PB84-212240
TOX (Total Organic Halogen)
Is It the Non-Specific
Parameter of the Future
(U.S.) Municipal Environmental Research Lab.
Cincinnati, OH
Jun 84
¦iwammsamsm
qyaaoammmmmassmm
r?
TMF¥7im^m /i v Hiii' fiMfFffrrryrra
U.S. Sepsri'!i€iit of C&nffiwee
Nations! Technical fefwmstioji Service
-------
EPA-600/D-84-169
June 1984
IS IT THE NON-SPECIFIC PARAMETER OF THE FUTURE?
by
Alan A. Stevens and Ronald C. Dressman
Drinking Water Research Division
Municipal Environmental Research Division
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
R. Kent Sorrel1 3nd Herbert J. Brass
Technical Support Division
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, D.C. 20460
MUNICIPAL ENVIRONMENTAL RESEARCH LABOIIMORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, Oil A526S
-------
TECHNICAL REPORT DATA
(Phaie trad fnwutiiom on the rr>rrn* bnt/rf cumpleiinz)
l. BtroriT IVO. 2.
EPA-600/D-84-16?
3. R'.CIPtfcNT S ACCESSION NO.
PR? i> 21 ??
4. TITLE AMD SUOTtTLE
TOX, IS IT THE NON-SPECIFIC PARAMETER OF THE FUTURE?
5. REPORT DATE
June 1984
6. PERFORMING ORGANIZATION CODE
?. AUTMORlS) a
Alan A. Stevens and Ronald C. Dressmai
Herbert J. Brass and R. Kent Sorrell
8. PERFORMING ORGANI/ATlOfJ REPORT NO.
9, PSRFOBM1MC ORGANIZATION name AND address
g
Drinking Water Research Division, MERL, ORD
^Technical Support Division, ODM
10. PROGRAM ELEMENT NO.
nmiA
11. CONTRACT/GRANT NO.
13. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory - Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 4526S
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
is. ass ;ract
A total organic halogen (TOX) analysis directly measures halogen (CI, Br, I)
covalently bound to organic molecules in a sample. Currently, the most popular
method for TOX in water involves carbon adsorption, oxidative combustion, followed
b; measurement of the formed hydrogen halide by roicrocouloraetry. TOX can be used
as an indicator of water quality and in water treatment for unit process design and
control. TOX can be a direct measure of disinfection by-product formation or, in
some circumstances, a surrogate measure of individual organic pollutants. Individual
circumstances must be carefully judged to determine the usefulness of TOX in these
applications.
17. KCY WOHOG AND DOCUMENT ANALYSIS
i. nrsc*<»pio"S
h.lDtNTiriCn'i/Of'CN fcNDt D TC F*M S
v. COS ATI 1 U'Ui,i»ioup
IB. Ui-llHitouTigNSTAHVlM
RELEASE TO PUBLIC
)», CLAS.J f/tut fxrpvrtf
ITNriASSTFTFn
*1. NO. Or PACES
31.
70 SlCuRlT-Y CLASS
UNCLASSIFIED
22. PRICE
CP* t—m 22JC-1 4-7?| *ftCv*Owtt COlTiOx il OftSOwITt
4
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention cf trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
TOX, IS IT THE NOi'-SFECIFIC PARAMETER OF THE FUTURE?
by
Alan A. Stevens and Ronald C. Dressman
Drinking Water Research Division. KERL, ORD
R. Kent Sorrell and Herbert J. Brass
Technical Support Division
Office of Drinking Water
U.S. Environmental Protection Agency
Cincinnati, OH 45268
ABSTRACT
A total organic halogen (TOX) analysis directly neasures halogen
(CI, Br, I) covalently bound to organic molecules in a sample. Current-
ly, the most popular method for TOX in water involves carbon adsorption,
oxidative combustion, followed by measurement of the forued hydrogen
halide by mi rocoulometry. TOX can be used as an indicator of water
quality and in water treatment for unit process design and control.
TOX can be a direct measure of disinsection by-product formation or.
In some circumstances, a surrogate measure of individual organic pollut-
ants. Individual circumstances must be carefully judged to determine
the usefulness of TOX in these applications.
INTRODUCTION
CARBON ADSORPTION' APPROACH TO TOX
In the early 1970's in an effort to find ways to measure halo-
genated organic compound removal by granular activated carbon (GAC)
filters at water treatment plants along the Rhine River, Kuhn and
Sontheimer developed a technique (1,2) to measure the total organic
.chlorine (TOO.) adsorbed on the GAC at various depths of the filter,
thus defining the adsorption wave front. The halogenated compounds
found in the Rhine (3) were not naturally occurring and were, there-
fore, considered undesirable in a finished drinking water. The
¦location of the adsorption wave front gave Important Information about
the status of the carbon filters' ability to continue to remove these
pollutants. If exhaustion were shown, it signaled the need for reacti-
vation. The procedure developed by Kuhn involved the combustion of a
one gran GAC sample in an atmosphere of O2 and steam, followed by
analysis of the liberated chloride with an ion-specific electrode.
Further investigation by Kuhn et. ai. (4) led to techniques for measure-
ment of the. TOO. content of the water itself.
1
-------
The first sT.ep for water analysis involved the adsorption of th»
crg?.r.ic compounds fron the water onto ground granular activated carbon.
Thii was accomplished by adding one gram of Che activated crfrboa to a
10-L water sample and stirring for one hour. This batch extraction was
repeated a second time after recovering the activated carbon by a
process of flocculation, sedimentation, siphoning off of the water, and
membrane filtration.
