0.0 - 400
(most 1.0)
10 - 79
0.2 - 7.0
75
0.3 - 7.7
15.3
0.06
0.0
1.4 - 1.07
1.3
0.0
0.17
0.89
0.48
0.11
0.0
0.0
0.0
1.6
2.71
Arsenic Content (ppn)
(105)*
15.9
16.0
45.8
25.0
22.1
19.9
(27)*
3
42
as high as 150
to 180
up to 17
(5)*
. 0.1 - 1.0
up to 200
up to 200
(119)*
6-45
46
37
24
0.07
0.15
1.14
0.22
0.50
0.96
0.08 - 0.60
0.40 - 0.60
0.75 - 1.20
*References
-133-
-------
Acute Effects
Symptoms of acute poisoning may occur as scon as 30 minutes after in-
gestion. Major early manifestations are burning and dryness of the mouth and
throat, dysphagia, colicky abdominal pain, projectile vomiting, profuse diarrhea,
and hematuria. Shock develops as a result of dehydration. If the patient sur-
vives, the recovery may be complicated by development of encephalitis, myelitis,
nephritis, or dermatitis. ' '
The fatal dose of arsenic trioxide for man is 70 to 180 mg., although
toxicity may result from much smaller amounts. Arsenical concentrations in
blood, urine, hair, and nails increase from 10 to 100 times normal in instances
(27)
of acute poisoning. Table 15 is a summary of toxicities of various common
arsenical compounds.
Arsine is the most toxic compound of arsenic; 250 ppm for 30 minutes has
been shown to be a fatal dosage, and 3 to 10 ppm can cause poisoning symptoms
in a few hours. Animals exposed for 3 hours a day to concentrations between
0.5 and 2 ppm have been shown to develop "blood changes" (unspecified) within a
period of several weeks. Typical arsine poisoning cases result in hemoglobinuria,
jaundice, and hemolytic anemia. Data on actual concentrations causing acute in-
toxication are lacking; however, post-event concentrations of 70 to 300 ppm,
5 ppm, and even as low as 0.5 ppm have been reported. Urine samples analyzed
at early stages of intoxication have contained arsenic concentrations ranging
from 0.5 to 2 mg/1 with occasional higher values being reported. The recom-
mended Threshold Limiting Value for arsine is 0.2 mg/m C0_063 ppml - less than
half the present limit of 0.5 ing As/fa for other inorganic arsenicals. ^104'
Arsine is the most dangerous form of arsenic and the most serious in
terms of industrial hazard. It has been referred to as the most powerful hemo-
lytic poison found in industry. Clinically, the resultant illness has sometimes
been referred to as "acid fume poisoning" or as "toxic jaundice',1. Arsine is
liberated whenever hydrogen is generated in the presence of arsenic; the element
may be a contaminant of either the metal or the acid used in the production of
hydrogen. Arsine evolution may also result from reduction of arsenious or
-134-
-------
TABU: is
TOKICITIES OF VARIOUS ORGANIC AND
INOBGAN1C ARSHttCAL COMPOUNDS
Compound
INORGANIC
Arsenic Acid
.
Arsenic Pentaxide
Arsenic Trioxide
Calcium Arsenate
' '
Ifiad Arsenate
Potassium Arsenite
Sodivm Arsenate
Sodiun Arsenite
ORGANIC
Caoodylic Acid
Manosodium Methanearsonic
Acid
Disodium Methanearsonic
Acid
Calcium Acid
Methonear sonata
Arsphenamine
Carbasone
(p-uridobenzenearsonic
acid)
Sodiun Arsanilate
Test Subjects,
Method of Intoxication
LD50 oral, rats
U)so oral, young rats
U>50 oral, old rats
U>50 oral, rabbits
ID i.v., rabbits
so
U>s rats
•1C 5° oral, rats
ID50 rats, mice
ID100 oral, man
ID oral, nan
Iti,0 (animal not specified)
ID,, oral, ratfl
50
ID nan
ID50 oral, rats
ICSO oral, rats
MID i.p., rats
I£j0 mammalian
M3L i.p. , rats
ID50 oral, rats
ICSO young rats
ID B.C., dogs
ID50 oral, rats
LDJo oral, rats
"'so
9 U
U)100 i.v., ratfl
ID5 oral, rats
LD° oral, rats
9 0
1C B.C., mice
Dose
(jngAg)
48 - 100
48
100
8
8
138
15
35-50
1 - 2.5
-1.4
35 - 100
20
10 - 50
100
14
50
10 - 50
10
75
830
1000 '
700
1000
4000
100
510
400
Reference
(5.2)
(103)
(103)
0-09)
(109)
(103)
-..(103)
(105)
(52) (110)
(111)
(52)
(52) .
(105)
(10D)
(103)
(52)
(10Q)
(103)
(3D (109)
(109)
(IOC)
(108)
(*?.)
(105)
(109)
. (109)
Abbreviations:
U>
LD
- lethal dose
50 - lethal dose for 50 percent of test animals
IO"0 - lethal dose for 100 percent of test animals
MIX) - minimum letlial dose
i.v. - intravenous
B.C. - subcutaneous
i.p. - intraperitoneal
-135-
-------
arsenic acid by means of nascent hydrogen, from electrolysis of arsenious solu-
tions, and from the action, of water or dilute acid upon metallic arsenides.
Dangerous quantities may even appear from the action of atmospheric moisture
(27)
upon arsenical-contaminated metallic sulfides.
Early symptoms of acute exposure include headache, anorexia, nausea,
vomiting, and paresthesia. Chronic exposure may be manifested by dyspnea on
exertion and palpitation resulting from the anemia. In large measure mortality
from arsine results from massive hemolysis. Survivors of acute arsine poisoning
usually regain a normal state after about tvro weeks, but residual EGG changes,
consisting of elevated T-waves in the procordial leads, have been reported to
persist for many months. If death occurs, it usually results from sudden heart
failure and pulmonary edema. At autopsy, the mucous membranes and serous surfaces
are found to be stained with hemoglobin, and myocardial and renal degenerative
changes have been observed. Arsenic tends to accumulate in the liver (up to 15
(27)
ppm), but large anounts are also found in the lungs and kidneys.
Chronic Effects
Polyneuritis and motor palsies may be the only manifestations of chronic
exposure. As in lead intoxication, weakness is most likely to affect the long
extensors of the fingers and toes. Arsenical neuritis is said to be more sym-
metrical, widespread, and painful than that seen with lead. Personality changes
may be included in the neurologic effects, along with headache, drowsiness,
memory loss, and confusion. Nerve biopsy specimens from neurologically affected
patients show degeneration. Chronic intoxication can also result in increased
salivation, hoarseness, cough, laryngitis, conjunctivitis, and abdominal pain.
Trophic skin changes with a purplish-red hue and smooth shiny finger tips are
(21)
frequently seen. *• '
The typical symptoms of severe chronic arsenicalism include nausea,
vomiting, diarrhea, hot flashes, and progressive anxiety. Such symptoms might
-136-
-------
(96)
continue intermittently. In one study cited in the NIOSH Criteria Document,
a worker exposed to arsenic for several years experienced a gradual darkening
of the skin, and a thickening and scaling of the skin on the soles of the feet.
An almost constant pain and feeling of pins and needles appeared first in the
feet and later in the hands. Muscular weakness became more apparent and the
extremities became numb in a glove and stocking manner. By three years after
the first symptoms, the skin of the trunk had darkened markedly, and there had
been a gradual loss of vision and increased pain. Attacks of the initial symp-
toms continued to occur three or four times annually for ten years, until the
patient was referred to specialists for management of severe heart failure and
muscular dystrophy. At that time, abdominal accumulation of fluid was evident
and severe ankle edema had developed. The patient was constipated except during
the episodes of nausea and vomiting, when he had diarrhea. He was emaciated
and had a diffuse tan pigmentation over the trunk. The palms and soles were
hyperkeratotic and Mees lines were present on the nails. All sensory functions
were diminished toward the extremities. The patient could not walk.
Urinary excretion of this patient was 0.140 mg/24 hours; the hair con-
tained 20.7 mg As/100 g of hair. The white count was low (2,174) with a slight
increase in monocytes. Both the EEG and BOG were normal. In an effort to in-
crease urinary excretion of arsenic, British Anti-Lewisite (BAL) was administered
but to no avail. After 3 months of hospitalization, functional use of the hands
returned and the patient could walk with the aid of leg braces and crutches;
urinary arsenic excretion was approximately 0.040 mg/24 hours. A follow-up at
one year revealed little, if any, improvement in the neuropathy. Deep tendon
reflexes were still absent and there was no proprioception beyond the knees or
/QC\
elbows. Pigmentation was still marked but the dermatitis cleared completely.
In a study of six patients exhibiting chronic arsenicalism, the symptoms
were, as above, nausea, vomiting, diarrhea, and peripheral neuropathy. In three
cases there was pigmentation, and in three cases there was hyperkeratosis of the
palms and soles. However, in two cases neither hyperkeratosis nor hyperpigmen-
tation were observed. Average urinary excretion was 1.87 mg As/1, with a range
of 0.348 to 3.46 mg AsA of urine. Arsenic in the hair averaged 4.88 mg AsAOO g
-137-
-------
of hair. Various blood abnormalities were evident such, as white cell counts of
less than 1000 (in three of the cases) and, in three of four of the patients
examined, and improper production red cells in the bone marrow. However, blood
abnormalities disappeared within several weeks.
Individual tolerance to arsenic intoxication varies considerably. Cer-
tain persons have reportedly been able to tolerate doses as high as 20 mg of
potassium arsenate three times daily without exhibiting signs of toxicity. The
"arsenic eaters" of Europe are reported to ingest as much as 400 mg of arsenic
trioxide once or twice a week without developing symptoms; and they experience
(27)
no withdrawal syndrome.
An allergic type of contact dermatitis is frequently seen where white
arsenic is handled. This dermatitis may be eczematous, follicular, erythematous,
or even ulcerative in character. In heavily exposed workers, mucous membrane
irritation, rhinorrhea, conjunctivitis, pharyngitis, and laryngitis are seen as
direct results of exposure to arsenic dust and are preventable with proper pro-
tective devices. Particulate matter absorbed into the nasal passages induces
inflammation and may result in ulceration and slough of cartilage leaving a 3-
(27)
to 8-mm punched-out area in the septum.
Accidental poisoning of agricultural animals and wildlife by solid ar-
senicals is reported occasionally, and it produces clinical syndromes and patho-
(27)
logic findings analogous to those in man.v
The NIOSH Criteria for a Reccmnended Standard . . . Occupational
/Qg\
Exposure to Inorganic Arsenicv makes reference to a 1945 study in which
medical records of workers in an arsphenamine plant were reviewed. Five types
of complaints were considered to be possible indicators of "subclinical or
borderline arsenicalism".
Hyperkeratosis - warts
cracking, chapped, dry, or thickened
skin
-138-
-------
Gastrointestinal - upset stomach
nausea
vomiting
abdominal pain
anorexia
Central Nervous System - headache
dizziness
fainting
Optic Nerve - blurring or diminution of vision
spots before eyes
Peripheral Neuropathy - shooting pains in extremities
numbness, tingling, sudden loss of
muscular power
Another symptom commonly associated with arsenicalism is hyperpigmenta-
tion. In one case cited in the NIOSH Criteria Document, 15 vinedressers and
cellarmen having symptoms of chronic arsenicalism had vascular disorders in the
extremities, and "all had varying degrees of hyperpigmentation and all but 2 had
palmar and plantar keratoses". Cold hands or feet or both were common to all and
apparently preceded the development of gangrene on the toes or fingers in 6 of
the 15 cases.(96)
Electrocardiograms also show changes possibly associated with arsenic
exposure. In a case where 170 soldiers had been chronically exposed to arsenic
in their drinking water, electrocardiograms were prepared for 80 of the soldiers,
45 of whom displayed abnormalities. Six weeks after the first examination, re-
peat ECG's were obtained in 47 cases, and the abnormalities initially observed
were absent or reduced. In another study of 192 vinegrowers suffering from chronic
arsenicalism, 56 percent had normal BCG's, 15 percent showed deviation from the
normal, but not sufficiently deviant to qualify as evidence of definite heart
muscle damage, and 29 percent showed definite changes - however, of this portion
approximately one-third (19 out of 55 men) of the BCG changes could also be at-
tributed to age, arteriosclerosis, or other disease. For abnormalities in the
-139-
-------
remaining two-thirds (36 men), no possible causes other than arsenic poisoning
could be detected. Follow-up examination showed a decrease in EGG abnormalities
in proportion to other symptoms of arsenic poisoning. Attempts have been made
to relate BCG changes to disturbances in serum electrolytes, but no relation
has been found; the changes are considered to be due to a toxic effect on the
(QC\
heart muscle.v '
Cirrhosis of the liver has also been associated with chronic arsenic
exposure via prolonged use of Fowler's solution (a dilute solution of potassium
arsenite previously used as a treatment for leukemia and various skin diseases).
Use of Fowler's solution has also been linked with "generalized mottling and
bronzing of the skin, palmar and plantar hyperkeratoses, ascites, and marked
edema".
Among workers exposed to inorganic arsenic, especially as airborne dust,
the chronic symptoms cartitionly found are perforation of the nasal septum, conjunc-
tivitis, and pharyngitis. There is reportedly a large degree of skin sensitivity
variation among arsenic workers; however, sensitivity of the skin to airborne
inorganic arsenicals is very common in moist skin areas or in areas where rub-
bing or chafing of the skin obcurs such as areas around the eyes and wrists,
or in facial areas where a respirator is likely to rub against the skin. Blond
and fair-skinned people have been reported as being especially sensitive to
arsenically induced dermatitis.
In one study cited in the NIOSH Criteria Document, dust-in-air measure-
ments were considered of limited value in predicting skin reactions, as were
levels of arsenic in urine; however, based on a study of 127 patients, dermatitis
was observed in 80 percent of those excreting 1 to 3 mg As/1 and in 100 percent
(96)
of those excreting more than 3 mg As/1.
the most controversial aspect of chronic arsenicalism is cancer and the
possibility that arsenic might be carcinogenic. Findings of excess cancer deaths
among workers chronically exposed to airborne concentrations of various inorganic
arsenicals have duplicated inorganic arsenic as an occupational carcinogen. Re-
sults of a number of studies have especially shown arsenic trioxide, lead ar-
senate, calcium arsenate, and sodium arsenite to be suspect carcinogens.
-140-
-------
In 1963, Pinto and Bennett analyzed the causes of death of 229 copper
and arsenic smelter workers at ASARCO's Tacoma plant, and on the basis of the
average urinary arsenic levels divided the workers into "exposed" and "non-
exposed" groups. Pinto and Bennett concluded that there was no significant
difference in the rates of cancer for the two groups. The findings in this
study became the basis for the present Federal standards for inorganic arsenic
exposure after they were accepted by the American Conference of Governmental
Industrial Hygienists, which until 1970 was the only organization setting stan-
dards for exposure to dusts and fumes in the workplace. With the passage of
the Occupational Safety and Health Act of ,1970, Pinto's findings were still
3 (98)
used as the basis for the still-prevailing standard of 0.5 mg As/m of air.
However, substantial controversy has come to surround the Pinto study during
the last few years.
In a 1972 study by Milham and Strong, urinary arsenic levels of children
living near the Taooma smelter were measured and correlated to the distance the
children lived from the Tacoma smelter.
Blood lead and urinary arsenic levels of third-
and fourth-grade children at Ruston School (located about
300 yards from the west border of the smelter complex) were
compared to those of similar students at another elementary
school about 8 miles away. Blood lead levels were essentially
the same for the two groups of children, but arsenic urinary
levels were considerably elevated among the Ruston children.
Hair specimen containers were sent home with children
at the end of the school year and were returned over the summer.
Hair arsenic levels were very high for Ruston children, averag-
ing over 50 ppm while the control school children averaged less
than 3 ppm.
A few weeks after the initial study, urines were sampled
along three downwind traverses starting at the smelter stack
and extending nearly 3 miles south and southwest. There was a
decline in urinary arsenic levels with distance from the stack.
-141-
-------
The findings of Milham and Strong suggest that Pinto's "nonexposed"
group/ since they did indeed work within the smelter complex along with the
"exposed" group (who worked in the actual smelting operations, as opposed to
office operations where the "nanexposed" group prevailed), probably had a sub-
stantial exposure to airborne arsenic trioxide dusts.
In 1969 Lee and Fraumeni studied the mortality statistics of white
male workers at the same ASARCO Tacoma plant for the years 1938 .through 1963,
and compared the results to the expected mortality rates for the general popu-
lation of the state. "The excess of respiratory cancer was as high as eightfold
among employees who worked more than 15 years and who were heavily exposed to
arsenic; it showed a gradient in proportion to the degree of exposure to arsenic
and sulfur dioxide. The findings support the hypothesis that inhaled arsenic
is a respiratory carcinogen in man, but an influence of sulfur dioxide or un-
identified chemicals, varying concomitantly with arsenic exposure, cannot be
discounted". Lee and Fraumeni also noted that "among the specific causes of
death, tuberculosis, respiratory cancer, diseases of the heart, and cirrhosis
of the liver showed a significant excess over expectation", based on mortality-
by-disease for the state as a whole. But, as they point out, it is difficult
to separate the effects of combined exposure to both As203 and SCL.
Animal experiments on the carcinogenicity of arsenic have generally
given negative results. Studies of the co-carcinogenic effects of arsenates
and arsenites with such materials as cotton oil, urethane, and dimethyl-
benzanthracene have also been negative.t11'10-5) However, in the summer of 1974,
Dow Chemical Company and Allied Chemical Corporation acknowledged that workers
in their inorganic arsenic pesticide plants were dying of lung cancer at 7 times
the expected rate, and of lymph cancer at 6 times the expected rate. As a re-
sult, some officials in NIOSH are now comparing industrial exposure to arsenic
(96 98)
to that of vinyl chloride. ' Some 15 copper, lead, and zinc smelters ship
their arsenic-containing flue dusts to the ASARCO plant at Tacoma where white
arsenic is produced. In total, about 40 different industries use white arsenic
(98)
in their manufacturing processes.v '
-142-
-------
The carcinogenicity of arsenic has not been proven in animal studies.
Even in combination with known carcinogens, animals exposed to various compounds
of arsenic in their drinking water showed no increase in cancer rate over that
expected for the non-^arsenical carcinogen alone. However, the relation between
arsenic and cancer in humans is considered by some to have been proven, especially
by studies of worker populations exposed to inorganic arsenicals. The proposed
standards for arsenic exposure cited in the NIOSH Criteria Document are based on
an assumed carcinogenicity of arsenic. The last two sentences of the Criteria
Document state that "because of the seriousness of [cancer] , prudence dictates
that the standard should be set at least as low as 0.05 mg As/m . It is be-
lieved that exposure at this level should, at the minimum, significantly reduce
the incidence of arsenic-induced cancer" . The proposed standards in the
1973 Criteria Document have since been further reduced to 0.004 mg As/m with
an action level of 0.002 mg As/m . In supporting the original proposed standard
of 0.05 mg As/m , NIOSH cited as evidence of the carcinogensis of arsenic three
epidemic-logical studies, two of which were made with respect to the ASAROO smelter
in Tacoma, Washington, while the other study was performed on workers in an
English sheep dip factory. One of the studies of the Tacoma smelting complex
and environs is the Lee and Fraumeni study where they state:
Arsenic has been suspected by many investigators as
a carcinogen in man, though there is no supporting evidence
from animal experiments. Skin cancer appears to be a definite
consequence of arsenic exposure among individuals exposed to
inorganic arsenic in industrial dusts, medicinals, and drinking
water. Less convincing is the clinical evidence suggesting
that long-term exposure to arsenic may give rise to cancer of
internal organs, notably the lung.
Lee and Fraumeni also point out in their study, "the greatest excess of
respiratory cancer occurred among smelter workers with high exposure to arsenic
accompanied by high or moderate exposure to S02. Although no studies implicate
S02 as a carcinogen in man, possibly this agent enhances the supposed carcino-
genic effect of arsenic or other substances. From laboratory experiments,
-143-
-------
inhalation of the known carcinogen benzo[a]pyrene, combined with the irritant
SO- produced squamous cell carcinomas of the lung in rats, whereas inhalation
of the carcinogen alone did not produce tumors. . . Our findings are consistent
with the hypothesis that exposure to high levels of As20_, perhaps in inter-
action with S02 or unidentified chemicals in the work environment, is responsible
for the excessive number of respiratory cancer deaths among smelter workers".
In the NIOSH-cited study of the English sheep dip workers, S02 was ap-
parently not involved in the worker exposure, "but the cancer mortality of the
(96)
chemical workers was significantly higher [than the control group]". '
The latest proposed standards, still under consideration at this time,
are based on the belief that exposure to airborne concentrations of inorganic
arsenic compounds are "strongly implicated as a cause in occupational carcino-
(97)
genesis".v ' Ten epidemiological studies are cited by OSHA as the basis for
this strong implication. Six of the studies show evidence of excess lung cancer
mortalities among worker populations having had exposure to inorganic arsenic
compounds. The authors of the other four studies concluded that there was no
significant excess of cancer mortalities among inorganic arsenic workers; how-
ever, in the analysis of three of these studies, both NIOSH and OSHA confirmed
that excess lung cancer mortalities were involved, but were not observed due to
inadequate study designs. (No definitive conclusions could be assigned to the
fourth study.) "Most of the available studies, including the data submitted by
Dow and Allied, do show significant excesses of lung cancer mortalities for work-
(97)
ers exposed to a variety of inorganic arsenic compourds".v There is no evi-
dence implicating the ingestion of organic arsenic as a cause of cancer.
Mode of Action
Trivalent arsenic can chemically combine with the sulfhydryl groups;
such groups are cormonly found in proteins. Enzyme deactivation can thus result
from the affinity of arsenic for the sulfhydryl groups which enzymes contain.
It has been demonstrated that trivalent arsenical toxicity can be reversed by
administering reduced thiol compounds, such as glutathione and cysteine. The
combining of arsenicals with tissue proteins and enzymes has actually been
shown to be accompanied by a loss of titratible sulfhydryl groups. It has also
-144-
-------
been shown that there may be a direct correlation between the number of sul-
(27)
fhydryl groups in an organism and its sensitivity to arsenical intoxication.
Studies involving the mode of action of the arsenical war gas lewisite
have resulted in the determination of a large number of enzyme systems sensitive
to arsenicals. An excess of simple thiol protects a variety of biological sys-
tems against the toxic inhibition of both organic and inorganic arsenicals - but
not uniformly throughout an organism; specifically a 200-percent monothiol ex-
cess fails to protect the cerebral pyruvate oxidase enzyme system from lewisite,
although other enzyme systems are completely protected by smaller concentrations.
Investigations of this problem determined that lewisite reacts with some proteins
in such a way as to bind two thiol groups, forming a stable compound not freely
reversible with monothiols. The protective action of various dithiol compounds
was therefore studied, and one compound, dimercaprol (2,3-dimercaptopropanol,
also known as British-Anti-Lewisite or BAL) was found to be an effective antidote,
even for protection of the pyruvate oxidase system of the brain. * '
In addition to the affinity of trivalent arsenic for tissue sulfhydryl
groups, arsenic may interact with biologic systems through other means. Ar-
senate or arsenite may compete with or substitute for phosphate in certain
enzymatic reactions.
In animals a direct relation between toxicity and strength of binding
to tissues has been shown for a large series of phenyl arsenoxide compounds.
Less firmly bound compounds are excreted more rapidly, and are less toxic at
comparable levels of administration. At dosages producing equavalent toxicity
(e.g., LDeg) tryparsamide, phenyl arsenic acid, and phenyl arsenoxide result in
comparable tissue arsenic concentration despite a 500-fold difference in absolute
(27)
amounts of arsenic administered. '
Arsenic is said to be a physiologic antagonist of iodine. The addition
of 0.02 percent arsenic to the diet of rats has been shown to more than double
(96)
their iodine requirement.v A high incidence of goiter and cretinism has been
reported among the so-called "arsenic eaters" of Europe and among dwellers in
(27)
the endemic zones of arsenical intoxication in the Cordoba province of Argentine.
-145-
-------
Arsenic is also an antagonist of selenium. Agricultural animals ex-
posed to selenium toxicity through forage in seliniferous areas are protected
f96)
by small amounts of arsenic (5 to 10 ppm) in their drinking water. Tungsten
(27)
is the only other element known to provide such protection against selenium.v '
Inorganic arsenic does not cross the blood-brain barrier in humans, though
it may do so in some anthropoids. In man and rats arsenic is transferred across
(27)
the placenta. It appears in cows' milk, but not in rodent milk.v '
The major toxicity of arsine is due to the hemolysis of the red blood
cells, but the exact reason for this effect is unknown. It occurs only under
aerobic conditions and involves only mature cells. Neither arsenic trioxide
nor arsenic pentoxide has this effect. Guinea pigs cronically exposed to arsine
(0.5 to 2 ppm) exhibit increased red cell fragility, leukopenia, and a rapid
fall of red cells to a stable level, roughly 80 percent of normal. The toxicity
of arsine and its clearance from the bodies of mice has been compared to that of
sodium arsenite; where arsenite is cleared exponentially from the animal with less
than 10 percent remaining after 24 hours, arsenic derived from arsine is cleared
(27)
more slowly, with about 45 percent remaining after 24 hours.
Oxidation State vs* Toxicity >''
Generally, but not invariably, inorganic arsenicals are more toxic than
organic, and trivalent arsenic is more toxic than pentavalent. Pentavalent ar-
senic, probably because of its lower affinity for thiol groups in protein struc-
tures, is excreted faster than trivalent arsenic, though evidence of rapid
excretion of all arsenicals has been shown. Pentavalent arsenicals, "although
physiologically inactive in this form", rapidly penetrate all parts of the body,
including the central nervous system. They are excreted otherwise unchanged,
but some tissues can reduce small amounts to trivalent arsenoxides, which can then
(27 59)
damage otherwise inaccessible cells. ' ' There is also evidence of in vivo
(96)
oxidation from trivalent to pentavalent forms, ' cited as a possible means of
natural detoxification. At least 15 strains of bacteria have been identified
which can oxidize trivalent arsenic (specifically sodium arsenite) to pentavalent
forms (arsenate); it is hypothesized that the bacteria somehow derive energy from
-146-
-------
the reaction. Inorganic trivalent arsenicals are cited as being between
5^ and 60^ times as toxic as pentavalent arsenicals in humans.
While arsenites are 10 to 60 times more .toxic to human and animals than
arsenates, arsine is even more toxic than arsenites. Methyl arsines are also
extremely toxic; trimethylarsine is the gas which was discovered in the end of
the last century to be the agent responsible for instances of mysterious deaths
reported in damp homes in Europe. Hie volatile trimethylarsine gas was geing
produced by the action of mold and dampness upon the arsenical-containing wall-
paper pigments. ' '
•The characteristic garlic-like odor of arsine and its methyl derivatives
has been found in many industrial and agricultural settings, especially in metal-
finishing industries where arsenically contaminated reactions between acids and
(27)
metals take place. Methyl arsines can only occur as a result of microbial
activity in both aerobic and anaerobic environments. ' Workers using caco-
dylic acid to control vegatation in forested areas have reported the charac-
teristic garlic odor, within as little as 48 hours after the thinning operation
out
(5)
and lasting for as much as three weeks. ' Virtually nothing is known about
the stability of methyl arsines with respect to oxidation in air and water.
However, although trimethylarsine is considered insoluble in water, it is signi-
ficantly more soluble in hydrocarbons, which may account for its accumulation in
certain organisms. ' Paradoxically, arsenic in shrimp (probably as also in other
marine life forms) is probably in the form of trimethylarsine, and when consumed
by rats is excreted much more rapidly and is much less toxic than arsenic tri-
oxide; this implies that trimethylarsine in food is much less toxic than in air.
Organic vs.. Inorganic Arsenicals
Methyl arsines are organic arsenicals but as such they are exceptions to
the general rule that organic forms are less toxic than inorganic forms; in fact,
they are as toxic as unmethylated arsine, AsH_, which, as has been pointed out,
is the most toxic arsenical compound. The general rule is that organic arsenicals
are between 10 and 100 times less toxic than inorganic arsenicals.
-147-
-------
As pointed out above, the organic arsenical trimethylarsine, when eaten
in shriitp meat, is not retained in the bodies of rats; it is excreted in the
feoes. Other organic arsenicals for which this is true are notably the four
feed additives used to improve feed efficiency and growth rate of poultry and
swine. These four aromatic arsenicals are arsanilic acid, 3-nitro-
4-hydroxvphenylarsonic acid, 4-nitrophenylarsonic acid, and p-ureidobenzenearsonic
acid. Table 16 lists the dosage levels and maximum allowable tissue levels for
the compounds.
The metabolism of ingested arsenic from the arsenical feed additives
has been investigated by a number of researchers. Chickens excrete arsanilic
acid largely unchanged; there is no evidence it is converted into any other
organic arsenical or to an inorganic form. Pour-nitrophenylarsonic acid, how-
ever, is converted to arsanilic acid, and 3-nitro is partly converted to 3-amino-
4-hydroxyphenylarsonic acid, but there is no evidence it is converted to an in-
organic form. In both poultry and swine, a high percentage of ingested arsenic
is excreted very rapidly. In a 5-day "balance trial" with growing sheep, 87
percent of all ingested arsenic was excreted. Tissue levels of arsenic in
arsenically-fed animals drop to well within the FDA-established tolerance levels
within the 5-day withdrawal period required before the animals go to slaughter.
In a study of arsenically-fed chickens - 50 ppm of 3-nitro for 70 days - tissue
(59)
levels of arsenic after 5- and 14-day withdrawal period were as follows;: '
Tissue level after (ppm):
kidney
liver
muscle
skin
70 days
of feeding
0.64
1.26
0.04
0.05
5-day
withdrawal
0.10
0.43
0.01
0.02
14-day
withdrawal
0.08
0.19
0.02
0.03
Controls
0.05
0.08
0.02
0.02
FDA Tolerance
2.0
2.0
0.5
0.5
-148-
-------
TABLE 16
Maximum Permissible Levels of Arsenicals in Animal Feeds
and Maximum Permissible Levels of Arsenic in Animal Tissue
VO
Compound
Arsanilic Acid
3-nitro-4-
hydroxyphenyl-
arsonic acid
4-nitrophenyl-
arsonic acid
p-ureidcbenzene-
arscnic acid
Species
Poultry*
Swine
Poultry
Swine
Turkeys
Turkeys
Maximum
Feed Level
90g/ton(100
mgAg)
90g/ton(100
mgAg)
45g/ton(50 mgA<
68g/ton(75 rrg/ki
170g/ton(187
mgAg)
340g/ton(375
mgAg)
Maximum
Tissue Arsenic Level
0.5 mgAg fresh,
uncooked muscle
2.0 mgAg fresh,
uncooked by-
products
0.5 mgAg fresh muscle
and by-products
other than kidney
& liver
2.0 mgAg fresh,
uncooked kidneys
liver
j) Same as arsanilic acid
j) Same as arsanilic acid
Same as arsanilic acid
Same as arsanilic acid
*Broilers, laying hens and turkeys.
-------
SECTION VII
ASSESSMENT OF HEALTH HAZARD
Arsenic is the most well known of the toxic elements, but the magnitude
of its reputation as a poison exceeds its level of potential hazard to the
general population. The greatest threat of arsenic to public health is in those
parts of the country where nonferrous smelting operations emit arsenic fumes
which cause an overall increase in the local (up to 10 to 15 miles) environmental
concentrations of arsenic.
Workers as well as people living in the vicinity of smeltering and re-
fining facilities are potentially affected and it is now generally conceded by
industrial producers and users of arsenic compounds that arsenic stimulates a
(99 112)
higher incidence of cancer than is found in the general population.***'*• *•'
The atmospheric concentration of arsenic in the area near one smelting
facility averages 2.3 pg/m over a 24-hour period, which is greater than the new
proposed OSHA standard of 2 yg/m for an 8-hour period in the workroom environ-
ment. The 2.3 yg/m exposure corresponds to an annual pulmonary absorption rate
3
(based on 20 m of air breathed daily, and an assumed 100-percent adsorption of
the entrained arsenic) of 16.8 mg for each adult in the local population, to
which is added, of course, the exposure from other sources such as food or water
which may have been contaminated by local high concentrations of arsenic.
The current worker exposure standard is 0.5 mg/m , which corresponds to
an annual pulmonary absorption of 3650 mg - more than 20 times the single lethal
dose level of arsenic in the form of arsenic trioxide. At this current exposure
level workers are experiencing increased rates of cancer.
Cancer is the biggest issue facing arsenic-dependent industries during
the last decade. The carcinogenicity of arsenic has been an active matter of
debate for more than half a century, and though industry is beginning to acknow-
ledge the findings of independent researchers showing that arsenic-exposed workers
/go 112)
face increased lung cancer risk, ' ' the debate continues, largely because
animal studi.es have not shown a relationship between arsenic and cancer. The
premise of the proposed new OSHA standards for workroom concentrations of arsenic
(96 97)
is that arsenic is a carcinogen. '
-150-
-------
At the current levels of exposure, some workerB experience dermatitis
on the moist areas of their skin and in areas where chafing of the skin is com-
mon, such as the point of contact between face masks and the face. The tissues
of the lung are constantly moist and they present a large surface area to arsenic-
laden air. Arsenic in smelter polluted air is in the trioxide form, one of the
most toxic of the inorganic arsenical compounds, and, because it is trivalent, it
presents those hazards especially associated with the affinity of trivalent ar-
senical compounds for sulfur-containing proteins. The moist condition of the
lung seems tailored to optimization of the toxic hazard of trivalent arsenic tri-
oxide to the delicate lung tissues.
Local populations and workers are also exposed to arsenic in higher than
natural concentrations in areas adjacent to industries producing arsenical pesti-
cides and other products out of powdered arsenical raw materials which are sub-
ject to dusting and becoming airborne; but the geographic area of exposure is
estimated as being 2 to 3 orders of magnitude less hazardous (in both geographic
extent and atmospheric concentration) than areas adjacent to nonferrous smelting
and refining facilities.
Persons living or working in areas where cotton is ginned face a possi-
ble exposure hazard from the airborne arsenical dusts generated by the ginning
process. Incineration of cotton gin trash also releases arsenic to the local
environment, though again, the amount is small in comparison to that released by
/Q\
smelting operations. ' (There are approximately 3000 ginning facilities in the
U.S., versus about 50 smelting and refining facilities.)
For the general population, the sources of arsenic exposure are mainly
food and drinking water, neither of which presents a hazard. The arsenic content
of seafood is higher than that of other foods but the arsenic is apparently of a
form that is rapidly excreted via the kidneys.
A large portion of the population faces potential exposure through the
use of arsenical pesticides, especially in the Texas-Oklahoma area where arsenic
acid is widely used to desiccate cotton prior to machine harvesting ("stripping").
-151-
-------
This level of exposure is approximately equivalent to that of city dwellers
(9)
living in the range of fallout from coal-burning power plants. (In one study
of arsenic pollution from coal combustion, it was calculated that a "standard
(21)
man" would breath about 0.5 mg of arsenic per year - less than one one-
hundreth of the amount necessary to produce a minimum single-dose toxic effect.)
Though arsenic has the reputation of being a cumulative poison/ the
human body does have mechanisms for controlling the body burden at moderate dose
levels. Inorganic arsenic (trivalent) accumulates about 20 times as rapidly
as arsenic which has been incorporated into such food organisms as shrimp,
chicken, or swine. Arsenic is removed from the body mainly through the kidneys.
If tihe dosage level is on the order of that experienced by smelter workers, then
urine remains the main node of arsenic loss, but the skin, hair, and nails ac-
cumulate excess arsenic which is eventually lost during normal tissue growth and
replacement processes. There is some evidence of long-term storage in bones,
but in no way does arsenic compare to such a material as cadmium which progres-
sively accumulates in the kidneys, with virtually zero loss.
Arsenic has been compared to mercury as a water pollutant; both are
subject to microbially-mediated chemical cycles involving methylation. But
unlike methyl-mercury which is absorbed readily at progressively increased con-
centrations up the food chain, arsenic does not undergo such a biomagnification.
Also, as mentioned above, arsenic in food organisms, even in high concentrations,
does not present any health threat yet identified.' ' ' '
The remainder of this section discusses in greater depth the health
implications of arsenic in air, water, and food, and the methods by which, and
sources from which, arsenic gets into these three important consumables. A
discussion of soil-pesticide interactions and of crop uptake is also included.
Arsenic in the Air
Arsenic pollution of the air derives largely from three sources: Non-
ferrous metals smelting, coal combustion, and cotton trash incineration. Lesser
sources include industrial operations where arsenical dusts are agitated into
atmospheric suspension, emissions from pesticide applications (including evaporation
processes), and emissions from the incineration of pesticide containers.
-152-
-------
Atmospheric emissions of arsenic fran agricultural operations are
probably responsible for the largest geographic distribution of arsenic in the
United States. However, coal, because it is used so widely as the energy source
for many urban electric power plants, probably provides the largest population
distribution of arsenic. The other emission sources, nonferrous smelting and
cotton trash incineration, are largely local problems.
Pesticides
In agriculture, spray applications of arsenical pesticides and herbi-
m'des may produce potentially hazardous exposure for nearby personnel - workers
as well as local populations. This is especially true in parts of the country
where machine-stripped cotton is grown, as in Texas and Oklahoma. Arsenical
herbicides control weeds in the early part of the crop growth and arsenical
desiocants and defoliants are used to prepare the cotton plant for harvesting.