The recovered carbon, in the form of a sludge frcn each extrac-
tion, was then pyrohydrolyzed to convert organochlorine compounds to
HQ. In the pyrohydrolysis procedure, superheated steam and oxygen
were passed through two furnaces in series having a common combustion
tube, the first heated from ambient temperature to 700 C and the second
maintained at 1000 C. The HQ. produced during the pyrohydrolysis was
collected as the pyrohydrolyzate by condensing the steam.
The measurement of the Q_ in the pyrohydrolyzate was performed
with a select-ion probe. To account for the interference from inorganic
Q~ adsorbed onto the activated carbon, a duplicate sample was extracted
In the sane way, and the recovered activated carbon sludge was washed in
a solution of sodium nitrate for a minimum of six hours. In this process,
N0j~ -displaced inorganic Q~ from the carbon, and the Q~ in the wash
water was then measured with the select-ion probe. The calculation of
Q~ contributed by organochlorine compounds was then made by subtracting
the amount of Q~ measured in the nitrate washed sample from the amount
of Q~ measured in the pyrohydrolyzate. This value was termed TOQ.
As a result of an evaluation (5) by the Drinking Water Research
Division (DWRD) of the Municipal Environmental Research Laboratory
(MERL), USEPA, Cincinnati, Ohio, several modifications were made to
the method of Kuhn and Sontheimer. Microcoulometry was selected to
measure the Q~ because of a positive bias associated with the use of
the select-icn probe that was recognized to have pro'.uced values up to
10 times the true value in some surface water samples. The selection
of the microcoulometric detector also increased sensitivity by 100-
fold thus reducing the necessary sample size to 100 mL or less. The
measurement was redefined as carbon adsorbable organohalides reported
as Q~ (CA0X as Q~). This definition recognizes that only the carbon
adsorbable fraction of TOQ (90-95% as inferred from measured TOC
adsorption) is being accounted for, and that Br- and I~ when present
were also being titrated. Next, inorganic X- was excluded from the
adsorption process by the addition of N0j~ to the sample; the need for
a duplicate sample to correct for inorganic Q~ interference was thus
eliminated, and with it the 6-hour delay in obtaining results. Finally,
the X~ contributed by organohalides was determined directly on the
activated carbon sludge from samples pretreated to exclude the adsorp-
tion of inorganic X- onto the carbon, thus providing for increased
reproducibility and ouch greater accuracy.
The evaluation of the Kuhn batch method led to the development of
a mini-column method (6) for adsorbing organhohalldes onto activated
carbon, the evaluation of which by DWRD led to modifications by
Dohraann that represented the state-of-the-art for analysis of OX in
water by oxidative combustion of the carbon. This measurement is
generally referred to as Total Organic Halide (TOX) and is the basis
of EPA Method 450.1, Total Organic Halogen, available from the Environ-
mental Monltorine and Support Laboratory, Qncinnati, Ohio 45268.
2
-------
In addition to Dohtmann, Mitsubishi Chemical Industries Limited,
through the COSA Instrument Corporation, is also marketing an instrument
systca consisting of a TOX analytical module and an adsorption module
based on Method 450.1.
Instrumentation that employs the carbon adsorption concept follow-
ed by reductive pyrolysis and conductivity detection of the halide has
has not proven to be reliable. In the authors' hands, problems included
poor precision and accuracy, insufficient inorganic chloride rejection,
and high variability of system blanks.
A small-batch (100 raL) extraction method using activated carbon
has also been developed (7) that is a hybrid of the modified Kuhn
method (5) and the mini-column method for water samples (6). The
excraction is scaled down to trco.t a 100-tnL sample. Activated carbon
recovery is by membrane filtration without the need for flocculation
and sedimentation. Pyrolysis and detection are the same as for the
mini-column method. The results obtained by this method are essen-
tially the same as those that would be obtained by the nini-coluain
method.
OTHER APPROACHES TO OX
Another method for OX is the mini-column method that employs XAD
resins as the adsorbent (8). It involves solvent desorption of OX from
the resin, controlled combustion of a portion of the solvent sample and
detection by microcoulometry. This method is not preferred because of
unfavorable adsorption characteristics of some compounds with the
resin.
Flash vacuun-svaporation is an OX method under study (9). It
involves first a desalting step via steam generation and recondensation
followed by the catalytic conversion of organohalides to Q-2. The Q-2
is then measured by a fluorometric derivatization analysis using
syringaldazine.
Neutron activation (NA) followed by gamma ray spectroscopy has
been used for the analysis of organohalogens. This technique was first
reported used in Sweden (10). The method has been receiving increasing
attention and is an excellent research tool. The organic halide can be
isolated' by either solvent extraction or carbon adsorption. Whether a
solvent extraction is involved, which can account only for the lipo-
philic fraction of OX, or adsorption onto activated carbon, is involved,
NA can differentiate between the halogens that comprise TOX. It does
not, however, fulfill the need for a routine analytical tool for use in
water laboratories throughout the country.
"Extractable organic chlorine" (EOCL) was suggested as a useful
indicator for chemical pollution in surface water (11). It was deter-
mined by extracting with petroleum ether and measuring the "total"
organohalides in the extract by microcoulometry.
3
-------
The ion chroraatograph offers another possibility for the detec-
tion ortd opisurement of individual halides such as vouli be recovered
In a pyrohydrolyzate and reported as OX (5). Interferences that must
be overcome occur as a result of the sulfur associated with carbon that
Is pyrolyzcd.