Arsenica] ly-desiccated cotton is grown almost exclusively in the Texas-Oklahoma
region. Arsenica] 1 y-defoliated cotton is grown largely in the 11 other cotton-
producing states.
(q\
The National Inventory of Sources and Emissions; Arsenic - 1968v '
estimates that in 1968 the total atmospheric emissions of arsenic due to pesti-
cides was 2973 kkgf including 17 kkg from cotton gins and 296 kkg from the in-
cineration of cotton gin trash. The amount of arsenic used in pesticides has not
changed very much in total amount since 1968. The increased use of arsenic acid
in the preharvest desiccation of cotton matches closely with the decreased use
of arsenic in other agricultural uses. The emissions factor given in the National
Inventory for the burning of cotton gin trash is 7*7 kg/1000 bales of cotton
ginned. For the 10,857,000 bales ginned in 1968 the amount of arsenic emitted
amounted to about 84 kkg. The peak year for cotton production was 1972 when 13
million bales were produced; in 1974 the amount of cotton was 11.5 million bales,
and the average annual production between 1968 and 1974 was 11.2 million bales.
Thus, on the basis of cotton production alone, the amount of arsenic emissions
from ginning and incineration of cotton trash would not have changed very much
between 1968 and the present. However, the portion of cotton upon which arseni-
cals were used has increased; arsenic acid production has increased by a factor
-153-
-------
of about 5 on the same interval, but its use is restricted to the Texas-Oklahoma
area where the type of cotton grown (accounting for about 40 percent of total
productionX requires desiccation prior to harvesting. A reasonable factor
of increase of overall arsenic use in cotton production would be 2, and assuming
no increase in emissions control from cotton ginning and trash incineration, the
amount of arsenic emitted to local populations living near ginning facilities
would be about twice the figures established in the National Inventory for 1968.
(Methanearsonates are used as selective herbicides throughout a large portion of
the entire cotton industry, and cacodylic acid is the defoliant most widely used
as a cotton harvest and in states other than Texas and Oklahoma.)
There are approximately 3000 ginning facilities distributed throughout
the southern and eastern states. The largest number of bales processed at any
one facility is about 10,000, corresponding to a potential local emission (based
on twice the emission factor given in the National Inventory) of 32 kg. Emissions
due to the incineration of gin trash (again, based on tfcLce the National Inventory
emission factor of 17 Ib As/1000 bales ginned) would be, for a 10,000-bale facility,
155 kg, assuming no emission controls.
With regard to emissions due to other arsenical pesticides, the total
amount of pesticides used annually has not changed very much and the National
Inventory figure of total emissions of about 3 kkg would still apply.
Coal
The National Inventory gives the average, arsenic content of American
coal as 5.44 ppm - ranging from a high for eastern coal of 9.95 ppm to a low of
1.18 ppm for coal from the western states. Approximately 450 million metric
(18)
tons of coal per year have been consumed, implying that 2,450 metric tons
per year of arsenic are associated with this coal.
During combustion, part of the arsenic is released with fly ash and part
of it stays with the bottom ash. Measurements of arsenic in stack gases from coal-
fired power plants (after fly ash collection) give a range of concentrations of
-154-
-------
0.021 mg/m to 0.3640 mg/m (volumes measured at standard temperature and pres-
3
sure), with an effective value of about 0.1456 mg/m . Assuming (after the
National Inventory) that 9.97 m of flue gas are generated by each kilogram of
coal consumed, the arsenic emissions are about 650 metric tons per year. This
is slightly more than one-quarter of the arsenic calculated above to be in the
coal.
In an English study cited by Vallee, et al., air measurements of arsenic
were made and ranged from 0.03 to 0.105 yg/m . It was estimated that the "standard
man" exposed to such concentrations would inhale about 0.5 mg/yr. Vallee, et al.,
also point out that in studies of dust taken from inside of buildings in towns
where large amounts of coal are consumed, the "content of copper, lead and zinc
in these dusts was much greater than that of arsenic (50 to 400 ppm) and correlates
(27)
to the content of these metals in coal."v '
Nonferrous Smelting
Arsenic is produced as a by-product of nonferrous smelting. But it is,
for most smelters, considered an impurity which presents a disposal problem. At
the present tine, there is no economic incentive for most smelters to remove ar-
senic from their flue gases. In areas such as Arizona where copper is produced
and the arsenic content of the ore is relatively low, virtually no effort is made
to control emissions. But in areas where high-arsenic ores are smelted, as in the
Pacific Northwest States, controls have been required, and Federal Legislation
has been directed at making such controls even more stringent.
Atmospheric emissions from nonferrous metals smelting is a local problem,
confined generally to distances of no more than 10 to 15 miles of the approximately
40 nonferrous smelting and refining facilities processing the ores of copper, leal,
zinc, and gold. All of these ores, especially those mined in the Pacific Northwest,
contain arsenic, and unless adequately controlled, arsenic trioxide fumes from
smelting facilities enter the air and settle gradually to earth, finding their way
onto local grazing and crop lands and onto animals and crops, thus setting the
stage for later ingestion in food. And to the extent that such arsenic fumes and
dusts are washed into or otherwise moved into water supplies, local - and to some
-155^-
-------
extent even distant - water can come to carry toxicologically significant amounts
of arsenic.(115)
Dusts collected from buildings located near smelters have been shown to
contain significant amounts (hundreds of ppm) of arsenic trioxide. To the ex-
tent such dust can be inhaled or otherwise inadvertantly ingested, it presents
a potential hazard.
Tacoma, Washington, and the Helena Valley in Montana are two areas where
researchers contend that a public health threat exists due to nonferrous smelting.
Nonferrous smelting, of course, takes place in other parts of the country as well,
but these two regions have the disadvantage of being close to population centers*
and they process high-arsenic-content ores. In a report of a government sponsored
study of the Helena Valley, ' it is stated that arsenic, cadmium, and lead,
"which are emitted as air pollutants from (the two plants in the region), settle
and accumulate in soil and on vegetation to an extent surpassing levels that are
toxic to grazing farm animals. Furthermore, evidence indicates that subclinical
effects could be occurring in humans". The study points out that though the
average soil content of arsenic is normally about 5 ppm, and the upper 4-inch
layer of soil outside the Valley has a geometric mean arsenic content of 6 ppm,
the concentration in the upper 4-inch layer within a mile of the smelter complex
averages 50 ppm, and sometimes measures up to 150 ppm.
Hie Tacoma, Washington area is by far the more controversial of the two,
especially since it contains the only arsenic refining facility in the nation.
A spokesman for the facility at Tacoma has conceded that arsenic exposure has
been associated with lung cancer among smelter workers, but that the atmospheric
concentrations in the areas adjacent to the facility are "hundreds and even
thousands of tines" less than in the actual smelting complex. Researchers with
that State's Department of Health have, however, found that children who go to
school near the plant excrete arsenic in their urine "at about the same level as
the workers at the smelter".^112*
Table 17 is a summary of ambient atmospheric concentrations of arsenic
for various urban and rural environments. Seven of the 13 measurements listed
-156-
-------
TABLE 17
OF AVAILABLE AMBIENT DMR FOR ARSENIC
(19)
Gonoentraticn
-J
I
Site Description
Nationwide (urban
and nonurban)
Urban, rural,
source orien-
tated
Specific measure-
ments from EPA
Data Bank of
unusually high
concentration
Tacona, Wash.
Center City
(70 ft. high
sancler)
Near ASARCO
Copper
snelter
•fear ASARCO
Copper
shelter
Near ASASCO
Copper
snelter
A jo, Arizona
Subucban/In-
dustrial (3
ft. high
sanpler)
Annaconda,
Montana
' Center City
(26 ft. high'
sarpler)
El Paso, Tx.
Nat'l Study
of 0.3. Cities
Year(s)
1964-65
Duourfi
April
1974
1965
1973
(Jan.
to
Mar.)
1973
(Aug.
to
Oct.)
1974 4'
Oul. •
to
Aug.)
1972
1962
1964
1953
Monitoring Agency
NASH
State Agencies of
Montana, Washing-
ton, Arizona,
Colorado
Washington
State Dept.
of Zoology
Puget Sound
CuiLcol Agency
Puget Sour-d
Control Agency
Univ. of Hash.
Arizona State
Department of
Health
JfaTtar-a State
Board of
Health
NAS8
T.i ^ 1 rqi a»mijii y
(\tq/ig?) No. of Observations
0.02 Average of yearly
averages for
154 locations
0.13 Average of max.
hr observations
24-
for
97 sites, 53 loca-
tions
1.3 88
4.16 NA
15.9
1.16 28
3.9
2.3 40
5.3 NA
4.17 18
2.5 23
1.4 NA
0.033 NA
Averaging Time
Quarterly Composite of
24-your values
24-hour observations
Analytical Itethod
Spectxoscopy
Bm'nsicn Spectroscccy
Atomic Absorption,
Arsine Co]crimetrie
Maxunaa 24-hour value
observed
3 month average of 24-hr
concentrations (selected
filters)
maxima value recorded
(24-hr)
3 north average of 24-hr
concentrations (selected
filters!
raaxisuE value recorded
(24-hour)
Average of 24-hour
Bnission Spectrosoopy
24-hour value
observed
24-hour value
observed
Atomic Absorption
Atomic Absorption
Arsine Colorimetric
KsadsiMi 24-hour value
observed
posite
Aocual awarage of 24-hr
«?ppl^ff (confiositie of
1£ cities)
Ririation SpectxoBooR'
Erdssion SpectroBCOpy
NA
-------
exceed the proposed standard of 2 yg/m averaged over an 8-hour period. 3Vro of
the measurements at Tacoma, Washington, have average 24-hour concentrations that
exceed the proposed standard.
m the nonferrous-metals portion of this study it is shown that the
total volume of arsenic in the ore concentrates of zinc, lead, and copper are:
zinc 525 kkg/yr
lead 1,060 kkg/yr
copper 35,000 kkg/yr
thus, copper concentrate carries the largest portion of arsenic in the nonferrous
metals industry; also, based on the Bureau of Mines production statistics for
1973, the amount of arsenic per metric ton of copper produced is significantly
higher than for lead and zinc.
zinc (1.36 x 106 kkg in 1973) 0.385 kg As/kkg Zn
lead (1.36 x 106 kkg in 1973) 0.78 kg As/kkg Pb
copper (2.2 x 106 kkg in 1973) 15.9 kg As/kkg Cu
The amount of this arsenic that is released to the air as a result of
smelting operations is about 5230 kkg, which is only about 14 percent of the total
arsenic in the ore concentrates, but the largest portion is from the copjier indus-
try - 2.2 kg As per kkg of copper produced, versus 0.176 kg/kkg for lead and
0.140 kg/kkg for zinc. Of this total annual atmospheric arsenic emission, 92
percent is due to copper smelting.
Arsenic in Water
Arsenic enters natural water systems from these sources:
1. Natural sources:
(a) Natural erosion processes including microbially-
(5)
mediated erosion.
(b) Geothermal processes which may lead to very high
arsenic levels in locales where hot springs carry
f5 22)
arsenic to the surface. ' '
-158-
-------
2. Artificial sources:
(a) Pesticide runoff
(b) Smelter fallout
(c) Erosion processes stimulated by mining and
excavation operations
Cd) Runoff of agricultural fertilizers containing
arsenic as an impurity
(e) Deep-well drilling, especially geothernal areas
where nonferrous sulfides reside; the heated water
mobilizes arsenic to the surface.
(Bores drilled for geothermal power in New Zealand
delivered 190/000 kg of arsenic in the year 1964.
Mobile arsenic becomes either locked into highly insoluble soil (or
sediment) complexes where it is effectively removed as an environmental hazard,
or it moves from the air and from soil into the water resources which carry it
to the oceans. While in fresh natural waters, arsenic poses a potential health
hazard to those who drink the water and to those who eat food that has been grown
in or near such waters. Water-borne arsenic is probably the main source of in-
gested arsenic for the general population - as opposed to local populations ex-
posed to industrial operations.
Chronic arsenic poisoning has been reported associated with drinking
water containing concentrations of arsenic ranging from 0.21 to 10.0 mg/1. Con-
centrations of 0.05 to 0.25 mg/1 have also been reported as having no ill effect.
The current standard for drinking water in the United States (established by the
Public Health Service in 1962) is 0.01 mg/1, recommended maximum concentration,
and 0.05 mg/1, the maximum permissible concentration. A 1969 survey of 969
drinking water supplies in the U.S. found arsenic exceeding 0.01 mg/1 in 0.5
percent of the samples, and exceeding 0.05 mg/1 in 0.2 percent. Ferguson and
Gavis, in reporting these figures, feel that these concentrations indicate no
current threat to the public health.t5)
In two surveys of fresh surface waters in the U.S. in 1970 and 1971, the
arsenic concentration exceeded the 0.01 mg/1 level in about 7 percent of 1500
-159-
-------
samples from 150 rivers and in 21 percent of 727 samples from rivers and lakes.
The mean arsenic concentration of the samples exceeding the 0.01 mg/1 limit in
the first survey was 0.1 mg/1, which is ten times the U.S.P.H.S. recommended
maximum concentration and 2 times greater than the permissible limit. '
The hazard of arsenic in water is a function of the chemical state of
the arsenic, trivalent versus pentavalent. The reduced, or trivalent inorganic
arsenite form is the most hazardous, 10 to 60 times more toxic than arsenate.
Arsenic compounds in water, whether they are organic or inorganic, tend to oxi-
dize to inorganic arsenates; but the chemical equilibrium relationship between
+3 As and +5 As in natural waters has not been adequately determined. It is
generally agreed that arsenates exceed arsenites in oxidizing aquatic environ-
ments, but the oxidation to arsenate rarely proceeds to completion. In one study
of ocean waters, the ratio of +5 As to total As was close to 0.8. However, in
lakes and rivers where residence times are short, Ferguson and Gavis state that
unless the oxidation is catalyzed by microorganisms, oxidation "cannot advance
very far". C5)
The minimum concentration of arsenic in drinking water for which a toxic
effect (chronic) has been noted is 0.21 mg/1. This is only 4 times greater
!. ' i
than the maximum permissible concentration established by the PHS. If, however,
this level of 0.21 mg/1 was largely of the reduced arsenite form, then the factor
of safety between the maximum permissible concentration and the approximate thresh-
old level of intoxication due to arsenate would be between 40 and 240, certainly
a comfortable margin for any eventuality of chronic arsenic ingestion from air to
food. However, the ratio of arsenite to arsenate in drinking water and in fresh
surface waters has not been measured to any significant extent. Braman and
Poreback measured the amounts of arsenate and arsenite in fresh waters in Florida.
(28)
Hillsborough River
Withlacoochee River
Well water near Withlacoochee
River
+3 As
<0.02
<0.02
<0.02
(ppb)
+5 As
0.25
0.16
0.27
-160-
-------
+3 As +5 As
(ppb)
Remote ponc'i, Withlacoochee <0.02 0.32
Forest
Univ. Research Pond, Univ. of 0.79 0.96
S. Fla.
Lake Echols, Tampa 2.74 0.41
Lake Magdalene, Taitpa 0.89 0.49
For these few samples, the ratio of arsenite to arsenate is near to zero for the
first three samples, and it varies from close to 1 up to nearly 7 in the remainder.
Mare information of this type is needed, and its relationship to human activity
and to the environmental circumstances of the bodies of water.
With regard to food - from both animal and plant sources - grown in
senic contaminated waters, studies indicate that while bioaccumulation of arsenic
does take place to a very high degree (5'22/27/28/LL6) (measured as high as
71,000 times the ambient concentration for dried seaweed) , ' the arsenic that
accumulates in both plants and animals is of a form that presents virtually no
hazard upon ingest ion. ' ' ' Wbolson cites a study in which shrimp contain-
ing 128 ppm As were fed to rats at a dietary level of 13.3 ppm; they were also
fed As.CL at the Seine level. The rats' livers contained 20-fold less ar-
senic with the shrimp diet, and more than 98 percent of the arsenic fed in the
shrimp was excreted within 4 days. Humans fed shrimp excreted all the arsenic
within 4 days. (5'116)
The technology currently exists for the monitoring of arsenic (both
organic and inorganic and in its tri- and pentavalent states) in fresh and salt
waters and in all aquatic organisms; lower limits of detection of 1 ng are pos-
/c 28}
sible. ' Surprisingly little work has been done, however, in measuring the
relative portions of arsenic compounds and valence states in water and in aquatic
organisms. To the extent that research has been carried out, though, the main
hazard of arsenic in water is evidently not through the eating of organisms from
aquatic waters (except possibly in cases of extreme arsenic pollution) , but rather
derives from the drinking of water containing inorganic trivalent arsenic.
-161-
-------
Arsenic in Food
The tissues of plants and animals grown in arsenically-polluted surround-
ings accumulate arsenic. Table 14 lists the arsenic content of foods as mea-
sured by various researchers.
Seafoods contain the highest levels of arsenic found in connonly avail-
able foods; this is especially so of seafood harvested in coastal waters located
adjacent to outlets of arsenic-contaminated rivers. Many marine organisms, both
plant and animal, bioaccumulate arsenic in their bodies hundreds and even thou-
sands and tens of thousands of times above the ambient levels. But there is
evidence such arsenic is not retained very long after ingestion of such organisms,
and studies to prove biomagnification (increasing tissue levels at higher posi-
tions up the food chain) have shown that such does not apparently take
place.C5,22,27,116)
Numerous cases of arsenic poisoning due to food contamination have been
(27)
reported,v ' but all of the reported cases were the result of contamination that
took place as a direct result of pesticide residues (apples have been reported
to contain 1 to 2 mg lead arsenate) or because of arsenic contamination resulting
from food processing (e.g., the use of arsenically contaminated sulfuric acid to
modify sugar used in the production of beer resulted in 70 deaths and 6000 ill-
/27\
nesses in England in 1900}. ' No evidence was discovered in this study that
arsenic taken up by food organisms in natural biological processes has caused
a toxic effect, either chronic or acute, though it .Ls conceivable that under
extreme conditions marine organisms such as shrimp .uid certain edible marine
plants grown in highly contaminated waters might pose a chronic health threat.
Arsenical feed additives are used in the feed of swine and poultry to
increase growth rates and feed efficiency. Federal law requires a 5-day with-
drawal period from arsenical feed additives prior to sending the animals to mar-
ket. Tissue levels of arsenic in poultry and pork reaching the market are well
within the Federally established standards. ^ '
-162-
-------
The incidence of arsenic intoxication due to pesticide residues has
been decreasing over the past three decades in relation to the decreasing use of
arsenical pesticides and the enforcement of residue standards. The largest area
of potential hazard due to arsenic residue on foods in in foods grown near non-
ferrous smelting facilities; the State of Washington has warned Tacoma residents
who grow vegetables in the region of fallout from the smelter there that there is
a potential hazard due to both arsenic and cadmium; but people continue to raise
and eat vegetables, with no evidence of arsenical intoxication having yet been
reported. Studies of arsenic uptake in vegetable crops( ' have shown that even
at soil levels sufficient to cause a 50 percent reduction in plant growth, the
uptake of arsenic is not appreciable. (This is discussed below;) The arsenic
is such crops is probably of a form which is much less toxic than the trivalent
inorganic form. Thus, the chief hazard potential to persons who eat crops grown
in the vicinity of smelters is from arsenic which has been deposited on the
vegetable surface as a result of fallout.
Arsenic in Soil
As discussed in the section on soil chemistry, the fate of arsenic in
soil is a function of soil content of iron, aluminum, and calcium adsorption
sites and of soil pH, humus content, and available phosphorus (i.e., phosphorus
in solution), which competes with arsenic both for adsorption sites within the
soil and for plant uptake. Additionally, the amount of available arsenic varies
with tine, increasing or decreasing in complex relationship to the other soil
variables, but generally reaching chemical equilibrium (i.e., reaching a fairly
constant ratio of available-to-total arsenic) within several months after initial
application.
Agricultural soils typically contain several ppm of total arsenic. In
a study by Vfoolson, three soil types were used to study phytotoxicity and
plant uptake in various crops. The initial, total levels (ambient levels) for
the soils were:
Soil Total Arsenic (ppm)
Lakeland (L) 1.2
Hagerstown 4.5
Christiana (C) 3.5
-163-
-------
Assuming a "worst-case" condition (i.e., soil from which there is no
losses of arsenic via leaching, runoff, volatilization, or any other means) how
much arsenic would have to be aided, and over what period of time, to double
the ambient levels? The answer is based upon these considerations:
1. An acre of agricultural soil, as measured to a depth of
(38)
a "furrow slice", is taken as weighing 1 million pounds,
or about 450 kkg.
2. The six most common arsenical pesticides, their dosage
ranges and arsenic are:
Dosage range Effective
(kg/acre) % As As dosage
Lead Arsenate 1 - 2.5 22 .22 - .55
Calcium Arsenate 1-2.5 38 .38 - .95
Arsenic Acid 2 53 1.06
Cacodylic Acid 1 54 .54
DSm 2 41 .82
MSMA 1.5 46 .69
Thus, for the three soils listed, the number of dosages of the above
compounds sufficient to double the initial (ambient) total arsenic content are:
Ambient (ppm) 1.2 4.5 3.5
Lead Arsenate (doses) 1 - 2.5 3.5-9 3-7
Calcium Arsenate .5 - 1.5 2 - 5.3 1.5 - 4
Arsenic. Acid .5 2 1.5
Cacodylic Acid 143
DSMA. .75 2.5 2
MSMA .75 3 2.5
-164-
-------
Thus, assuming one dose per year for each arsenical, in seme cases
(e.g., arsenic acid in soil L) half a dose will double the total soil arsenic
level, while in others, such as lead arsenate on soil type H, 9 doses (or 9
years) are required.
In the real world, however, soil arsenic levels, both total and available,
decrease from initially high levels. And since it is the available (soluble)
arsenic which determines plant toxicity and uptake, available arsenic is tabulated
below for the three soils for various initial arsenate application amounts as a
function of time.(35J
Arsenate Available Arsenic After:
Applied
Soil (ppm As)
10
L 50
L 100
500
10
H 5°
H 100
500
10
C 5°
c 100
500
The significance of these numbers is apparent in terms of phytotoxicity
to the various crops studied by Wcolson. For example, the most "arsenic-
tolerant" soil is soil H - that is, the six crops tested by Wcolson, grew best
at any given applied arsenate level in soil H, which, as can be seen above, has
the lowest available arsenic levels. Thus, in soil H, at 100 ppm arsenic ini-
tially applied, all six crops tested (green beans, lima beans, spinach, cabbage,
tomatoes, and radishes) would be able to produce a crop at the 4-itonth and
9-month levels of available arsenic. (Considering arsenic acid, which has the
highest recommended arsenic dosage of the agricultural arsenicals considered
0 Months
1.4
20.0
48.3
384.0
1.0
6.0
18.3
276.0
2.6
18.3
52.3
429.0
4 Months
3.0
18.0
35.0
377.0
0.6
4.1
5.7
126.00
2.1
19.3
22.0
260.0
9 Months
3.7
20.7
55.0
288.0
<0.1
4.0
9.8
120.0
1.7
8.2
19.2
138.0
-165-
-------
here, the nunber of doses (years) necessary to achieve a total arsenic level of
100 ppm would be about 43 - and again, this assumes no losses of arsenic from
the soil by any means, whereas in reality much arsenic would be removed by the
natural methods listed in the soil chemistry section of this study; and even if
not removed from the soil, the large portion of the applied arsenic would not
be available for phytotoxicity or plant uptake in any way, as evidenced in the
above table.)
However, other soils, represented here by soil L, are less tolerant of
arsenic; phytotoxicity takes place at lower total applied levels. In another
study, by Wbolson, et al., soil types L and H were given initial doses of
100 ppm of sodium arsenate, and the available water-soluble arsenic was recorded
as a function of time. The available arsenic in soil H dropped to about 10 ppm
within 1 week while soil L required more than 24 vjeeks to decrease to 10 ppm
available arsenic.
The L soil, however, because of the longer period required for the avail-
able arsenic to become unavailable (i.e., fixed in the soil), has the advantage
of being able to more rapidly reduce its total arsenic content; the water-soluble
arsenicals are more prone to leaching deeper into the soil or of being carried
I- • < i
away with water runoff than are the fixed arsenicals in the H soil where fixation
takes place rapidly.
Soil H, because of its rapid soil fixation of arsenic, is more prone to
accumulation of arsenic if the annual doses are of a sufficiently high volume.
That is, the soil cannot continue to fix arsenic indefinitely - each year the
rate of fixation would tend to decrease if the amount of applied arsenic exceeds
the amount that can be annually removed by the various natural processes. Re-
search as to a threshold value fox such an annual dosage volume has not been un-
covered in this study, probably because it does not exist. However, since in
any real soil upon which arsenicals have been used the ratio of available-to-
total arsenic is always greater than zero (the actual chemical equilibrium values
of available-to-total soil arsenic are a function of soil type), total soil
arsenic will always tend to decrease toward the ambient level, which is the level
corresponding to the. natural movement of arsenic into and through the soil.
-166-
-------
In Wbolson's study of six crops, ' the amounts of available arsenic
necessary to cause a 50 percent growth reduction (GRcn) were:
(PPM)
Green Bean 6.2
Lima Bean 10.9
Spinach. 10.6
Cabbage 48.3
Tomato 25.4
Radish 19.0
The concentration of arsenic in the dried edible portions of these crops at the
GR~Q level of available soil arsenic were:
Dried edible
portion (ppm)
Green Bean 4.2
Lima Bean 1.0
Spinach 10.0
Cabbage 1.5
Tomato 0.7
Radish 76.0
The GRcQ level is effectively the economic limit at which a crop can be
grown; greater growth reduction will not result in a marketable crop.
Comparing the available soil arsenic to the plant arsenic (edible portion)
at GRcQ (above two tables) , it is evident that, except for the radish, bioaccumu-
lation does not take place as happens with freshwater and marine plants. Bio-
accumulation ratios (HR, ratio of water concentration to plant concentration of
a material) for fresh water plants range from 3 to 20,000 and for marine plants,
from 50 to 70,000. Of course, for plants grown in soil, the concentration of
available arsenic is based upon the soil itself rather than upon the water in
the soil, for which the bioaccumulation ratio would undoubtedly be higher, but
not as high as for aquatic plants.
-167-
-------
The identity of the arsenic compound or compounds in land-grown crops
is not known. However, studies of aquatic plants have shown it to be present
as both water- and lipid-soluble arseno-organic ccnpounds. Wbolson, referring
to Lancaster, et al., points out that lake weeds containing 288 ppm arsenic were
fed to sheep as 20 percent of the total diet for 3 weeks with no ill effects on
the animals' health. Tissue residues of arsenic did increase during the period,
but decreased when the weed was removed from the diet.
In the case of the 76-ppm radishes, assuming that radishes normally con-
(23)
tain 93 percent water, the arsenic concentration would be on the order of
5.3 ppm in a corresponding fresh, undried radish. Assuming further (worst case)
that the arsenic in such radishes is of a form having a toxicity equivalent to
trivalent inorganic arsenic and that the arsenic would be completely absorbed
from the alimentary canal upon ingestion, then the .imount of radishes required
to produce a minimal toxic effect - on the order of 10 mg arsenic - would be
about 150 3-cnwiiameter radishes. It is likely the symptoms of arsenic toxicity,
even in this worst-case situation, would be masked by those of ordinary overin-
dulgence.
1 With regard to plant uptake of soil arsenic, Vallee, et al. observe that
"soil concentrations of arsenic may rise to many hundred parts per million after
years of spraying with lead arsenate and other pesticides", and "experimental
attempts to sterilize fresh soil have sometimes required huge amounts of arsenical
compounds". Nevertheless, small amounts of arsenic may be taken up by plants
grown in heavily contaminated soil, "but rarely in quantities sufficient to con-
(27)
stitute a human risk".
-168-
-------
SECTION VIII
THE MARKET FOR ARSENIC
Domestic Arsenic Supply
All of the domestic arsenic production is in the form of white arsenic,
AsJD , and is derived from collected flue dusts from copper smelting. Moreover,
the entire U.S. production since 1959 has been at one location - the Tacoma,
Washington plant of ASAHCO. Up until 1959, white arsenic was produced at two
other plants, the Anaconda Co. plant at Anaconda, Montana, and the U.S. Smelting,
Refining, and Mining Co. plant at Midvale, Utah.
Because there is a single producer of white arsenic in the U.S., recent
production data has not been released by the Bureau of Mines. However, Table 18
lists these data up until 1968 as reported by the Bureau of Mines. In addition,
Table 18 also lists the price history for white arsenic, and the production of
primary refined copper from domestic ores. The price for white arsenic is listed
in terms of constant 1974 dollars; up until 1968, the correction for inflation
(17)
was based upon Bureau of Mines data, and later corrections were based upon the
Bureau of Tabor Statistics wholesale price index for intermediate industrial
materials.^142*
The U.S. production of white arsenic in any year is dependent upon the
quantity of copper ore smelted, and upon the world price for white arsenic. The
data of Table 18 from 1949 through 1968 (20 years) were analyzed to quantify this
dependence. The results were a regression equation:
/Xn - 227 \ /X - 947,000
Yc = 5,160 + 3,300 ^-AsO—y/+ 2'31° ( 500,006
where Y = Calculated U.S. Production of White Arsenic, Metric Tons Per Year
c
X, = White Arsenic Price, Constant 1974 Dollars Per Metric Ton
Xj = U.S. Production of Refined Primary Copper from Domestic Ore,
Metric Tons Per Year
-169-
-------
TMOR 18
SUPPLY STATISTICS FOR ARSENIC TKKKEDE
SOURCES: BUREAU OF MINES^17'18'70'141',
Year
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
U.S. As 03
Production,
kkg/year
5,580
5,790
7,400
6,820
4,830
5,760
4,740
5,480
4,570
5,060
2,340
6,000
5,100
5,400
4,900
4,500-i
7,000 !
5,500
2,900
3,500
6,700*
6,700a
6,100a
7,000a
6,900a
8,700a
Price, Actual
Dollars per
WogAs,©,
117
132
139
132
123
123
123
123
123
123
101
' 104
93
93
116
118 '
126
118
126
133
143
143
143
143
143
286
Price, Constant
1974 Dollars
per Jdog As2O,
271
310
314
279
258
253
250
243
235
232
183
183
157
157
196
199
212
199
204
206
216
208
200
190
178
286 :
U.S. Cu Production
from Domestic Ores,
Hog/year
630,000
835,000
864,000
837,000
845,000
764,000
904,000
979,000
952,000
910,000
723,000
1,018,000
1,071,000
1,102,000
1,104,000
1,142,000
1,211,000
1,227,000
769,000
1,053,000
1,331,000
1,380,000
1,280,000
1,523,000
1,557,000
37,440,000
U.S. Imparts
of ASjO,,
kkg/year
8,300
24,100b
19,400**
6,000**
9,800*
9,400**
12,900b
17,800**
16,600?*
13,700**
22,100**
16,100**
17,600
14,300
13,200
16,500
14,100
16,900
24,500
22,800
16,500
17,000
15,700-
12,300
10,400
13,600
TI.S rmampt-inn
of As,0,,
kkg/year
13,900
29,900
26,800
12,800
14,600
15,200
17,600
23,300
21,200
18,800
24,400
22,100
24,600
24,400
25,300
29,400
28,900
28,600
31,300
28,800
23,200°
23,700°
21,800°
19,300°
17,300°
22,300°
World Production
of As20,,
Hog/year
53,500
45,000
48,400
52,700
51,000
52,000
58,800
61,200
49,800
49,500
50,300
45,400
47,400
47,600 . .
aEstimated ty Regression Analysis of Prior Years
estimated from Oonsuqption and Production Data
rrnm Production aai liqaoirts
-------
Figure 1 compares Y / the calculated domestic As00 production level/ with the
C- 4U J
reported production data of Table 15. The standard error in Y is 960 metric
C
tons per year.
The influence coefficients in the regression equation indicate ( as they
should) that the U.S. production of white arsenic increases as both the price
and the copper production increase. The two coefficients (of the two normalized
variables) are of comparable magnitude, implying that both variables are of
similar importance.
The regression equation was then used to calculate domestic As?0
duction data for the years 1969 through 1974. The calculated 1974 level, 8,700
kkg of As_0_ (equivalent to 6,550 kkg of arsenic) is lower than other values in-
dependently estimated elsewhere in this report (8,180; 8,300; and 12,250 kkg of
arsenic) . As an upper limit, the plant capacity of ASARDO/Tacoma was reported
as 33 metric tons of As00_ per day, or 12,000 metric tons of As00, per year on
f£Q-\ * J * J
a 365-day basis. (*yc)
Figure 2 shows supply curves for domestic white arsenic, at several levels
of domestic copper production, derived from the regression equation for AsJD-
production. If the market price is less than $50 per metric ton, it is likely
that domestic production will cease; e.g., it will no longer pay to refine and
sell the white arsenic. Although the 1974 level of primary copper production
from domestic ores was about 1.5 mill ion metric tons per year (at a price of
$l»70/kg) , the Bureau of Mines projected that the copper production would reach
2.275 million kkg at $2.57 per kg (in constant 1974 dollars) . * '
There are many important and new factors affecting domestic white ar-
senic production which the historical data (and so the regression equation and
Figure 2) do not take into account:
1. The past few years have witnessed a rapid increase in
the quantities of copper ore which are leached. In 1968,
12 percent of the total domestic mine production was
cement copper; this had grown to 15 percent by 1972.
Since much less arsenic, accompanies cement copper from
the mine than copper ore concentrates from the mine, the
-171-
-------
FIGURE: i
CALCDIATED VS. REPORTED
U.S. FPD0UCTIGN Cfl? Mffiffi
1949-1968
5
1
I
i
8000
7000
€000
6000
4000
3000
8000
1000
0 1000 2000 3000 4000 9000 6000 7000 8000
REPORTED AleO, PRODUCTION, KKG/YEAR
-172-
-------
u>
400
390
300 -
. 150
o
*
100
0 2000 4000 6000 8€00 tOflOO KflOO 14,000 KyQQO
DOMESTIC A%03 PW3DUCTION, KK6/YEAR
o
a
O
-------
ratio of total available arsenic to copper product
should be decreasing. On the other hand, since the
recovered As 0 is such a small fraction of the total
available arsenic, the impact of leaching nay not be
significant with respect to the domestic supply of
2. The past few years have also witnessed a marked improve-
ment in controlling sulfur dioxide emissions from copper
smelters. Previously, relatively few smelters had acid
plants (for reasons stated earlier in this report) . The
impact of environmental forces upon the industry has been
to extend SCL control (and subsequent acid production) to
a target of 90 percent capture. With greatly increased
emission control, the particulates (which include As2O,)
formerly released to the atmosphere are now being cap-
tured to a greater degree. Hence, gas cleaning processes
upstream of acid plants should be recovering much greater
quantities of arsenic. The portion recovered via dry dust
collection (electrostatic precipitators and baghouses)
should increase the commercial supply of arsenic, while the
portion removed via wet scrubbing should not. It also
appears that the enforced recovery of more SCL has played
a part in the increase in copper ore leaching, by pro-
viding a source of sulfuric acid.
3. The recently-proposed arsenic standards by the Occupa-
tional Safety and Health Administration, if promulgated,
could drastically affect the commercial supply of arsenic
in either direction. Enforced recovery of arsenic to meet
tighter ambient standards could increase the supply. How-
ever, the tighter standards may conceivably shut down the
sole producer of white arsenic, if ambient standards in
the As2°3 Plant are to° expensive to meet.
-174-
-------
4. The discussion of the cnrunerclal occurrence of arsenic
(elsewhere in this report) revealed that potential sources
for arsenic other than copper ores amount to tad.ce the
copper-related resource. Much of these potential (but
1
unexploited) sources are not technically or economically
practical under any reasonable set of market circum-
stances, either because the arsenic is at extremely small
concentrations or because the arsenic is tightly bound.
However, there remain relatively large and feasible un-
tapped resources for arsenic (should there ever be suffi-
cient demand for arsenic). Among these feasible resources
are phosphoric acid (555 kkg/year) and Searles Lake brines
(2,160 kkg/year). Feasible arsenic resources which should
become very sizable in the next few decades are associated
with coal gasification and with geothermal energy.
For the above reasons, the domestic supply of arsenic may conceivably
increase by dramatic proportions in the next few years, or may conceivably be
reduced to zero. The conclusion is therefore reached that no long-term (i.e.,
ten years) projection may be made with any meaningful certainty.
World Arsenic Supply and Total U.S. Supply
Table 18 also lists the world production of white arsenic from 1961
through 1974. For those years, the U.S. production amounted to about 10 percent
of the world production. The world production data is shown in Figure 3. It is
apparent that the quantity of white arsenic produced is not strongly correlated
to price. Hence, the world supply curve of Figure 3 was drawn vertically (e.g.,
completely inelastic supply) in the range of the historical data ($150 to $300
per metric ton).
At the lower range of Figure 3, the world supply curve was drawn with in-
creasing supply elasticity; marginal refiners will drop out of the marketplace as
the price approaches $50 per metric ton. Above $300 per metric ton, those with
crude arsenic resources (either as a waste material or as a byproduct) will be
-175-
-------
T
400 -
390
300 -
fe 250
200
o
iso H
s
0- 100
I
50
'TOTAL
U& SUPPLY
(IMCUOES
IMPOnT9
WORLD
PRODUCTION
o o
iO 20 30 4O 90 60 7O 60
SUPPLY THOUSAND KKG/YEAR
u>
-------
induced to refine and market these resources. It has been, shown that non-copper
resources of arsenic are potentially large, and that unrecovered copper resources
of arsenic are also large. Hence, it is implied by the arrow at the upper end of
Figure 3 that an. entirely different supply regime for arsenic exists but has yet
\
to be quantified.