Organic compounds in aqueous solutions can be separated on the
basis of molecular weight using ultrafiltration membranes. These
separations are not absolute, because some organic matter can be re-
tained due to electrical charges on the organics. McCahill, Conroy
and Mater (12) have used this concept to separate OX according to
molecular weight. Following separation, the solutions are irradiated
with UV radiation to cleave the C-X bond. The Q~ is determined by
select-ion electrode, whereas ion chromatography is used to determine
the a- and Br" simultaneously. The photolysis required 1.5 hours,
followed by a typical ion chromatographic analysis tioe of 0.3 to 0.5
hours. No detection limits were given, however, the Ion chromatoeraph
can detect 10-"? to 10-oM halogen with a 1 mL injection.
PURCEABLt- UHCANIC HAL1DE (POX) AND NON-1'UKCEAELE ORGANIC HALIDE (NPOZ)
Analysis for POX is conducted directly by purging a 10-raL aliquot
of the water sample directly into the pyrolysis furnace and titrating
the HX in the effluent gas during the purge. Commercially available
instruments have a built in capability for doing this. Instrument
instruction manuals detail the use of this feature. POX can also be
estimated by converting THM or other volatile organics data to units of
Q~ ion measurement (13).
Experience in the DWRD laboratory has shown that for most accurate
results separate THM analysis for purgeable compounds and NPOX should
be performed on disinfected water having an OX concentration of 150
ug/L or higher. This is desirable to avoid loss of some THMs that
would occur during a TOX analysis. When it is necessary to determine
POX and NPOX individually and directly, separate aliquots of the sample
should be taken for each analysis. The aliquot taken for NPOX analysis
can first be purged free of trihalomethanes. The NPOX aliquot can then
be analyzed by the carbon adsorption method for TOX. The results for
POX and NPOX analysis can then be combined to give a measure of TOX in
the sample.
Generally, whenever either POX or NPOX is determined analytically,
the other can be deteroined by subtraction from TOX. The failure of
TOX measurement to account completely for THMs, however, means that
NPOX determined by subtracting converted gas chromatographic results is
biased low 10 to 20 percent (14).
MEASUREMENT OF DISINFECTION BYPRODUCTS
Disinfectants react with organic materials in drinking water
sources and thus form byproducts of largely unknown .composition. That
these reactions occur has been long recognized in the form of disin-
4
-------
fectant "demand" and natural organic color bleaching by chlorine,
chlorine dioxide and ozone. Most of the demand is oxidative; that is,
susceptible sites on the organic molecules are oxidized to new func-
tionalities. In the process, the disinfectant is chemically reduced.
On the order of 90X of chlorine demand, for example, can ultimately be
accounted for as chloride ion. The remaining 10Z or so may react by
substitution or addition reactions and become incorporated into the
organic molecular structure. Where brc.nlde ion is present, oxidation
to active bromine species also occurs, and bromine will also be found
in the organic byproduct mixture (15).
Some specific compounds that result from this type of reaction
have been identified. These include the trihalomethanes that are now
regulated. Some others are dihaloacetonitriles, trichloroacetic acid,
dichloroacetlc acid, and 1,1,1-trichloroacetone. The above compounds
can be Identified and measured by gas chromatographic techniques.
These and related compounds that are detectable by GC techniques do
not, however, account for all of the organic halogen byproducts. Some
of these other materials are high molecular weight nor.-volatile species
that are not amendable to GC or other available analytical techniques
capable of detecting and measuring specific compounds. Thus, the TOX,
especially the NPOX parameter, becomes a useful non-specific measure
of the degree of organic attachment of halogen (Q, Br, I) in organic
molecules as a result of the disinfection procers and, therefore, a
useful tool in unit process design and control.
Based on recent findings (16,17) that organic halogen formation
is best avoided In drinking water treatment, the non-specific measure-
ment technique can be used as a group parameter to measure treatment
improvement or deterioration. This is illustrated by example in
Figures (1) and (2). The data used co develop these figures were
generated during studies in the DWRD Laboratory (18).
Figure 1 displays the results of free chlorination of 5 mg/L
Aldrich humic acid solution at three temperatures, three different pH
values over three different reaction periods and two chlorine concentra-
tions. POX, determined by a THM measurement (essentially all chloro-
form) is displayed below the abscissa, and NPOX above. TOX is repre-
sented by the total bar. Bromide was not involved in this experimental
oatrlx. The units are in ug/L as CL"".
In general, THMs are shown in Figure 1 to have increased with
reaction pH, while NPOX decreased dramatically. At low pH, a tempera-
ture increase caused an increase in NPOX that did not occur at neutral
and high pH values. Chloroform yield increased at all pH values
with increasing temperature. NPOX concentrations approached their
final values faster at high pH than at low pH. This effect was not
observed for chloroform. Chlorine dose had a larger effect on NPOX
concentrations than on chloroform. All of these results of chlorina-
tion are typical of those reported for humic materials obtained from
other sources (19).
-------
Figure 2 compares the NPOY and trihalomethane formation at 20C
at 3 different pH levels for each of the three disinfectants: chlorine,
chloramines, and chloride dioxide. NPOX formation was reduced by 85%
when comparing the use of chloramine to chlorine as the disinfectant,
and is still much lower when chlorine dioxide was ur.ed. Trihaloaethane
formation was reduced by greater than 95% when chloramir.es were used,
and no trihalomethanes were detected when chlorine dioxide was used.
Unlike with chlorine, which increases the formation of trihaloraethanes
with increasing pH, the use of chloramines and chlorine dioxide as
disinfectants resulted in a decrease in the formation of all organic
halides with increasing pH. NPOX formation with chlorine dioxide was
rapid, with no difference observed between 2 hr and 144 hr sample
concentrations.