'Ihe present world supply is not totally related to copper ore, as it is
in the United States. As the data in Table 19 indicate, several countries im«-
portant in arsenic production are not important in copper production, and visa
versa.
Also shown in Figure 3 is a curve for the total 1974 U.S. supply (domestic
production plus imports) of white arsenic. This latter curve was constructed as
parallel to the domestic production curve of Figure 2 (at a copper production rate
of 1.5 million metric tons per year).
Demand for Arsenical Insecticides
Prior to 1952, very large quantities of arsenical insecticides ware
consumed:
Insecticide
Paris Green
Lead Arsenate
Calcium Arsenate
Composition
Cu(COCCH3)2-3 Cu(As02)2
PbHAsO
4
Ca3(As04)2 + CaO
As20_ Content
55%
32%
50%
Applications
Potatoes, Mosquitos
Potatoes, Apples
Cotton, Apples
In 1940, the consumption of lead arsenate and calcium arsenate were
(respectively) 34,100 and 22,700 metric tons, ' equivalent to a total of 22,300
metric tons of white arsenic. The demand dropped drastically from 1940 to 1949;
in the latter year, the U.S. demand for all uses of As2CL was only 13,900 metric
tons. Higher cotton prices and increased cotton planting in 1950-1951 temporarily
increased the demand for calcium arsenate for boll weevil control; the total U.S.
demand for As2°3 in the 1950's was dependent from year-to-year upon the degree of
boll weevil infestation.
-177-
-------
TABLE 19
PRODUCTION OF ARSENIC AND COPPER BY COUNTRY IN 1972
^143)
SOURCE: BUREAU OF MINES
JILER
Country
United States
France
U.S.S.R.
Mexico
Svreden
Peru
W. Germany
Japan
S.W. Africa
Brazil
Portugal
Canada
Spain
Chile
South Africa
Zaire
Zambia
Phillipines
Australia
World
As.OjProduction,
Metric Tons/Year
7,000
10,000
7,200
591
16,000
1,020
500
427
4,000
164
190
27
-
-
-
-
-
. -
-
45,400
Copper Production
Metric Tons/year
1,510,000
500
664,000
78,600
24,700
225,000
1,320
113,500
32,400
4,300
4,800
725,000 ,
32,100
724,000
162,000
429,000
716,000
205,000
185,000
6,630,000
-178-
-------
In the 1960's, the arsenical insecticides received strong competition
from organic insecticides, so that by 1971 only 1.4 percent of all insecticides
used on crops were arsenicals; and the quantity of arsenical insecticides used on
crops amounted to only 1,260 metric tons (90 percent of which was used on apples and
other fruits - virtually none was used on cotton or potatoes).™ ' However, non-
crop consumption of arsenical insecticides amounted to more than an equal quantity,
and were typically used by homeowners. From 1965 through 1972, the average con-
sumption both for crops and for other uses was:
Lead Arsenate
Calcium Arsenate
Paris Green
Insecticide Consumption,
Metric Tons/Year
3,500
900
7,300
Equivalent AsJD_,
Metric TonsASar
1,100
450
4,000
Hence, the. equivalent AsJD_ demand during this period, for insecticides,
amount to 5,550 metric tons per year. The wholesale prices during the 1970 to
1972 period were relatively stable at about $800 per metric ton for lead arsenate
and about $410 per metric ton for calcium arsenate (about $1,100 and $550,
respectively, in constant 1974 dollars).
(144)
The higher price of the lead
arsenate reflects, of course, the high cost of lead oxide or lead nitrate.
Both lead and calcium arsenates are manufactured from arsenic acid, which
in turn is made from white arsenic. In 1970-1972, the prices for arsenic acid .(100
percent basis) and for white arsenic were respectively (in constant 1974 dollars)
about $750 and $200 per metric ton. An analysis of ingredient costs is:
Product
Arsenic Acid
Lead Arsenate
Calcium Arsenate
Paris Green
Ingredient
White Arsenic
Arsenic Acid
Arsenic Acid
White Arsenic
kkg Ingredient
kkg Product
0.70
0.46
0.50
0.55
Cost of Ingredient
Per kkg Product
$140
$345
$375
$110 ,
Product Price,
$/kkg
$ 750
$1,100
$ 550
$ 400
-179-
-------
It is apparent that the arsenic-bearing starting material IB an impor-
tant factor in determining the price of lead arsenate and paris green, and is
the critical factor in determining the price of calcium arsenate. The demand
for arsenical insecticides has been decreasing recently because of competition
from organic insecticides, because of the cancellation of crop registrations,
and because of tighter OSHA constraints upon the manufacture of arsenicals.
The 1975 price for lead arsenate is about $600 per metric ton, much less than
in the 1970-1972 period, despite the increased cost of the arsenic-bearing
ingredients, and despite the increased price of the organic insecticides which
compete for its use.
Based upon the above discussion, the As2°3 Demand curve for insecticides
was constructed as shown in Figure 4. This curve passes through the 1970-72
point of 5,550 kkg/year at a price of $200 per metric ton of As20_. The curve
was drawn to be relatively inelastic from $100 to $400 per metric ton on the
rationale that the use of arsenical insecticides is already a small fraction of
the total insecticide demand; so that its use is highly selective and not likely
to change drastically with, price. At higher As2°3 P1^068/ the curve reflects
greater demand elasticity, as alternate (i.e., organic) insecticides may be sub-
stituted for arsenicals. At lower As2p3 prices, sane increase in demand is shown,
but it is not anticipated that arsenicals would make major inroads on the total
insecticide market.
Demand for Arsenical Desiccants and Defoliants
Arsenic acid use for the desiccation of cotton has rapidly increased since
thenid-WeO's:'49'53'146'147'
Year
1968
1969
1970
1971
1972
1973
1974
1975
Arsenic Acid Used,
kkg/year (100% Basis)
965
1,290
1,620
1,950
2,760
3,600
4,400
(5,600)
Equivalent As90,,
kkg/year ^ J
724
969
1,212
1,460
2,070
2,700
3,300
(3,920)
Arsenic Acid Price,
1974 $/kkg
$ 620
650
690
740
760
760
780
(880)
-180-
-------
o
(000 2000 3600 4000
WHITE
9000 SOOO TOD WOO 9090 10000 \\ftt) BjXX>
QUANfTITY, KK6/YEAR
-------
The use of arsenic acid to terminate the cotton plant is considered
mandatory in the Blackland area of Texas. ('1/l8'149) The convincing arguments
include:
1. Early maturity and harvest followed by early stalk
destruction is an integral part of the pest management
program. Desiccation removes the available food supply
for boll weevils. Arsenic acid is the only desiccant
that constantly terminates the cotton plant, especially
in higher rainfall areas, and arsenic acid is the only
effective compound for controlling re-growth.
2. Arsenic acid renders plant parts brittle for mechanical
stripping of the bolls. Mechanical spindle picking is too
expensive, and waiting for frost would incur heavy insect
losses. A self-propelled stripper costs approximately
$20,000; a spindle picker costs twice as much. A stripper
can harvest as much as 70 bales per day, while a mechanical
picker can only harvest 20 bales per day in good^yielding
cotton. Pickers are more useful in high yield areas (2
bales per acre or more). ,
3. Without effective desiccation, stripping results in large
quantities of "green trash" which, cause major problems.
at the cotton gin. In addition to stoppages ("choke-ups11),
heating would decrease lint and seed quality.
Cacodylic acid is also used as a cotton harvest aid, in the cotton states
other than Texas and Oklahoma. Its use as a defoliant is new, dating back only
(149)
to 1972. In 1974, about 230 metric tons were used, ' equivalent to 125 metric
tons of white arsenic.
Based upon the above discussion, the desiccant and defoliant demand curve
(for 1974) was constructed, as shown in Figure 4, as inelastic from $150 to $400
per metric ton of As-O . Above $400, some elasticity is indicated, but it is
judged that arsenic acid would still be used to some extent to $700/kkg of As2^V
Below $150, the use of arsenic acid should increase to include other cotton-
growing areas and to more completely saturate the Texas-Oklahoma region.
-182-
-------
In the 1970-1973 period, arsenic acid waa priced at about $730 per metric
ton (100 percent basis, 1974 dollars). During this period, As-O., was priced at
about $200 per metric ton; the AsJD- ingredient cost was about $150 per metric
ton of arsenic acid.
Demand for Arsenical Herbicides for Weed Control'
The nonsodium and discdium salts of methanearsonic acid, MSMA and DSMA,
are widely used on cotton for weed control. In 1974, about 3.0 million hectares
of cotton land were treated with methanearsonates as directed postemergence
sprays, in one or two applications, at a level of about 3 kg per hectare per
application. An addition! 1.3 million hectares were treated with topical poste-
mergence applications, at about 1 kg/hectare; and 0.2 million hectares were
treated prior to crop planting at about 3 kg/hectare.
(149)
This usage data im-
plies a consumption of close to 10,000 kkg/year. Additional quantities of MSMA
and DSMA were used for weed control on turf, on lawns, and on ornamental shrubbery.
Independently, the consumption of methanearsonic acid salts was estimated
as:
.(43,49,150)
Year
i
1970
1971
1972
1973
1974
MSMA and DSMA,
kkg/year
11,000
9,000
11,200
9,100
9,800
Equivalent As90_,
kkg/year ^ J
6,400
5,200
6,500
5,300
5,700
However, the use of organic arsenicals amounted to only 6.7 percent of all
the herbicides used on cotton in 1971, and to only 2.1 percent of all the herbi-
(49)
aides used on all crops. ' One alternative herbicide for cotton, both for pre-
plant applications and as a post-emergent directed spray, is paraquat, which is
also used as a harvest aid on cotton.
The prices for MSMA and DSMA have generally been as follows, on a 100
percent basis:^ *
-183-
-------
DSMA Price, 1974 $/kkg
kkg AsJDVkkg DSMA
As203 cost/kkg DSMA, 1974 $
MSMA Price, 1974 $/kkg
kkg As oykkg MSMA
As203 cost/kkg MSMA, 1974 $
1968-J.969
$1,350
0.54
$ 115
$1,000
0.62
$ 130
1975
$2,200
0.54
$ 155
$2,100
0.62
$ 180
Figure 4 shows the demand curve for arsenical herbicides. It is con-
structed to pass through the point of 5,800 metric tons per year of As203 at
a price of $200. The demand is shown to be moderately elastic: at higher prices,
the price differential between arsenicals and organics will be smaller; and at
lower prices, the arsenicals should command a greater portion of the total herbi-
cide market.
Demand for Arsenical Soil Sterilizers
i In 1972, the quantity of sodium arsenite shipped was 4,200 metric ipns;
compared to 5,300 metric tons in 1967. The prices (in 1974 dollars per kilogram)
were $415 in 1972 and $210 in 1967.^' Since 0.76 kkg of AsJX are equivalent
to one kkg of NaAsCL, the ingredient cost per metric ton of sodium arsenite (in
1974 dollars) was $145 in 1972 and $155 in 1967.
The herbicidal uses of sodium arsenite include soil sterilization such as
for railroad rights-of-way, for tank farms, for parking lots, for electrical sub-
stations, and for ornamental uses under trees and shrubs. The total demand for
sodium arsenite, however, includes some insecticide uses such as animal dips and
termite control.
The demand curve of Figure 4 indicates that the use of arsenical soil
sterilizers is already highly selective, and therefore inelastic.
-184-
-------
Demand for Arsenical Wood Preservatives
The demand for chromated copper arsenate (CCA) and for fluor chrome
arsenate phenol (PCAP) has been as follows:
Year
1967
1968
1969
1970
1971
1972
1973
OCA,
kkg/year
1,060
1,460
2,120
2,740
3,960
4,430
5,320
PCAP,
kkg/year
2,430
1,800
2,060
1,220
987
870
767
As203 Equiv. , kkg/yr
OCA
330
460
670
860
1,250
1,330
1,640
PCAP
520
390
440
260
210
190
160
Total
850
850
1,110
1,120
1,460
1,520
1,800
The changeover from PCAP to CCA has also been accompanied by an increase
in the total consumption of wood preservatives, in the total consurptian of ar-
senical wood preservatives, and in the consumption of white arsenic. The CCA
price, however, has remained fairly stable at about $2,100 per metric ton of
CCA "oxide" (in constant 1974 dollars).(57'152)
For each metric ton of CCA, about 0.45 metric tons of arsenic acid is
consumed in its manufacture. At a price of $750/kkg for arsenic acid (constant
1974 dollars), the arsenical ingredient cost is about $340. Similarly, the ar-
senic acid for PCAP Is about 0.31 kkg/kkg PCAP, at a cost of about $230/kkg PCAP.
The ingredient cost for copper and for chromium (CCA-B is 20 percent CuD and 35
percent CrCO and is more important than the arsenic acid cost.
Since no real substitutes can be found for arsenical wood preservatives,
and since the cost of arsenical ingredients is a relatively small fraction of the
preservative price, the demand curve of Figure 4 has been drawn to be relatively
inelastic.
-185-
-------
Demand for Arsenical Feed Additives
The two important arsenical feed additives are Poxarsone (4-hydroxy-
3-nitrobenzenearsonic acid, or "3-nitro") and arsanilic acid (p-anunobenzene-
arsonic acid) :
AsO(OH)
AsO(OH),
Roxarsone Arsanilic Acid
One metric ton of Roxarsone is equivalent to 0.378 metric tons of
and
one metric ton of arsanilic acid is equivalent to 0.459 metric tons of As.0_.
The consumption and price data are as follows: ^153' 154/155)
Year
1968
1969
1970
1971
1972
1973
1974
1975
Arsenical Feed
Additives,
kkg/yr
1,160
1,230
1,320
1,300
1,360
1,330
1,360
Equivalent
As203,
kkg/yr
464
491
528
521
544
531
544
Price, 1974 $/kkg
Roxarsone
$6,500
6,600
6,600
6,800
6,500
6,000
5,500
4,900
Arsanilic Acid
$4,100
4,000
3,800
3,300
The cost per metric ton of arsenical feed additive for the arsenic acid
ingredient is approximately $425, a small fraction of the product price. The
above data indicate that the price of Roxarsone (in constant 1974 dollars) was
relatively stable through 1972, but has been decreasing since; the price of ar-
sanilic acid has apparently been dropping steadily.
-186-
-------
•Hie substitutes for arsenical feed additives are antibiotics, which range
in price from $20,000 to $60,000 per metric ton., *• ' an order of magnitude
greater than the arsenicals. The demand curve of Figure 4 is constructed to be
inelastic, reflecting this price differential and also reflecting the relatively
small iirpact of As203 cost upon feed additive price.
Demand for As203 in Glass Manufacture
Data on the consumption of As203 in the glass industry indicate a major
reduction from 1968 (5,100 kkg) to 1974 (2,400 kkg). White arsenic was used in
the past at a level of about 0.5 percent in decorative glass such as crystal
tableware. Substitutes for oxidizing and firing are generally available, so that
the demand curve of Figure 4 was constructed to be moderately elastic.
Demand for As203 in Miscellaneous Uses
This category of use includes specialty items such as As2S3 for special
pigments, gallium arsenide semiconductors, and arsenide for light-emitting
diodes. The special nature of these uses (which now amount to about 500 metric
tons per year of As203) indicates that the demand is inelastic, as shown in
Figure 4.
Sunroary of Demand for As203
The estimated demand for As203, as taken from the demand curves of
Figure 4, is listed in Table 20. The total estimated domestic demand for
As203 is relatively inelastic from $100 to $400 per metric ton.. It should
be emphasized that the curves of Figure 4 and the data of Table 20 .are based
upon historical data only in the neighborhood of $200 per metric ton; ex-
trapolations from this level are based largely upon qualitative information.
Earlier in this section, the conclusion was reached that no long-term
(i.e.. ten years) projection of U.S. arsenic supply may be made with any meaning-
ful certainty. The prospect for making meaningful U.S. demand projections is
equally bleak for two critical reasons:
-187-
-------
M
00
TABLE 20
ESTIMATED U.S. DEMAND FOR WHITE ARSENIC (1974 BASIS)
Use
Insecticides
Dessicants and Defoliants
Herbicides for Weed Control
Soil Sterilizers
Wood Preservatives
Feed Additives
Glass Additives
Miscellaneous Uses
Total
As2O3 Demand, kkg/yr, at As2O3 Prices (1974 Constant Dollars)
$50
6,250
4,250
7,150
4,450
1,750
1,150
3,100
700
28,800
$100
5,830
3,720
6,500
4,250
1,560
620
2,780
500
25,770
$200
5,500
3,500
5,800
4,200
1,550
550
2,400
480
23,980
$300
5,200
3,500
5,150
4,200
1,550
550
2,000
480
22,630
$400
4,800
3,500
4,550
4,180
1,550
550
1,610
480
21,220
$500
4,100
3,130
3,800
3,750
1,550
550
1,200
480
18,560
$600
2,700
2,250
2,400
2,800
1,400
550
630
480
13,210
$700
0
0
0
0
980
480
0
380
1,840
$800
0
0
0
0
230
140
0
100
470
-------
1. Arsenical products compete directly with petrochemicals in
virtually every use category except for glass additives and
some miscellaneous uses. Petrochemical products may be said
to dominate the arsenical markets for insecticides, for des-
sicants and defoliants, for herbicides, for soil sterilizers,
for wood preservatives, and for feed additives. The large
price increases in 1974 and 1975 for arsenicals were likely
the result of large price increases for petrochemicals in
these markets. With additional time, it is possible that
the production capacity for arsenicals will be increased
so that arsenicals can comnand greater shares of the markets
(at lower prices). The volatility of petrochemical prices
and supplies, however, would make such projections extremely
imprecise.
2. The future for arsenical products lies to a great measure
upon actions to be taken by a number of government agencies.
The Occupational Safety and Health Administration, the U.S.
Environmental Protection Agency's Office of Pesticide Pro-
grams, and State agencies have the mechanisms for banning,
severely restricting, or otherwise drastically influencing
the demand for arsenicals or for their market competitors.
For the above two reasons, historical market data (largely the result of
uncontrolled commerce) provide little basis for projecting the future arsenical
market, which promises to be a controlled market.
-189-
-------
SECTION IX
nDEOTIFICATIQN AND SCREENING OF
CANDIDATE CONTROL ALTERNATIVES
In previous sections, the role of arsenic in the U.S. economy has been
discussed in detail. The release of arsenic and its compounds to the environ-
ment has been addressed, with, emphasis upon identifying the specific sources of
such releases and upon quantifying these releases. An assessment of the health
hazards resulting from such releases has teen made.
In this section, various control alternatives for reducing these health
hazards are presented, and evaluated from the standpoints of feasibility, neces-
sity, and effectiveness. Those alternatives passing this screening process will
be evaluated from a cost standpoint in the next section.
Existing Control Programs
Many suitable control alternatives are already in effect for reducing
the dangers from arsenic. These include:
1. The dangers to workers from arsenic exposure are being
suitably addressed by the Occupational Health and Safety
Administration of the Department of Labor.
2. The potential dangers from arsenic in water supplies are
being suitably addressed by the standards for drinking
water and by monitoring water supplies, by the U.S.
Environmental Protection Agency and by State and local
governments.
3. The potential dangers from arsenic in food supplies are
being suitably addressed by the standards and monitoring
activities of the Pood and Drug Administration.
4. The U.S. Environmental Protection Agency and State agencies
are active in limiting arsenic discharges via wastewater
effluents from point sources. As the results of the
study show, arsenic in wastewaters are the least of all
emissions and dissipations.
-190-
-------
5. The Office of Pesticide Programs of the U.S. Environmental
Protectfon Agency, and appropriate State, agencies, are
active in pesticide registration programs. These programs
limit the use of a given pesticide Cinsecticide, herbicide,
defoliant, dessicant, soil sterilizer, fungicide, etc.) to
a specific and finite combination for crop Cor application)
and pest. These programs require positive Government
actions for the use of a pesticide in new applications;
conversely, the cancellation of a specific pesticide regis-
tration is equivalent to a selective use ban. For registered
uses, the pesticides must be appropriately labelled, and
with information made available as to proper handling,
proper use, use precautions, chemical, physical, and bio-
chemical behavior, behavior in or on soils, and toxico-
logical properties.
6. The U.S. Environmental Protection Agency and State and
local agencies are active in reducing the dangers from
arsenic air pollution. The actions limiting the sulfur
oxide emissions from primary copper smelters have been
effective in reducing arsenic emissions from these sources.
The actions limiting particulate emissions f ran power
generation stations and other stationary sources have
reduced arsenic emissions as well. Since the arsenic con-
tent of coal is keyed to the sulfur content, actions re-
sulting in the use of low-sulfur coal have also resulted
in reduced arsenic emissions.
Oontrol Alternatives for Specific Emissions or Dissipations
Several control alternatives have been formulated to reduce specific
emissions or dissipations or arsenic:
1. Requiring all phosphoric acid manufactured in the United
States to be processed for arsenic removal prior to its
-191-
-------
use in manufacturing fertilizers or any other phosphate
products. The technology used for making food-grade
phosphoric acid could be adapted.
This candidate control was rejected from further con-
sideration for three reasons.
First, the arsenic in fertilizers, dissipated on land,
has been shown to present no imminent hazard. Second,
the arsenic in animal feed phosphatea has not resulted
in dangerous additions to the human food supply. Third,
the arsenic in phosphate detergents, while constituting
the largest source of arsenic water pollution from point
sources, has not led to dangerous levels of arsenic in
fresh waters nor to dangerous levels in public water
supplies.
2. Banning the intentional use of arsenic as an alloying ele-
ment in non-ferrous metals.
This alternative was rejected for two reasons. First, the
only hazards appear to be the emissions to the air and the
wastes destined for land disposal in the secondary metals
processes. These losses are more directly and appropriately
controlled with specific air and land regulations than with
a blanket ban. No health hazard is apparent from the use
of products (batteries, cables, radiators) containing ar-
senic. Second, the arsenic alloys serve useful commercial
purposes, and substitutes for arsenic alloys are not ap-
parently available.
3. Banning the intentional use of arsenicals in consumer
products (other than non-ferrous alloys). These would
include arsenical wood preservatives, arsenical fungi-
cides for vinyl plastics, and arsenical herbicides in
home-lawn-care products.
-192-
-------
This alternative was rejected because there is no apparent
health hazard to the consumer via vaporization, leaching,
or othar mechanism.
4. Stringent emission standards for the release of arsenic
•t-T-inyffto to the atmosphere frdrti' high^teropprgiti'tFg indus-
trial processes. It is now generally accepted that air-
borne arsenic trioxide is a carcinogen. The primary copper
industry, other-primary and secondary non-ferrous metals
industries, cotton trash incinerators, and some glass manu-
facturing plants, all emit arsenic trioxide from high-
temperature processes. It was concluded from this study
that even the best of the dry dust collection techniques
fall far short of effective ASo^3 capture.
This control alternative is deemed to be needed, feasible,
and effective in reducing health hazards so that it will be
considered further. Based upon the technology of As2°3
removal from flue gases (as demonstrated in the gas-cleaning
parts of byproduct sulfuric acid plants at copper smelters
and other non-ferrous metal smelters), high pressure-drop
venturi scrubbers can achieve 99+ percent removal of As2°3*
The control measure, therefore, would be an air quality stan-
dard based upon such high removal efficiency.
5. Stringent emission standards for the release of arsenic tri-
oxide to the atmosphere from the combustion of fossil fuels.
The rationale for this alternative is the same as for the
previous alternative, in terms of reducing health hazards.
The feasibility of an emission standard for stationary
sources, based upon the very high As2°3 ranoval capability
of high pressure drop wet scrubbing systems, is similar to
that for industrial sources of ASoQ, air pollution. This
alternative will therefore be further considered.
-193-
-------
However, it does not appear that a stringent As2°3
emission standard for mobile sources is feasible. More-
over, the quantity of emitted arsenic estimated from
petroleum combustion, 108 metric tons per year, while
significant, is not an extremely large fraction of the
total atmospheric emissions of arsenic. Hence, a control
alternative for mobile sources of air pollution will not
be further considered.
6. Regulating the land disposal of arsenic-bearing slags,
flue dusts, sludges, and other residuals from industrial
sources.
Large quantities of arsenic and its compounds are in the form of indus-
trial and commercial wastes destined for land disposal. Obese slags, sludges and
collected flue dusts are derived from the primary and secondary non-ferrous metals
industries, from the primary ferrous metals industry, and from the phosphorus
chemicals industry. The wastes are of varying physical forms, chemical forms,
and concentrations of arsenic; and represent correspondingly varying dangers to
the environment. Almost always, arsenic is but one of several or many hazardous
constituents in these land-destined wastes; toxic heavy metals often accompany
arsenic and a
-------
containers, but refuse large amounts. When the hazardousness level is relatively
low, due either to the inherent characteristic of the compound or its low con-
centration in the overall waste mass, even large quantities of hazardous wastes
may be accepted. Some arsenic-bearing slags may be disposed of in general pur-
pose landfills; since the arsenic constituent may be at a very low concentration,
it may be virtually insoluble (as a stable arsenate, for example}, and it may
be in a fixed physical form (in a stable aggregate, for example).
Each general purpose landfill has its own ambience - geologically, hy-
drologically, and environmentally. Ideally, a general purpose landfill would be
located in an isolated, dry part of the country with a thick layer of impermeable
soil between the waste and the water table. Such areas are plentiful in the western
part of the U.S., but not in the east. However, many existing and future landfill
sites throughout the U.S. can approach conditions which would classify them as
approved landfills, by meeting the following criteria:
(a) The composition and volume of each hazardous waste is
known and approved for site disposal by pertinent regu-
latory agencies.
(b) The site should be ambiently suitable for hazardous wastes.
(c) Provision is made for monitoring wells, rain water diversion,
and leachate control and treatment, if required.
The advantages of approved landfill sites include:
(a) Many hazardous wastes may be disposed of in a controlled
and environmentally safe fashion.
(b) Selection of landfill sites and disposal technology for
ambience suitability still leaves a great number of available
landfill sites.
(c) Disposal costs, for both transporting the waste to the site
and the landfilling itself, are kept to levels close to those
for general purpose sites and still much lower than for secured
landfill.
-195-
-------
From a practical standpoint many local regulatory agencies and landfill
site owners are informally practicing much of this tliscrimination by selective
acceptance of waste materials. Sites with known high potential for surface and
ground water contamination are thereby avoided.
Secured landfills involve additional safeguards beyond those described
for approved landfills. Criteria for secured landfills include:
(a) The. composition and volume of each extremely hazardous
waste is known and approved for site disposal by pertinent
regulatory agencies.
(b) The site should be geologically and hydrologically approved
for extremely hazardous wastes. Included in the criteria
would be a soil or soil/liner permeation rate of less than
10 cm per sec, a water table well below the lowest level
of the landfill, and adequate provision for diversion and
i control of surface water.
(c) Monitoring wells are provided.
: (d) Leachate control and treatment (if required).
(e) Records of burial coordinates to avoid any chemical
interactions.
(f) Registration of site for a permanent record once filled.
A number of landfills which meet the physical requirements (if not all
the regulatory criteria) are located around the country. California has a number
of Class 1 impermeable landfills which accept extremely hazardous materials.
Texas has similar sites. A number of low level-radioactive waste landfill sites
accept industrial hazardous wastes. In addition to the radioactive waste sites
various other private secured landfills also take extremely hazardous wastes.
At the present time secured landfills are scattered and not fully utilized. Part
of the lack of utilization stems from the fact that the majority of the sites are
in isolated western areas away from industrial centers. Another reason for the
lack of utilization is the high cost as compared to other available disposal methods.
-196-
-------
Relatively Isolated inpermeable soil conditions exist in many areas of
the country. If impermeable soil is not available then clay, special concrete,
asphalt, plastic and other liners and covers are available to accomplish similar
containment and isolation of wastes.
A number of practices are being used to ensure the environmental ade-
quacy of hazardous waste disposal.
Direct hazardous wastes encapsulation in concrete is now practiced by
at least one contract disposer. The practice is used for small quantities of
containerized miscellaneous hazardous wastes.
Steel drums, alone or with plastic liners, not only provide some long-
term containment but also are the most convenient storage and transportation for
relatively small quantities of wastes. The ultimate problem involved is the
eventual decay of the steel drums. Therefore, unless disposed of in an appropriate
landfill site, future release to the environment is likely.
In wet climates, sections of or entire landfill areas are encapsulated
by adding clay or asphalt "caps" or "covers" to impervious isolation cells or land-
fill liners.
The impervious cover is necessary to protect the hazardous waste from
rainfall flooding. Neutralizing or pH control ingredients such as lime may also
be used to encase or surround the hazardous waste to avoid solubility, decom-
position or other change in the character of the waste to increase its environ-
mental damage.
In dry climates, there is no need to encapsulate the entire landfill
since rainfall and water buildup is not a problem. Isolation cells may still be
constructed, however, for specific hazardous waste containment.
In wet climates, particularly, both private and public landfills are
paying increasing attention to leachate collection, monitoring and treatment.
Tiandf 111 areas in the State of Pennsylvania are representatives of those in a wet
climate and leaching treatment has been initiated in some public landfill areas.
The vast majority of the landfill operations handling hazardous wastes, however,
do not have any leachate control and treatment provisions.
-197-
-------
Hazardous sludges are being increasingly treated either on-site or in
collection areas by mixing them with, inorganic chemicals and catalysts to set
up the entire mass into solid structures with low !<>achability and good land
storage or landfill characteristics. There are a number of such processes which
produce solids ranging from crumbly soil-like materials to concrete to ceramic
slags.
Once a landfill area has been isolated from surface, and groundwater
contact and leachates are being handled satisfactorily, almost any non-flammable,
non-explosive and non-air polluting hazardous waste can theoretically be disposed
of safely. There are a number of practical restrictions, however, to this ap-
proach:
(a) In wet climates the impervious landfills are flooded with-
heavy rainfall. Dumping of liquids or sludges into the land-
fill only accentuates the problem.
(b) Some hazardous wastes create hazards for landfill personnel
or give air pollution problems.
(c) Chemical interactions with both, other materials and the
: liner can cause undesirable side effects. i
Control Alternative Aimed at the Commercial Use of White Arsenic
The most direct control alternative is a ban on white arsenic consumption,
either on the basis of selective uses, or upon all uses (i.e., a total ban). There
are several strong arguments against banning white arsenic use:
1. Any actions directed at commercial white arsenic and its
derivatives, even if totally effective in halting all
emissions and dissipations related to commercial uses,
would only address a small fraction of the total arsenic
quantities mobilized in our economy. Much more arsenic
is unintentionally mobilized than is intentionally mobilized.
Of all the arsenic that is mobilized, the comparatively small
quantity intentionally used is the only portion that serves
useful purposes in our society.
-198-
-------
2. Even if such, actions could be effective in halting emissions
and dissipations, they could not Be justified (except for
the case of airborne emissions) in terms of demonstrated
health, hazards. The data gathered in this study indicate
the opposite: that arsenic in water, in food, in the soil,
and generally dissipated in the environment as it is now
presents no identifiable health hazard.
However, tlie emissions to the. air from the intentional commercial use
of white arsenic are sizable and do involve a potential hazard to health. There
is sufficient justification, therefore, to retain the alternatives (for further
consideration) of selective or total bans upon white arsenic use.
The one exception is the use of white arsenic for very small-volume
and specialized items. These are included in the "miscellaneous uses" category
in this report, and include semiconductors, light-emitting diodes, and special
glasses for infrared applications. The health hazards from such uses appear
negligible, while the usefulness of arsenic appears quite important.
Needs for Additional Research
One of the results of this study is that very large quantities of arsenic
are mobilized by the primary iron and steel industry. Further research is needed
to validate the quantity estimates made in this study; to validate the hypotheses
made in this study of the distribution of the arsenic to end products, to land,
to water, and to ths air; and to determine the environmental adequacy of the wide-
spread use of arsenic-bearing steelmaking slags.
This study made apparent that several emerging technologies will mobilize
very large quantities of arsenic, comparable in magnitude to all the arsenic
mobilized by existing commercial activities. These emerging technologies are coal
gasification, oil shale processing, and geothermal energy recovery. Since the
Government is playing an active role in the research and development of these
emerging technologies, appropriate Government agencies (Environmental Protection
Agency, Energy Research and Development Agency, and Department of the Interior)
could take the initiative in developing effective arsenic removal and disposal
-199-
-------
techniques as integral parts of these processes/ highlight the fate of arsenic
in Environmental Impact Statements, and develop appropriate regulations as the
technologies emerge.
The application by spraying of arsenical pesticides results in relatively
large quantities atomized or evaporated. Since airborne arsenic trioxide is an
identified danger to human health/ further research is deemed necessary to quantify
the hazards to the general population (other than farm workers) from such prac-
tices, and to seek, techniques for pesticide application which reduce the quantities
lost to the atmosphere or which reduce the range of travel of these airborne
pesticides.
-200-
-------
SECTION X
COSTS OF AlflEHSIATIVE KEGULMTICNS
Bans Upon White Arsenic Use
i
The most direct ccntrql alternative upon the ccmnercial flew of white
arsenic and its compounds is a ban upon its use. Such a ban can be all-encompassing,
or it can be for selective uses. This section is intended to estimate the costs of
such bans, to provide information (along with a separate assessment of the feasi-
bility, effectiveness, and benefits of such bans) for evaluating alternatives.
The costs estimated for each control alternative are in terms of dollars
per kilogram of white arsenic diverted from dissipation via the use in question.
A comparison of control options on the basis of dollars per kilogram of white
arsenic diverted is potentially misleading unless recognition is made of the bene-
fits to human health and to environmental quality from each such diversion. The
eventual choice of control measures should ideally be based upon the cost per unit
reduction in health damage. Although the correlation between quantities of white
arsenic emitted and health damage has been discussed, the basis for a quantitative
estimate of healtii damage does not yet exist. For the purposes of this section,
therefore, the benefits of a control alternative will be assessed in terms of
quantity of white arsenic diverted from dissipation, with the results regarded
as the results of a screening mechanism of candidate options. Without a more
precise measure of health benefits, it is not possible to identify the most cost-
effective options or to determine the amount of diversion societally desirable.
The purpose of this section is to provide an indication of the available options,
their likely effects, and the probable costs — important steps in selecting con-
trols to be instituted.
The control options evaluated here are those which seemed most feasible
in preliminary review. Many alternatives were considered and some were rejected
for detailed analysis because the costs appeared too great for the perceived bene-
fit on an a priori basis; others were rejected because their effectiveness was
shown to be too small.
-201-
-------
The costs of a control alternative can be broken into two broad categories.
Long-run costs are derived from the differences between the two "steady states",
one without a control alternative and one with; these costs extended indefinitely.
The short-run costs are those incurred while moving from the steady state without
a control alternative to one with a control alternative; these costs have a
termination when the steady state with a control alternative is reached.
A second important distinction among costs is that between those involving
direct monetary outlays and those which are felt in other ways. If an emission
standard were adopted, then the control and treatment cost is an out-of-pocket
expense. If the quantity produced decreases, however, then the foregone consumer
surplus is a cost despite the fact that there is no direct monetary outlay.
A ban on white arsenic and its derivatives results in a cost in forcing
people to use substitutes; e.g., in forcing users to forego the benefits of arsenicals
over and above the next best substitute. Although there are substitutes in virtually
every major use of white arsenic, they are not perfect substitutes. Sometimes
they cost more, sometimes they don't provide the same quality product, and some-
times they don't last as long. The mere fact that white arsenic is being used
verifies that it has advantages ever the next best substitutes. It is possible
that some uses are not justified at current market prices, but it is inconceivable
that all uses are unjustified.
The long-run cost to society of foregoing the present and future benefits
of arsenical products is called the foregone benefits cost, and occurs each year
a ban is operative. It is made up of the foregone benefits to users and the fore-
gone benefits to producers. The former is the difference between the market price
and the value of white arsenic in various uses (the amount that users would have
been willing to pay to have white arsenic available for each purpose.) The latter
is the difference between the market price and the cost of producing white arsenic
for the market.
Figure 5 is a simple market description illustrating foregone benefits
from a ban which prohibits the Q, consumption of white arsenic for, say, herbicides.
Users forego benefits equal to area P2P3A whil® producers forego benefits equal
-202-
-------
FIGURE 5
FOREGONE BENEFITS
i
-203-
-------
to area P1P2A* ^ a^]er words» users pay only the P, market price, but the actual
value/unit of arsenic to them is the average value on the demand curve between
P_ and A. Likewise, producers receive price P2, but the average cost/unit is only
the average value on the cost curve between P, and A.
By determining the value of white arsenic from demand curves, we auto-
matically consider the possibility of white arsenic substitutes. The difference
between the D.. specified demand for herbicides and the lower D2 demand for arsenicals
as a specific herbicide reflects opportunities for substitutes. Stated otherwise,
the foregone benefits to users from a ban on white arsenic would be P2P4B i113*63^
of the smaller P2P3 ^ arsenic had no substitute. Moreover, the slope of the demand
curve Dj is determined by the relative price, availability, and effectiveness of
substitutes. However, although presently-available substitutes are represented
in the demand curve, new substitutes that could be developed are not generally
included, even though they can reduce the long-^run foregone benefit cost signi-
ficantly.