These figures illustrate the variation in organohalide species
¦ formed under different conditions, giving clues of how to minimize OX
' formation during treatment. One must keep in mind, however, th.it this
group parameter is measuring only one specific attribute of the
sample's organic matrix; i.e., halogen content. The beginning assump-
tion was that this was the most important byproduct characteristic of
concern. TOX is not a total measure of disinfectant byproduct and may
not even be proportional to It. This is clear from the data presented
In figure 2 when one considers that the oxidant demands were similiar
for chlorine and chlorine dioxide. Oxidant demand may indeed be a
better non-specific parameter for disinfection byproducts If non-halo-
genated byproduct species are important. Thus, the usefulness of the
TOX parameter as a measure of disinfection byproduct is limited by a
lack of full knowledge of its significance.
USES AS A SURROGATE FOR SPECIFIC CHEMICALS
TOX has been Incorporated into federal regulations for monitoring
ground waters near hazardous waste sites (20). Its use as a surrogate
has also been suggested for monitoring individual halogen containing
VOCs in waters intended for drinking (21). This section will focus on
TOX and POX as a measure of the presence of halo-VOCs in such waters.
For proper Interpretation of TOX or POX results, it is necessary
to understand the ability of these analytical techniques to reflect the
concentration of specific halogen containing chemicals; i.e. the appro-
priate mass balances. Few investigations have been published correlat-
ing TOX and POX analyses with individual VOCs. Early data of TOX
accuracy for volatile organ.ohalides ranged from 73% to 110%, Table 1
(22,23). They included such compounds as chloroform, brouoform and
bromobenzene in reagent wator at concentrations of 98 to 443 ug/L.
The POX data in Table II from the same authors (14,24) indicated 98%
recovery for bromoform and 80% recovery for THMs. It should be pointed
out that the 80% recovery for THMs was obtained with a vitrified insert
tube in the measurement system, which was demonstrated to have caused a
reduced recovery. A more nearly complete recovery of chloroform was
obtained when the tube was replacad.
Additional POX recovery data were obtained for a wider variety of
volatiles at.higher concentrations by Riggin et a,., Table III (25).
These recoveries were also from reagent waters and dosed at 1000 ug/L.
' • - 6 -
-------
They reported compound recoveries to be from 47 to 106*, which were
from 112 lower (chloroform) to 513! lower (bromodicbloromethane) than
those previously reported (22,23).
At lower concentrations of the spiked VOCs, variable recoveries
were observed by Riggin et al, Table IV (25). For 12.5 ug of chloro-
form per liter, a complete recovery was found. The other four com-
pounds all afforded lower recoveries, with bromoforn not being detected
at a 30-ug/L concentration.
More recent data f->r VOC recoveries, Table V, were ob'iaineo. using
VOC-free ground water (26). This ground water.was spiked, head-space
free, with the volatile organics of interest. Each compound was
studied at three concentrations, 10, 30 and 109 ug/L. The samples were
analyzed in triplicate by both TOX and POX. Purge-and-trap/gas chroraa-
tography/eleotrolytic conductivity detector (P&T/GC/E1 CD) analyses were
also performed to ensure the concentrations in the sample bottles were
close to their calculated values. Analysis by P&T/GC/E1CD also ensured
the integrity of the spiked samples by revealing the presence, if any,
of other VOCs.
The compounds at 10 ug/L, which is near to quantitation limit of
the procedures, gave a recovery range of 79-140% for TOX and 25-88% for
POX. The recoveries of the chlorinated compounds for the 30 and 100-
ug/L concentration in Table V ranged from 55-802 recovery. There does
not appear to be a clear relationship between concentration and
recovery. However, there may be a more efficient recovery by TOX than
for POX at the concentrations tested.
Results for an interlaboratory study (*able VI) provided similar
findings, with TOX recoveries ranging from 73-91% and POX recoveries
from 56-77% for the three volatile compounds (27). These data also
seem to demonstrate slight recovery bias in favor of TOX. The results
obtained by the two laboratories generally agreed witnin 10% of each
other.
TABLE I
REPORTED ACCURACY Or TOX FOR VOCs IN REAGENT WATER
Average
Model
Cone.
Recovery
Compounds
ug/L
per cent
Reference
Chloroform
9f
89
(22)
11U
94
(23)
Bromodichloronethane
160
98
(22)
Dibromochlorooethane
155
86
(22)
374
73
(22)
Bromoform
160
110
(22)
238
100
(23)
Bromobenzene
443
95
(23)
7
-------
TABLE II
REPORTED ACCURACY OF FOX FOR VOCs IN REAGENT WATER
Model
Compounds
Cone.
ug/L
Average
Recovery
per cent
Reference
Iroitoforra
TOHs
Chloroform
100
140
98
80a
100
(24 j
(14)
(14)
aVitrlfled combustion tube insert
TABLK III
RECOVERY FOR VARIOUS rURGEABLE ORGANIC HALIDE (POX) COMPOUNDS SPIKED
AT 1000 ug/L INTO REAGENT WATER - DOHRMANN DX-20 SYSTEM (25)
Compound
Methylene chloride
Gilorofora
Trans 1,2-Dichloroethylene
1,1-Dichloroethane
1.1-Dichloroethylene
1,1,2,2-TetraehIoroethane
Tetrachloroethylene
Carbon Tetrachloride
1,1,2-Trichloroethans
1.2-Dichloropropane
Tr ichlorofluoronethane
Trichloroethylene
1,1,1-Trichloroethane
1.2-Dichlor;oe thane
1.3-Dichloropropene
Chlorobenzene
1.2-Dichlorobenzene
1.3-Dichlorobenzene
1.4-Dichlorobenzene
Br oinod i chlo rone t hane
Bromoform
Recovery
per cent
87 (4)a
81 (3)
106 (0.2)
84 (3)
78 (2)
88 (6)
86 (0.6)
86 (1)
76 (8)
76 (2)
79 (7)
76 (2)
68 (10)
70 (4)
60 (4)
48 (7)
65 (5)
51 (7)
61 (7)
47 (9)
5* (3)
aRelative standard deviation (%} in parentheses
8
-------
TABLE IV
AVERAGE RECOVERY FROM 7 POX ANALYSES OF SELECTED COMPOUNDS (25)
Amount of POX
Spike Level
Conpound *ig/L
Spike Level
ug/L as CL~
Found
ug/L as Cl~
Recovery
per cent
Chlorofona
12.5
11
11
(1.4)*
100
Trichloroethene
14
10
6
(0.7)
60
Tetrachloroethene
14
10
5
(1.0)
50
Chlorobenzene
25
8
3
(0.6)
38
Broaoforn*
30
13
ND
<10
ND - Hot Detected.