There are two basic ways to determine foregone benefits to white arsenic
users and producers. One approach is an engineering analysis—an analysis that
a user himself would employ in determining what he is willing to pay for white
arsenic and its substitutes, or that a producer himself would employ in determ-
mining hew much to produce at each price. A second approach is to trace out
the demand curve (e.g., the curve P.,A in Figure 5) from (1) observed changes in
market prices and quantities and (2) opinions of experts among suppliers and
consumers.
The estimates for this study were developed under the second approach.
The first approach is very expensive and subject to significant errors from in-
accurate or incomplete information.
Arsenic demand curves developed for this study indicate that sub-
stitutes for most uses are much more expensive or are so inferior that the current
$200 per metric ton market price could increase to $700 to $800 per metric ton
before users would completely cease using arsenicals. By definition, the white
arsenic user who would pay as much as $700 per kkg would forego a minimum of $500
benefit per metric ton if the As20_ he can now purchase at $200 per kkg is banned.
-204-
-------
The white arsenic demand curve P^ in Figure 5 is drawn to indicate that
market price is nearly equal to consumer value for part of the uses, but con-
sumer value is significantly above market price in other uses.
In the long run, foregone benefits are paid by the many consumers of
products in which white arsenic is a component. It is erroneous to think that
the benefits of white arsenic uses are obtained by a few producers acting against
the public interest while the benefits of less arsenic pollution are to be
enjoyed by the general public,, In the short run, producers of arsenicals and
manufacturers that use arsenicals in their products will suffer losses from a
ban. However, in the long-run, suppliers reach a new equilibrium via copper
prices and the consumers bear the loss of foregone benefits via higher copper
prices and via higher prices or lower quality of products containing sub-
stitutes for white arsenic.
Estimates of foregone benefits are certainly subject to error. However,
they are often a significant cost to society whenever there is a ban on products
for environmental or any other reason; therefore, foregone benefits must be
estimated to provide a complete accounting of social costs and they must be
analyzed if we expect to make rational decisions on arsenic controls. Estimates
in this report are objective estimates of cost consequences of specified arsenic
control alternatives.
Another long-run cost, in addition to the foregone benefits cost, is the
cost of disposing of the excess white arsenic in environmentally-adequate ways.
Since white arsenic is a byproduct, As2°3 and its derivates which cannot be sold
(because of a ban on use) must continuously be collected and disposed of in a secured
landfill. It should be pointed out that, by placing a ban on a certain form of
consumption, total dcroestic production is unlikely to fall by an amount equal to
that form of consumption. Slacks in demand are more likely to result in decreased
imports. Carried further, sufficient slacks in demand would result in exports
of white arsenic. This aspect further confuses an estimate of this cost.
-205-
-------
There are several types of short-run costs tJiat nust be accounted for:
1. Capital that becomes obsolete or reduces in value — If arsenicals
are not available for a specific production process, then either
the capital will be used for other purposes in its present
state, it will be converted for other purposes, or it will lie
idle, depending upon the costs of conversion and the perceived
productivity of the capital in a new function. The cost of a
white arsenic ban to society depends upon how much of the bene-
fits which could have been provided by the existent capital
can be reclaimed. By introducing a time lag between the announce-
ment of a ban and its institution, these costs can be reduced
significantly.
2. Unemployment — As a specific production is halted by a ban
on white arsenic, the labor involved in that production could
become unemployed. The cost depends upon the amount of time
unemployed and their productivity in new jobs relative to
the old jobs. By introducing a time lag between the announce-
ment of a ban and its institution, these costs can be reduced :(
significantly.
3. Stockpiling — If the demand for white arsenic is reduced,
stockpiling is likely to occur as a short-run response. The
cost is the opportunity cost of using the resources which go
into stockpiling. The drop in demand will result in fewer
imports and a short-run stockpiling at the smelter for a
selective ban. For a total ban, stockpiling would probably
not occur since the smelters would have no reason to refine
the white arsenic to stockpile.
Because an estimate of the amount stockpiled is dependent on
so many unknown variables, this cost will not be quantified.
However, since the cost is probably small and it only exists
in the short-run, this will not affect the results appreciably.
-206-
-------
Using the above outline as a guide, the long-run and short-run costs
resulting from a selactive ban on each of the primary uses of white arsenic
will be considered, as well as a total ban on all forms of white arsenic con-
sumption.
Estimation of Foregone Benefits (Long-Run costs)
Using the estimated white arsenic demand curves of Figure 3, the foregone
benefits of ban on each of the uses was calculated as the area between the de-
mand curve and the supply curve. The results are tabulated below in terms of
the annual foregone benefits for each selective ban, and the foregone benefits
per kilogram of arsenic trioxide diverted for each ban. The miscellaneous uses
of white arsenic were not included in this analysis, as substitutes are not
generally available for specialized uses.
AszOi Use Ban
Assured
Insecticides
Dessicanta and Defoliants
Herbicides (Weed Control)
Soil Sterilizers
Wood Preservatives
Feed Additives
Glass Additives
1total Ban
AsaOs Diverted
Wcgyfyear
5,500
3,500
5,800
4,200
1,550
550
2,400
23,500
Foregone Benefits,
Million Dollars/yr
$ 2.94
2.01
2.87
2.33
1.06
0.41
1.11
$12.73
Foregone Benefits
$/kkg Diverted
$530
570
500
560
680
740
460
$540
The foregone benefits per metric ton of white arsenic diverted amount to
$540 for a total ban, and to approximately that amount for individual bans upon
the agricultural uses. The foregone benefits per metric ton are somewhat lower
for glass additives (reflecting the moderate elasticity of this curve in Figure 3);
and are somewhat higher for wood preservatives and feed additives (reflecting the
inelasticity of these curves in Figure 3).
-207-
-------
Estimation .of Disposal .Costs .for .Excess
(long-Run)
The 1974 domestic white arsenic production was estimated at 8,700 metric
tons. The total domestic demand for white arsenic (24,000 metric tons in 1974)
would not be reduced to the point where domestic production would be curtailed for
any one individual use ban; it appears reasonable that reduced imports would be
the result of any single ban. Hence, there would be no disposal cost associated
with any single assumed use ban.
However, a total ban on white arsenic would mean that the 8,700 metric
tons must be disposed of in an environmentally-adequate manner. A unit cost for
such land disposal (i.e., secured landfill), including transportation, is about
$50 per metric ton,(75'123'136) so that the total cost would be $435,000 per year.
In lieu of analyzing the world market to determine if any or all of the excess
white arsenic could be exported, the maximum costs for disposal will be assumed.
Estimation of Short-Run Costs for As2°3 Use Bans
The short-run costs for a ban on white arsenic include the idle capital
and unemployment in the manufacture of white arsenic, and the idle capital and
unemployment in the industries using white arsenic.
The 1967 and 1972 Census of Manufacturers for SIC 2819, Industrial
Inorganic Chemicals, N.E.C. (in which white arsenic manufacture is classified);
and for SIC 2879, Agricultural Chemicals, N.E.C. (in which most arsenical products
are classified); contain the following statistics:*151'156'157)
SIC 2819, 1972 Census
SIC 2819. 1967 Census
SIC 2879, 1972 Census
SIC 2879, 1971 ASM
SIC 2879, 1970 ASM
SIC 2879, 1969 ASM
SIC 2879, 1968 ASM
SIC 2879, 1967 Census
Valua of Shipments
Million Dollars
3,657.5
4,248.4
1,150.8
963.9
859.0
976.7
902.4
817.0
Nurtoer of
Bnployees
60,600
•81,200
12,200
11,900
12,200
12,300
12,100
11,500
Payroll,
Million Dollars
666.9
662.4
116.5
102.6
101.9
94.6
85.2
80.7
Gross Value of
Fixed Assets,
Million Dollars
-
-
-
410.5
410.5
374.4
314.6
272.9
-208-
-------
These statistics were converted into the following average ratios:
a) 14 employees per million dollars/year of shipments
b) Annual wsges per employee are about $10,000.
c) The gross value of the fixed assets are 40 per cent of
the annual value of shipments.
Applying these ratios to the manufacture of white arsenic, where the
value of shipments in 1974 (8,700 kkg at $200/kkg) was $1.74 million, yields the
following values:
25 employees
$250,000 annual payroll
Gross Value of fixed assets = $700,000
Since the domestic white arsenic production, is only 36 percent of the domestic
consumption (the balance being imports), only a total ban would result in un-
employment or in idle capital (selective individual use bans should instead
result in decreased imports). If the average length of unemployment caused by
a total ban were one year, the associated unemployment cost would be $250,000.
The white arsenic production facilities would probably have no salvage value.
On the other hand, 2iese facilities are not new, and the present (depreciated)
value, taking into account possible recent additions for pollution control and
other reasons, is crudely estimated at $350,000. If this present value is
amortized over 5 years, the annual idle capital cost would be $70,000.
Table 21 lists the arsenical derivative products discussed in Chapter
VIII, along with their quantities, prices, and value of shipments. The substitutes
for these products ate organics, and it is assumed that the equipment used for
manufacturing the arsenicals could not readily be converted for manufacturing the
substitutes and that unemployment would result. Apply the ratios developed above,
the arsenical derivatives industries, with a value of shipments of $52.9 million,
is estimated to have:
740 employees
$7.4 million annual payroll
Gross Value of fixed assets = $21.2 million
-209-
-------
Table 21
Economics of Arsenical Derivative Products
1974 Basis
it,
I-1
o
Product
Insecticides
Dessicants
Herbicides
Soil Sterilizers
Wood Preservatives
Peed Additives
Total
Quantity
kkg/year
11,700
4,400
9,800
4,200
6,000
1,360
37,460
Avg. Price
$Akg
625
730
2,100
415
2,100
5,500
-
Value of Shipments
Million Dollars/Year
7.3
3.2
20.6
1.7
12.6
7.5
52.9
Employees
100
45
290
25
175
105
740
Annual
Payroll
$1,000,000
450,000
2,900,000
250,000
1,750,000
1,050,000
$7,400,000
Fixed
Assets
$ 2,900,000
1,300,000
8,300,000
700,000
5,000,000
3,000,000
$21,200,000
Unemployment
Oasts
(6 mos)
$ 500,000
225,000
1,450,000
125,000
875,000
525,000
$3,700,000
Idle Capital
OostsAr
(10 yrs)
$ 290,000
130,000
830,000
70,000
500,000
300,000
$2,120,000
-------
The short-run costs are estimated based upon a 6-roonth employment period; and
upon a present value of assets equal to 50 parosnt of the gross value and a
5-year amortization period for this present value. The results of this estima-
tion are shown in Table 21.
More preexist estimates of the short-run costs would of course be desirable,
but the e:cror should not affect the final cost estimate appreciably.
Summary of the Posts for Banning White Arsenic Use
Table 22 summarizes the costs for each selective ban and for a total ban
on arsenic use. As Table 22 shows, the foregone benefits are the predominant
costs for all but feed additives (where the value added is very large compared to
the arsenical raw material cost).
For a total ban on arsenic use, the overall first-year costs would be
$20.0 million. The costs for each of the next four years would be $16.0 million,
and the annual costs thereafter would be $13.2 million.
Based upon a consumption of 24,000 metric tons of white arsenic, the costs
of a total ban per iretric ton of white arsenic are $830 for the first year, $665
for the next four years, and $550 thereafter.
Posts of Controlling Industrial Arsenic Emissions to the Atmosphere
The most important need for additional controls is the reduction of arsenic
trioxide emissions to the atmosphere from high-temperature industrial processes.
The primary copper industry is the largest source of such emissions; an estimated
6,300 metric tons per year of As20_ are emitted. The total AsJD., in copper roast-
ing and smelting flue gases amounts to an estimated 30,000 kkg/year, implying that
23,700 kkg/yr are collected and that the collection efficiency is 79 per cent. An
additional 11,600 kkg/yr of AsJD., in converter flue gases are removed in byproduct
acid cleaning plants.
The roasting and smelting flue gases are typically passed through dry dust
collection systems, which fall far short of effective (i.e., 99+ per cent) As203
capture. Some high-residence-time devices such as "balloon flues" are used but
also with limited success. It appears that high-pressure drop venturi scrubbing
-211-
-------
Table 22
Sumnary of Costs of Selected and Total Bans
Costs in Millions of 1974 Dollars Per Year
AsaOj Use Ban
Assumed
Insecticides
Dessicants
Herbicides
Soil Sterilizers
Wood Preservatives
Feed Additives
Glass Additives
As2O3 Production
Total Ban
long-Run Costs
Foregone Benefits
2.94
2.01
2.87
2.33
1.06
0.41
1.11
-
12.73
Disposal
-
-
-
-
-
-
-
6.44
0.44
Short-Run Costs
Umirployrrunt
(Firt;t Year Only)
0.50
0.23
1.45
0.13
0.88
0.53
-
0.25
3.97
Idle Capital
(Five Years)
0.29
0.13
0.83
0.07
0.50
0.30
-
0.70
2.82
-212-
-------
systems are the technically-feasible controls for As2°3 in these flue gases.
Such systems are used in cleaning the converter flue gases prior to manufacturing
sulfuric acid/ and are effective to the point where the commercial acid contains
only 0.5 ppm arsenic/ implying 99+ percent removal of As2°3* ^e f°llowin<3
analysis leads to an estimate of the costs for controlling As2^3 eroissi0318 from
copper smelters.
The sulfur/copper ratio in copper concentrates is nominally about 1.15,
and about one-third of the sulfur is lost in the roasting and reverberatory
(smelting) steps. S0_ in these flue gases is nominally at about a 4 percent
volumetric concentration, although newer plants are being designed to yield
higher SO2 concentrations so that it may be captured more economically. For the
conventional plants, the above data permits the estimation of the quantity of
flue gases from roasting and smelting: about 6,700 cubic meters (STP) per metric
ton of copper. If these gases are passed through a waste heat boiler and an
electrostatic precipitator, they should be at about 250°C and 1 atmosphere, so
that the gas volume prior to wet scrubbing would be about 13,000 actual cubic
meters per metric ten of copper.
A "typical" smelter is defined as having an annual copper production of
100,000 metric tons (there would be 16 such typical smelters equivalent to the
current U.S. copper production of 1.6 million metric tons). The throughput of
this typical smelter is on the average about 0.20 metric tons of copper per minute;
the roasting and smelting flue gas flow rate would then be 2,600 actual cubic
meters per minute (92,000 actual cubic feet per minute).
The 1967-68 total capital cost (purchase cost plus installation cost) for
a high-efficiency (99.5 percent) venturi scrubber with a capacity of 92,000 ACFM
was $220,000. Updating this cost to 1975 with the Chemical Engineering Plant
Cost Index results xn a capital cost of $365,000. The annual operating cost is
about 5 percent of the total capital cost, or about $20,000 per year.* * ~
The scrubber liquor would likely be recirculated, with a relatively small
fraction bled for removal of arsenic and other contaminants. Hypothetically, the
scrubber bleed may be treated with lime followed by sedimentation; alternately,
-213-
-------
it may be treated with sodium sulfide or sodium hydrosulfide (as in food-grade
phosphoric acid manufacture) with subsequent removal of As2S3 ^ filtration.
Such a conventional system for treatment of the scrubber bleed should cost
approximately $150,000 (installed) for the "typical" plant; with annual operating
costs of perhaps $20,000.
The "typical" plant would then have solid wastes from the scrubber liquor
treatment of perhaps 500 metric tons per year. At a disposal cost of $50 per
metric ton in a secured landfill, these costs would amount to $25,000 per year.
In this analysis, no credit will be taken for possible recovery values from these
wastes.
In summary, then, this "typical" plant would have overall capital costs
of about $515,000 and annual operating costs of about $65,000. For the entire
primary copper industry, made up of 16 such typical plants, the costs for re-
moval of most of tire 6,300 metric tons per year of As-O- would be a capital cost
of $8.3 million and an annual operating cost of $1.0 million. If the capital
investment were amortized over 10 years, the total annual cost would be about
$1.8 million; or about $300 per metric ton of As203 removed.
Very little of the As20_ should pass through such a high-pressure-drop
wet scrubbing system. In actuality, the major emissions of As203 should then be
attributable to flue gases which never are collected; i.e., the leaks and spurious
emissions from the smelting process equipment. Since the total quantity of AS203
in all copper flue gases amounts to about 42,000 metric tons per year, a one per-
cent loss of such gases is equivalent to an emission of 420 metric tons per year.
Costs of Controlling Arsenic Emissions from Fossil Fuel Combustion Stationary
Sources
The same control technology, i.e., high-performance wet scrubbing systems,
could be applied to the flue gases from electric power generating stations and
other stationary sources which burn fossil fuels.
Using a factor of 10 cubic meters (SIP) of flue gas generated per kilogram
/q\
of coal burned,l ' a "typical" power plant that burns 100,000 metric tons of coal
-214-
-------
per year (190 kg/mir) generates 1,900 cubic rasters (STP) per minute of flue gas.
Assuming a flue gas tetrperature of 250°C, the flue gas flow rate would be 3,650
actual cubic meters per rainute, or 129,000 actual cubic feet per minute.
The 1967-68 total capital investment (purchase cost plus installation
cost) for a high-efficiency (99.5 percent) venturi scrubber with a capacity of
129,000 ACEM was $300,000.(159) Updating this cost to 1975 with the Chemical
Engineering Plant Cost Index results in a capital cost of $500,000. The annual
operating cost is nbout 5 percent of the total capital cost, or about $25,000
per year.(159)
As in the case for the previous analysis (for the capper smelter), a system
for treating and recirculating the scrubber liquor would be required. In a less
demanding situation than exists at a copper smelter, the installed cost of this
system may amount to $100,000, with annual operating costs of perhaps $15,000.
The "typical" power plant would then have hazardous wastes from the scrubber liquor
treatment of perhaps 100 metric tons per year. At a disposal cost of $50 per
metric ton in a secured landfill, these costs would amount to $5,000 per year.