aRecovery for bromofora at 300 ug/L (130 ug/L as a-) was 487. and 49%
for duplicate analyses.
^Standard deviation in parenthesis.
TABLE V
AVERAGE ACCURACY OF TOX, POX AND PURGE-AND-TRAP
FOR DOSED VOC GROUND WATER SAMPLES3 (26)
TOX POX P&T Confirmation
Calc. Cone. Recovery Recovery Analysis, per cent
ug/L per rent per cent of dosed quantity
Compound
Broaofortn
Chloroform
t-Dichloroethylene
Tetrachloroethylene
10.0
140b
30.5
93
100
120b
10.0
79b
29.9
. 76
100
81
10.1
84b
30.1
63
98.4
60
9.83
79b
30.2
75
101
78
62b
120
66
100
91
110
25b
94
43b
110
76
96
53
97
55
90
59
99
88b
91
70b
100
70
110
aBased on triplicate analyses for TOX, POX, duplicate for P&T.
^Precision was >10%; the ranges of these percent restive standard
deviations were; 15Z-61Z at.10 ug/L: 22Z--25Z at 30 tig/T..; 27% at
100 ug/L.
9
-------
TABLE VI
INTERLABORATORY VOC RECOVERY STUDY (27)
Recovery3
per cent
Compound Spike
ug/L
Ti>Dc
TOX
DWRDC
TSD
POX
DWRD
CHCI3 (50)
CHQ3 (150)
CHBr3 (150)
t-DCEb (150)
85 (1.3) 87 (14.)
81 (3.7) 90 (4.9)
76 (0.1) 91 (17.)
77 (1.0) 73 (9.7)
73 (3.1) 61 (8.2)
60 (1.5) 58 (11.)
70 (1.6) 66 (4.2)
77 (5.1) 56 (6.2)
aRelative standard deviation for triplicate measurements are given in
parentheses.
^t-DCE is trans-Pichloroethylene.
CTSD and DVJRD are Technical Support Division and Drinking Water
Research Division Laboratories respectively.
The VOC data presented thus far has predominately been from spiked
reagent water or a single source of ground water. The ability of TOX
and POX methods to measure the presence of VOCs in a variety of ground
waters, and the methods' ruggedness, have not been fully investigated.
Some information that has been gathered is shown in Table VII (26).
All the cities listed in the table, utilize ground water sources. The
TOX concentrations of these raw waters ranged from <5 ug/L as CI- to
300 ug/L as CI-. The POX results ran from <2 ug/L as Cl~ to 300 ug/L
as Q~. VT/GC/E1CD was used to compare the TOX and FOX concentrations
with concentrations reported as chloride for individual VOCs. This
compirison indicated that POX gave a positive result only when VOCs
were present. TOX analysis on the other hand afforded positive results
in all but one of the 11 samples. Five of these positive TOX samples
contained no detectable VOCs.
Information collected by a USEPA survey, Table VIII (28), also
indicated that a non-VOC background could interfere with the use of TOX
as a surrogate for low concentrations of VOCs. The survey found an
average TOX of 19 ug/L as Q~ in raw ground water, ranging from <5 ug/L
as CL~ to 85 ug/L as CL~. A high TOX, 85 ug/L as OL~, did not indicate
a high VOC concentration, as less than 1% could be accounted for by
halogenated VOC in this saaple. Surface waters, also seem to possess a
variable TOX background, ranging from <5 to 49 ug/L as Q-. In both
ground and surface water, the TOX analysis would have to detect VOC
concentrations above the background interference, whose nature and
variability are yet undefined.
10
-------
SAMPLE STABILITY
Recent Information has. shown Ehat, after chlorlnation, certain of
the constituents comprising TOX are not stable when a t.'ater sample is
stored. This aspect of TOX loss during sample storage was first re-
vealed in a paper presented at an American Chemical Society meeting in
1981 (29). The paper dealt with storage of chlorinated drinking waters
The first study, using samples from a surface water source stored at
20*C, revealed a 30% decrease in -OX concentration 15 days after addi-
tion of sulfite to reduce the chlorine residual (Figure 3). A 502
increase of TOX was observed when the chlorine residual was not reduced
because of the dominance of continued TOX formation. The storage at
subanblent temperatures (5 C) of sulfite-reduced drinking water frcas
the same source, demonstrated a slower rate of decline and a smaller
percentage of loss of only 10%. Analysis of THKs, the main component
of the purgeable fraction revealed no significant change in concentra-
tion, therefore, the decline was from the loss of nonpurgeable organo-
halides (NPOX). These losses are suggested to be due to decomposition
of oetastable organohalides formed during the chlorination process
(30). Similar behavior has been observed in chlorinated ground water
and tertiary effluents, with a 20% loss of TOX after reduction of
the chlorine residual (31).
The storage of sulfite-reduced samples at subambient temperatures
is not, however, a guarantee of low NPOX losses. Data for other
surface supplied drinking waters stored under identical conditions shou
significant loss with time (Table IX) (30).
A suitable approach to solve this problem rray be to acidif} the
sample and store It at reduced temperatures. In an experiment where
one set of samples was acidified with nitric acid to pH 2, the TOX
declined by only 15% after three months at r-oom temperature (Table X)
(30). A second set, whose pH was allowed to remain at 7.9, had a 51%
loss during the same period. When acidification was used in combina-
tion with refrigeration, better preservation was afforded (Table XI).
The acidified samples had no loss of TOX by day four and only a 6%
loss by day 48. The unacidified samples
-------
TABLE VII
COMPARISON OF
TOX, POX AND
PURGE-AND-
TRAP V0C
CONCENTRATIONS
(ug/L as Q~) FOR SE
LECTED GROUND WATERS3 (26)
City 0
TOX
rox
VOCs via P&T
66951
21 (21)
<2
ND
66989
10 (56)
<2
0.3 (4)
66952
18 (46)
<2
0.6b
66990
27 (10)
21
(5)
20 (7)
66954
30 (15)
<2
ND
66991
20 (12)
<2
ND
66993
25 (27)
<2
ND
66998
9 (50)
<2
ND
66999
10 (5)
2
(50)
6 (6)
67000
<5
<2
ND
Rivb
300 (2)
300
(2)
330b
aEased on Triplicate Analyses for TOX and POX; Duplicate for P&T;
the percent relative standard deviation (or per cent difference
for P&T) are given in parenthesis.
''Denotes single analysis only
ND «• Not Detected
TABLE VIII
TOX CONCENTRATION OF GROUND AND SURFACE WATER
CUSS RESAMPLt SURVEY (28) ,
No. of Sites Water Type TOX
ug/L as Q~
12 Ground <5-85
(avg «• 19)
6 Surface <5-49
(avg - 2C)
IN THE
voc
per cent3
<1-39
aDefined as the percentage of TOX accounted for by VOCs.
12
-------
TABLE IX
TOX DECLINE FROM SULFITE REDUCED DRINKING WATERS
(SURFACE WATER SOURCES) WITH SAMPLES STORED AT 5°C. (30)
Sanple
£
Date
Saopled
Initial Gone.3
ug/L as Q~
Storage
days
Loss
per cent
001-06
May 81
110
(3%)
191
39
006-03
Jan 81
430
(6Z)
213
43
006-04
Feb 81
410
(2%)
183
41
006-05
Mar 81
520
(2%)
196
42
006-06
Apr 81
370
(1Z)
165
39
008-01
Dec 80
120
(1Z)
254
44
008-07
Jul 81
140
(15%)
30
28
017-02
May 81
120
(2%)
176
38
XWS
Jul 80
250
(2%)
21
18
£Based on duplicate analysis with precision given as
A-B
- % Diff x 100
A+B
2
13
-------
TABLE X
COMPARISON OF TOX DECLINE FOR NITRIC-ACIDIFIED AND NON-ACIDIFIED TA!
WATERS (SURFACE WATER SOURCE) STORED AT AMBIENT TEMPERATURES (30)
pH 7.9 pH 2.0
Day (bnc.a Loss Cbnc.a Loss
ug/L as Q~ per cent ug/L as CL~ per cent
0 160 (4.8%) - 160 (4.8%)
10 120 (2.8%) 25 140 (5.52) 8
89 76b 51 130b 15
aBased on duplicate analysis with precision given as
A-B
X Diff x 100
A+B
2
^Single analysis
TABLE XI
COMPARISON OF TOX DECLINE FOR NITRIC-ACTDIFIED AND NON-ACIDIFIED
TAP WATER (SURFACE WATER SOURCE) STORED AT 5°C. (30)
pH 7.9 pH 2.0
Day (bnc.a Loss Cbnc.a Loss
ug/L as CL~ per cent ug/L as CL~ per cent
0 140 (2.0%) - 140 (2.0%)
4 110 (2.3%) 17 140 (9.7%)
48 94 (10%) 32 130 (3.2) 6
Based on duplicate analysis with precision given as
A-B
% Diff 100
A+B
— "
-------
AVAILABILITY AND COST
Due to the Resource Conservation and Recovery Act (RCRA) require-
ments (33) and general interest in TOX as a surrogate, the use of TOX
has become nore. widespread. .At present, it is estimated there
between 50 and 75 commercial laboratories.and univsvsiLid f.baf: perform
TOX analysis using the Dohrnann DX-20 system (34).
A survey was conducted by the Technical Support Division (TSD) in
order to determine the capabilities of these laboratories. Thirty-
three laboratories were contacted. This is about half of the labora-
tories that perform TOX analyses using the DX-20 system. Of these 33,
only 18 performed POX analyses on a routine basis. Based on this
survey, it is estimated that approximately half-of all laboratories
performing TOX analyses analyze for POX.
These same laboratories were asked to provide price quotes for the
measurement of TOX, POX and trihalomethanes (THMs) by purge-and-trap
and liquid-liquid extraction techniques. Also requested were quotes
for the analyses by purge-and-trap of volatile organic chemicals con-
taining halogen (VOCs). These VOC quotes were further identified as to'
the type of detector being used; namely, electrolytic conductivity
(ElCD) or mass spectrometer. The cost information is shown in Table
XII. TOX by Method 450.1 requires duplicate analysis for each sample,
and this was also applied to POX. The prices for THM and VOC analyses
are based on a single measurement because a duplicate is not required
according to their respective methods. No information was obtained on
the qualLty of analytical data generated by these laboratories.