This "typical" power plant would have, then, a total capital cost of
$600,000 and annual operating costs of $45,000. For the entire U.S. population of
coal-burning power plants (4,500 such "typical" plants), the required capital cost
would amount to an estimated $2.7 billion, and the annual operating cost to $200
million. If the capital investment were amortized over 20 years, the total annual
cost would be about $335 million. Even if these costs were apportioned among all
the hazardous materials removed by such control systems, an estimated 10 percent
accountable to arsenic would be $33.5 million per year. Since the total quantity
of arsenic in present atmospheric emissions from coal combustion is 650 metric tons
per year, the costs of such a control measure would be about $50,000 per metric
ton of arsenic removed, or $39,000 per metric ton of As2°3 removed-
Costs of Safe Disposal of land-Destined Wastes
Large quantities of arsenic and its conpounds are in the form of industrial
and commercial wastes. Slags, sludges, and collected flue dusts from a variety
-215-
-------
of sources contain arsenic and other hazardous substances. An estimate of the
total waste quantity, the arsenic content, and the total hazardous constituents
is:
Source
Primary Zinc (Pyro) (75)
Primary Lead Industry (75)
Primary Copper Industry
Other Pri. Ncn-flerrous Metals (75)
Phosphoric Acid Sludges
Manganese Smelting Dusts
n m
Iron and Steel Dusts lijj;
Goal Ocnbustion Ash
Totals
Itotal Hazardous
Wastes,
kkg/year
288,000
542,000
6,089,000
30,000
1,000
5,000
1,951,000
45,000,000
54,000,000
Total Hazardous
Constituents,
kkg/year
47,200
61,200
95,200
500
150
1,000
20,000
15,000
245,000
Arsenic
Quantity,
kkg/year
120
800
12,000
50
90
350
1,350
1,800
17,560
The costs for environmentally-adequate disposal range from 0 to $50 per
metric ton of wastes. The lower costs are applicable to slags, where the arsenic
and other hazardous constituents may already be chemically fixed (as arsenates,
etc.) and so not susceptable to leaching. The higher costs are applicable to
lined ponds, impervious landfills, concrete pits, collection and treatment of
leachates, surface protection from dispersion of dusts, chemical fixation of
sludges and dusts, etc.
In the major non-ferrous primary metals industries (zinc, lead, and
copper) the overwhelming majority of the total wastes are slags, rather than
sludges or dusts. Costs for environmentally-adequate land disposal have been
estimated.
(75)
Industry
Primary Copper
Primary Lead
Primary Zinc (Pyro)
Metal Production
kkg/yr
1,600,000
610,000
290,000
Disposal Cost
Per kkg Product
$1.29
1.37
4.20
Disposal
CostAr
$2,060,000
840,000
1,220,000
-216-
-------
These costs, and the disposal costs for the arsenic-bearing wastes from
other industrial soc'rces, are only partially attributable to the control of
arsenic pollution, since /other hazardous constituents are in these wastes.
For the primary copper industry, which is the source of three-fourths
of the arsenic in all land-destined wastes, the total estimated costs are $2,06
million per year. If these costs are apportioned among all the hazardous con-
stituents (totalling 95,200 kkg/year), the share to be borne by controlling arsenic
(13,000 kkg/year) would be $280,000 per year, or about $22 per metric ton of
arsenic. Using this unit cost to extrapolate to other sources, the total apportion®!
cost for environmentally-adequate disposal of arsenic wastes would be about
$380,000 per year.
-217-
-------
SECTION XI
REFERENCES
1. Carapella, S.C., Jr. Arsenic. In: Encyclopedia of Chemical Technology,
R.E. Kirk, and D.F. Othmer (eds.I. John Wiley and Sons, Inc., New York,
1964. pp. 711-717.
2. Sullivan, Ralph J., Air Pollution Aspects of Arsenic and Its Compounds.
1969. (NTIS No. PB 188 071).
3. Lansche, Arnold M., Arsenic. In: Mineral Facts and Problems, preprint from
Bulletin 630. Bureau of Mines. 1965.
4. Schneider, Robert F., The Impact of Various Heavy Metals on the Aquatic
Environment. National Field Investigation Center. Denver, Colorado.
Feb., 1971. (NTIS No. PB 214 562).
5. Furguson, John F., and Jerome Gavis, A Review of the Arsenic Cycle in
Natural Waters. Water Research. 6/11):1259-1274, Pergamon Press, 1972.
6. Swain, Robert E., Wastes Problems in the Nonferrous Smelting Industry.
Industrial Engineering and Chemistry. 31X11):1358-1360, 1939.
7. Angino, E.E., L.M. Magnuson, T.C. Waugh, O.K. Galle, and J. Bredfeldt,
Arsenic in Detergents: Possible Danger and Pollution Hazard.J Science. ,
168_: 388-90. April 1970.
8. Anon., Trace Metals; Unknown, Unseen Pollution Threat. Chemical and
Engineering News. July 19, 1971, pp. 29-33.
9. National Inventory of Sources and Emissions: Arsenic - 1968. W.E. Davis
and Associates. Leawood, Kansas, 1971. (NTIS No. PB 220 619).
10. Anderson, David, Emission Factors for Trace Substances. Environmental
Protection Agency, EPA-450/2-73-001, December, 1973.
11. Miscellaneous data supplied by EPA, January 1975.
12. Phosphorus and Its Compounds, Vol. II, John R. Van Wazer, ed., Interscience,
New York, 1961.
13. Sauchelli, V., Manual on Fertilizer Manufacture, 2nd ed. Davison Chemical
Corp., Baltimore, 1954.
14. Sauchelli, V., Chemistry and Technology of Fertilizers, Reinhold. New York,
1960.
15. Superphosphate, Its History, Chemistry and Manufacture. GPO, Washington,
D.C. 1964.
-218-
-------
16. Arsenic Trioxide Salient Statistics. Chemical Economics Handbook. Stanford
Research Institute, Menlo Park, Calif., Sections 710-5020 A-D, January
1973.
17. Mineral Facts and Problems. Bureau of Mines. 1970.
18. Commodity Data Summaries 1975. Bureau of Mines. 1975.
19. From EPA Data Bank.
20. Dennis, W.H., Metallurgy of the Non-Ferrous Metals, 2nd edition. Pitman.
London. 1961.
21. Goodale, C.W,, and J.H. Klepinger, The Great Falls Flue System and Chimney.
Trans. AIME 46_: 467-647. 1913.
22. Reay, P.F., The Accumulation of Arsenic from Arsenic-Rich Natural Waters by
Aquatic Plants. J. Appl. Ecol. 9_(2): 557-565. 1972.
23. Lange, N.A., Handbook of Chemistry, 9th edition. Sandusky, Ohio, Handbook
Publishers, Inc. 1956. p. 59.
24. Arsenic. In: Encyclopedia of Science and Technology, Vol. 1. McGraw-Hill,
New York, 1971, pp. 598-599.
25. Sisler, H.H., et al., General Chemistry. The Maonillan Company, New York.
1959.
26. Doak, G.O., et al., Arsenic Compounds. Ins Encyclopedia of Chemical Tech-
nology, 2nd edition. John Wiley & Sons. New York 1963, pp. 718-733.
27. Vallee, B.L., D.D. Ulmer, and W.E.C. Wacker, Arsenic Toxicology and Bio-
chemistry. A.M.A. Archives of Industrial Health. 21^:132-151. Feb., 1960.
28. Braman, R.S., and C.C. Foreback, Methylated Forms of Arsenic in the En-
vironment. Science. 182_: 1247-124 9. Dec. 21, 1973.
29. Patty, F.A., Arsenic, Phosphorus, Selenium, Sulfur, and Tellurium. In:
Industrial Hygiene and Toxicology, 2nd edition, Vol. II Toxicology. John
Wiley and Sons. New York. 1962. pp. 871-880.
30. Whitacre, R.W., and C.S. Pearse, Arsenic and the Environment. Mineral
Industries Bulletin. 17(3). Colorado School of Mines, Research Institute.
May 1974.
31. Principles of Plant and Animal Pest Control, Vol. II, Weed Control.
Washington, D.C., Nat. Acad. of Sci., Pub. No. 1597. 1968.
-219-
-------
32. Water Pollution Control in the Primary Nonferrous-Metals Industry - Vol, I,
Copper, Zinc, and Lead Industries. EPA: Office of R and D. No.: EPA-R2-73-
247a. September 1973.
33. Wood, J.M., Biological Cycles for Elements in the Environment, and the
Neurotoxicity of Metyl Alkyls. In: Proceedings, 16th Water Qual. Conf.,
Trace Metals in Water Supplies: Cccurance, Significance, and Control.
Univ. of 111. Feb. 1974. pp. 27-37.
34. Wbolson, E.A., J.H. Axley, and P.C. Kearney, Correlation Between Available
Soil Arsenic, Estimated by Six Methods, and Response of Corn (Zea mays L.).
Soil Science Society of America Proceedings. 35(1): 101-105. Madison, Wise.,
1971.
35. Wbolson, E.A., Arsenic Phytotoxicity and Uptake in Six Vegetable Crops.
Weed Science. 2M6): 524-527. November 1973.
36. Woolson, E.A., J.H. Axley, and P.C. Kearney, The Chemistry and Phytotoxicity
of Arsenic in Soils: II. Effects of Time and Phosphorus. Soil Science
Society of America Proceedings. _37_(2):254-259. 1973.
14
37. Woolson, E.A. and P.C. Kearney, Persistence and Reactions of " C-Cacodylic
Acid in Soils. Environmental Science and Technology. 7/l):47-50. 1973.
38. Woolson, E.A., personal communication, May, 1975.
39. Statement to OSHA by Crosby F. Baker for the American Smelting and Refining
Company: In Re Proposed Standard for Occupational Exposure to Inorganic,
Arsenic.
40. Federal Register issues: May 11, 1968 and May 15, 1969.
41. Recommended Methods of Reduction, Neutralization, Recovery, or Disposal of
Hazardous Waste, Vol. VI, National Disposal Site Candidate Waste Stream
Constituent Profile Reports: Mercury, Arsenic, Chromium, and Cadmium Com-
pounds. EPA. August 1973. (NTIS No. PB 224 585).
42. Wood, R.A. Arsenic - Utilization/Availability. Metals and Ceramics Infor-
mation Center, Battelle Columbus Laboratories. Columbus, Ohio. July 1975.
43. Fowler, D.L. The Pesticide Review 1973. U.S. Department of Agriculture.
Washington, D.C.
44. Putman, R.B., Kpppers Company, Inc., statement to OSHA on arsenical wood
preservatives (made with regard to the proposed standard for occupational
exposure to inorganic arsenic).
45. Gill, T.C. and R.B. Phelps, Wood Preservation Statics 1973. Forest Service
(USDA) in cooperation with the American Wood-Preservers' Association.
Washington, D.C.
-220-
-------
46. Personal coimiunication with Paul Lepore, Food and Drug Administration,
April 1975.
47. Personal ocnupunication with Jerry Smith, Lead Industries Association,
May 1975.
48. The Pesticide Manufacturing Industry - Current Waste Treatment and Disposal
Practices. EPA, Water Pollution Control Research Series. 12020 FYE 1/72,
Chapter 3, Pesticide Production and Use.
49. Andrilenas, Paul A., Fanner's Use of Pesticides in 1971...Quantities.
Agricultural Economic Report No. 252, USDA. 1974.
50. Personal comnunication with Paul Andrilenas, USDA, March 1975.
51. 1975 Weed Control Manual. Agri-Fieldman Magazine, February 1975. Meister
Publishing Co. Willoughby, Ohio.
52. Farm Chemicals Handbook. Meister Publishing Co. Willoughby, Ohio, 1975.
53. Personal communication with Jim Campbell of the Texas Branch of the Pennwalt
Corporation, June 1975.
54. Personal communication with Jim Brown of the National Cotton Council,
August 6, 1975.
55. Personal cxanmunication with Parker Stokes, Ventron Corporation, July 1975.
56. Wood Preservative Standards. American Wood-Preservers1 Association.
Washington, D.C. 1974.
57. Personal communication with R.B. Putman, Kbppers Company, Inc., July 1975.
58. 1974 Feed Additive Compendium. Miller Publishing Co. Minneapolis, Minn.
59, Calvert, C.C., Arsenicals in Animal Feeds and Wastes. ACS Symposium Series,
No. 7, Arsenical Pesticides. Am. Chem. See., Wash., D.C. 1975.
60. Metals Handbook, 8th edition, Taylor Lyman, ed. American Society for Metals.
Metals Park, Ohio. 1961.
61. Personal comtrunication with Bill Hetrick, ESB (Electric Storage Battery),
Inc., Philadelphia, May 20, 1975.
62. Personal communication with Fay Tooley, Editor. Handbook on Glass Manufacture,
April 1975.
63. Personal communication with Cooper USA, Research Triangle Park, N.C.,
April 1975.
-221-
-------
64. Personal communication with Glen Schubert, U.S. Department of Agriculture,
April 1975.
65. Personal conrounication with Richard Murray, National Paint and Coatings
Association, April 1975.
66. Anon., Arsenic Makes a Comeback. British Medical Journal, 3/5929) :487.
August 1974.
67. AIME World Symposium on Mining and Metallurgy of Lead and Zinc, C.H. Cotterill
and J.M. Cigan, ed., AIME., N.Y. (1970):
a. J.L. del Valle, A. Fernandez, A. Rovira, J. Moreno, and R. Guzman,
Treatment of the Leaching Residues and the Electrolytic Zinc Plant
of Espanola del Zinc, S.A.
b. K.H. Heino, R.T. McAndrew, N.E. Ghatas, and B.H. Morrison, Fluid
Bed Roasting of Zinc Concentrate and Production of Sulfuric Acid
and Phosphate Fertilizer at Canadian Electrolytic Zinc Ltd.
c. S. Ganci, A Comparison of Ore Dressing Practices at Broken Hill,
Australia.
d. M.K. Foster, Horizontal Retort and Acid Plant, ASARCO Mexicana,
S.A., Neuva Rosito Plant.
e. R.E. Lund, J.F. Winters, B.E. Hoffacker, T.M. Fusco, and
' D.E. Warnes, Josephtown Electrothermic Zinc Smelter of St. i;
Joe Minerals Corporation.
68. Hallowell, J.B., J.F. Shea, G.R. Smithson, Jr., A.B. Tripler, and B.M.
Gonser, Water Pollution Control in the Primary Nonferrous-Metals Industry,
Vol. I., Copper, Zinc, and Lead Industries, EPA-R2-73-247a (September 1973).
69. Development Document for Interim Final Effluent Limitations Guidelines and
Proposed New Source Performance Standards for the Nonferrous Metals Manu-
facturing Point Source Category (February 1975):
a. Zinc, EPA 440/1-75/032
b. Lead, EPA 440/1-75/032-a
c. Copper, EPA 440/1-75/032-b
70. Minerals Yearbook 1972, Vol. I, U.S. Department of the Interior, Bureau of
Mines.
71. Anon., The Crisis in U.S. Zinc Smelting Spells Trouble for the Mining Indus-
try, Eng. & Mining Journal 173(2):69-74. Feb. 1972.
-222-
-------
72. American Metal Market, Fairchild Publications Inc., NYC. July 17, 1974.
73. Draft Development Document for Effluent Limitations Guidelines and Standards
of Performance for the Ore Mining and Dressing Industry Point Source Category.
EPA. April 1975.
74. Pyrometallurgical Processes in Nonferrous Metallurgy, J.N. Anderson and
P.E. Queneau, ed., AIME, Gordon and Breach Science Publishers, N.Y. 1965.
a. M. Okazaki, Y. Nakane, and H. Noguchi, Roasting of Zinc Concentrate
by Fluo-Solids Systems as Practised in Japan at Qnahama Plant of
Toho Zinc Ltd.
b. L. Harris, Lead Smelting Developments at LaOroya.
c. J.I. Elvarjder, The Boliden Lead Process
d. B. Glaus and A. Guebels, Copper Blast-Furnace Practice at Union
Miniere Du Haut-Katanga.
,e. W.H. Peck and J.H. McNicol, An Improved Furnace for Continuous
Copper Dressing of Lead Bullion.
75. Assessment of Industrial Hazardous Waste Practices in the Metal Smelting
and Refining Industry, Primary and Secondary Ndn-Ferrous Smelting and
Refining (Draft), EPA Contract No. 68-01-2604. April 1975.
76. Addicks, L., Copper Refining. MbGraw-flill, N.Y. 1921.
77. Development Document for Effluent Limitations Guidelines and New Source
Performance Standards for the Phosphorus Derived Chemicals Segment of the
Phosphate Manufacturing Point Source Category. EPA-440/l-74/006-a. Jan. 1974.
78. Lowell, J.D., Copper Resource in 1970. Mining Eng. 22_(4) :67-73. April 1970.
79. Beall, J.V., Copper in the U.S. - A Position Survey. Mining Eng. 25_(4):35-47.
April 1973.
80. Wright, H.M., The Bethlehem Project of Bethlehem Copper Corporation, Ltd.
Instn. Min & Met-Trans 74_:177-253. 1963-4.
81. Encyclopedia of Chemical Technology, 2nd edition. R.E. Kirk and D.F. Othmer,
eds. John Wiley and Sons, Inc. New York. 1964.
82. Butts, A., Copper. Reinhold, N.Y. 1954.
83. Gregg, J.L., Arsenical and Argentiferous Copper. The Chemical Catalog Co.,
Co., N.Y. 1934.
-223-
-------
84. Treilhard, D.G., Copper - State of the Art. Eng. & Mining Journal 174:P-Z.
April 1973. ~^~
85. Yazawa, A., and T. Azakami, Thermodynamics of Removing Impurities During
Copper Smelting. Canadian Met. Quarterly 8_(3): 257-61. July-Sept. 1969.
86. McMahon, A.D., Copper - A Materials Survey. U.S. Bureau of Mines 1C 8225.
1965.
87. Dunn, E.M., Determination of Gases in Smelter Flues. Trans. AIME'46:648-89.
1913.
88. Beals, G.C., and J.J. Cadle, Copper and By-Product Sulfuric Acid Production
by Palabora Mining Company. J. of Metals 20(7):86-92. July 1968.
89. Lutjen, G., The Environmental Confrontation in Copper. Eng. Min. J. 174(4):
E-I. April 1973.
90. Milliken, C.L., Copper Smelter Design for the 70's. Mining Eng. 23(5):48-50.
May 1971. ~~
91. Dayton, S., Inspiration Design for Clean Air. Eng. Min. J. 175(6):85-96.
June 1974; (8):63-74. Aug. 1974.
92. White, L., The Newer Technology: Where It Is Used and Why. Eng. Min. J.
174_:AA-HHH. Apr. 1973.
93. Elton, J.O., Arsenic Trioxide from Flue Dust. Trans. AIME. '46:690-702. ''
1913.
94. Bender, L.V., and H.H. Goef Production of Arsenic Trioxide at Anaconda.
Trans. AIME. 106:324-8. 1933.
95. Pesticides. Chemical Process Industries, 3rd edition. R. Norris Shreve,
ed. New York, McGraw-Hill, 1967.
96. Criteria for a Recommended Standard...Occupational Exposure to Inorganic
Arsenic. U.S. Department of Health, Education, and Welfare, NIOSH. 1973.
97. Federal Register. 40(14), January 21, 1975.
98. Richards, Bill and Rachel Scott, Arsenic, Industry and Cancer. Washington
Post. p. B5, January 12, 1975.
99. Anon., Tight Rein on Arsenic. Chemical Week. JL16J7): 16-17. January 29, 1975.
100. Utidjian, H. Michael D., Criteria Documents. J. Occupational Medicine.
16(4):264-269. 1974.
u.S. EPA
-224-
------- |