The average cost for TOX and POX analyses were $99 and $77 respec-
tively. The ranges of co3t for sample analysis (Table XII) are great,
varying by approximately an order of magnitude. Considering the
fact that TOX measurements are much more routinely made when compared
to POX, may account for the apparent elevation in the cost of POX
analysis.
B*sed on the limited number of laboratories responding, THM
measurements by purge-and-trap had an average price of $86 per sample
and was more costly than measurement by liquid-liquid extraction, which
had an average price of $59 per sample. The ranges of cost for THM
sample analysis varied approximately five fold for purge-and-trap and
two fold for liquid-liquid extraction. These cost data show that there
is no cost saving in using TOX or POX as surrogate measures of the
presence of THMs. In fact, based on average costs of analysis per
sample, the determination of the four THMs is less expensive than the
cost for total organohalide measurement.
Analysis of halogen containing VOCs had an average price of ($114)
per sample with a range of $65 to $150. Based on information from a
limited number of responding laboratories, analytical cost is more
expensive than for either TCX and POX measurements.
General caveats must be placed on the above cost comparisons. The
basis for contacting laboratories was in their ability to perform TOX
15
-------
TABLE' XIL
COST COMPARISON OF ANALYSES
Analysis
Average Cost
dollars
Cost Range
dollars
Number
of labs
TOX
99
30-200
33
POX
77
15-140
18
THM (P&T/GC/ELCD)a
86
42-200
15
THM (LLE/GC/ECD)b
59
A0-S5
8
VOC (P&T/GC/ELCD)c
ll4 . .r
65-150
5
VOC (P&T/GC/MS)d
207
100-350
18
aTrihalonethane analysis by EPA purge-and-trap methods using an
electrolytic conductivity detector.
^Trihalonethane analysis by EPA liquid-liquid-extraction methods
using an electron capture detector.
cVolatile organic analyses by EPA purge-and-trap methods using an
electrolytic conductivity detector.
^Volatile organic analyses by EPA purge-and-trap methods using a
mass spectrometer detector.
measurements using the. Lohrraann, DX-20 system. THM and VOC cost data
from these laboratories may not accurately reflect analytical costs for
a wide range of laboratories. These costs appear reasonable, however,
based on the direct experience of the authors in dealing with labora-
tories specializing in THM and VOC measurement.
Another point is that while these cost data were valid at the
time the survey was performed (early 1984), little is known about
changes in the cost of sanple analyses in the future. Sudden increases
or decreases in demand for a particular measurement could radically
alter these cost data. Also, while not being actively considered at
this time, lifting the requirement for duplicate analyses could con-
siderably reduce the cost of sample analysis by the OX methods.
SELECTION OF ANALYTICAL APPROACH
Discussion in the previous sections focused on the following
aspects of organohalide as a nonspecific parameter: the measurement
procedures; use as a surrogate for disinfection byproducts; use as a
surrogate for specific organic chemicals, principally VOCs; sample
stability; and cost aad relative ease of analysis. The emphasis in
16
-------
this section is Co relate what is known about organohallae ana specific •
chemical analyses in an attempt to evaluate the utility of each measure-
ment approach for activities such as surveying water quality; compli-
ance monitoring; and unit process design, control and monitoring.
A word about matching methods to objectives is in order here. A
key consideration is to establish the needs of the project, be it
research, surveillance, or compliance monitoring. In some applications
the selection of methodology capable of providing data of fine distinc-
tion and the highest possible precision and accuracy with stringent
quality assurance requirements would be the primary consideration. In
more routine applications, perhaps because of high saople volume, cost
may be the primary consideration, and a method capable of providing
data to a satisfactory but somewhat lesser degee of reliability and
less stringent quality assurance requirements may be chosen. Should
any uncertainty arise in the course of this latter approach, a more
definitive method can be called into use as needed to resolve the
question.
In either case, another primary consideration may be the amount
of information that the method can provide; such as, specific compound
identification and quantification vs the collective measurement of
groups of compounds, a consideration that may be determined by the use
of the data.
Figure 4 is a schematic diagram that links the relationship of
water sample type to classes and subclasses of target organohalides.
In addition, the diagram links the target to the selection of an appro-
priate method based on the type of response characteristic for that
method.
Response characteristics are generally of three types. The first
type is a combination of non-specific and non-exclusive, meaning that
it does not measure specific compounds and can not be made to measure
individual subclasses of compounds to the exclusion of other subclasses
present to which the method responds. For exaaple, POX will collec-
tively measure THMs and halogenated volatiles such as TCE, whether ona
is interested in both subclasses or not.
The second type is a combination of specific and exclusive, mean-
ing that it measures individual compounds to the exclusion of compounds
present and belonging to other subclasses to which the method responds.
For exaaple, P&T/GC can be made to measure individual THMs to the
exclusion of volatiles such as TCE, simply because of the gas chromato-
graphic character of the method, even though both subclasses nay be
present in the sample. Each method of the second type is specific and
exclusive only to a degree, however, that Is determined by details of
test procedures employed. The fine distinctions among these method
types is beyond the scope of this paper.
The distinction about exclusivity is of special Importance when
using the OX methods as surrogates to estimate the presence of a
defined subclass of organohalides. For example, in groundwater conspli-
17
-------
FINISHED
DRINKING
WATER
SOURCE WATER
(Ground & Surface)
WASTE WATER
DISINFECTION BYPRODUCTS
INDUSTRIAL SYNTHETIC ORGANICS
Non-Volatiles
Serai-Volatlles
Volatlles
Volatlles
Seml-Volatllos
Non-Volatllos
Me t hod
(e.g., halomhumics)(e
•g., haloacetic aclds)(
e.g., TIIMs)
(e.g., TCE)
(e.g., PCB)
(e.g., Dalapon)
TOX
HSNE
NSNE
NSNE
NSNE
NSNE
NStlE
POX
NA
NA
NSNE
NSNE
NA
MA
NPOX
NSNE
NSNE
NA
NA
NSNE
NSNE
P&T/GC
NA
NA
SE
SE
NA
NA
Ll.F./CC
NA
SE
SE
SE
SE
NA
OClkje.g. i
•
HPLC
Undefined
SE
NA
NA '<
SE
SE
D/CC
NA
SE
NA
HA
SE
SE
Method Response
FICURE 4. SCHEMATIC LINKING THE RELATIONSHIPS OF WATER SAMPLE TYPE TO CUSS
AND SUBCLASS OF TARGET ORGANOHALIDES TO METHOD RESPONSE.
NSNE ¦ non-specific and non-exclusive - docs not measure Individual compounds and cannot be made to \uoasure individual
subclasses to the exclusion of others present to which the method responds.
NA » not applicable - does not measure the Indicated subclass or Individuals therein.
SE " specific and exclusive - measures individual compounds and can be made to measure compounds In individual
subclass to the exclusion of compounds present belonging to other subclasses to which the method responds.
aThe dotted line linking the classes of target organohalides suggests the possibility of both classes of compounds
being present in some samples of all types.
''Other chromatographic techniques where D/GC ¦ Derivltization/Cas Chromatography and HPLC Is High Performance
Liquid Chromatography.
-------
ance iBonitortng,-«nless -o«e'h«'; good #t • ion (to ...believe - perhaps by
having previously characterized the v„cer by specific compound analysis
- that only the subclass 'bf organohalldes of interest are present, one
oust recognize that an OX measurement may be positively biased by the
presence of organohalldes other than those of the subclass of interest.
The third type of response characteristic is that a method is not
.applicable, meaning that it does not measure a particular subclass of
.compounds or any 6uch individual compounds under the subclass.
The target organohalldes defined in Figure 4 are differentiated
based on chemical nature and origin. A class separation has beep made
between chemicals generated by disinfection, and chemicals that
generally are regarded as industrial in origin. In each category there
are volatile, serai—volatile and non—volatile halogen containing
components. Of course, certain chemicals may be of volatility inter-
mediate to these subclasses. While these cases are not directly dealt
with here, they may be Important in certain situations. Table XIII
presents those methods considered by the authors to provide the most
reliable measure of the various subclasses of organohalldes for the
activities discussed.
For disinfection byproducts, the volatile compounds are princi-
pally comprised of trihaloraethanes. Other seml-volatlle chemicals
such as the dihaloacetonltriles and chloroacetones are known to readily
occur at much lower concentrations than THMs. Although reporting on
conditions far more severe with regard to precursor and oxidant concen-
tration than encountered in normal treatment practice, a substantial
portion of this fraction has recently been Identified as being con-
prised of lower molecular weight haloacetic acids (35). The non-
volatile fraction includes halogenated precursor material, much of
which is of high molecular weight.
For the analysis of disinfection byproduct volatiles, purge-and-
trap (P&T) or liquid-liquid extraction (LLE) gas chromatography (CC)
are preferred over a TOX, NPOX or POX measurement because GC analysis
provides significantly more information; namely, the identities and
concentrations of specific compounds. The analysis time and cost are
approximately the same for organohalide and GC procedures because EPA,
EKSL Method 450.1, unlike the other methods, requires duplicate
analyses on every sample.
This requirement for duplicate analyses on every sample was
originally stated in the context of some drinking water research
activities where small differences and difficult to measure OX concen-
trations required greater stringency of quality control. The require-
ment could be lifted for some surveillance or monitoring activities
when fine distinctions are not to be made between measurements. When
fine distinctions need to be made, this requirement for replicate
analysis may be applied to all analytical methods.
For many non-volatile byproducts, no alternative to organohalide
measurement (NPOX) exists because either analysis for specific organics
cannot be performed, or the compounds are either or both unknown and
1 ft
-------
TABLE XIII
MOST RELIABLE MATCH OF METHOD TO SUB-CLASS OF TARGET ORGANOItALIDES
• m m *- m m -m m rr- -i« m -m » m-m m » —¦ — —
-------
sunr-fi.-cv uu»Uj.K»/y«..
difficult to measure. Certain semi-volatiles such as the chloroacetlc
acids, however, are amenable to LLE and derlvlclzatlon/gas chromato-
graphy (D/GC) analysis, which Is listed with other chromatographic
techniques (OCT) In figure 4. OCTs such as column chromatography and
high performance liquid chromatography (HPLC) av? being used to identify
higher molecular weight byproduct fractions. Generally, these OCTs are
being used in research investigations.
, The only way to directly measure non-volatile byproducts (NPOX)
Is to measure the NPOX by the method for TOX on a sample that has been
purged completely free of volatilcs. If a POX measurement Is then
made on a duplicate sample, the sura of the POX and NPOX thus obtained
is equivalent to a TOX measurement (14). Alternatively, one cr.n obtain
an indirect measure of NPOX by directly measuring and subtracting the
POX concentration from the TOX concentration (14).
j One should keep In mind that both TOX and NPOX are measured to an
unknown degree of accuracy because there exists no reference material
;of a humlc/fulvlc acid nature with which to assess recovery efficiency
.by carbon adsorption. The only estimate of accuracy cooes from corapar-
.lng TOC measurements before and after a sample has been put through the
The efficiency, of removal has be«;n determined to be 90 to 95Z
.(22) and is assumed to be the same for TOX.
In assessing contaminants which are industrial in origin, POX and
TOX have been evaluated as surrogate measures of VOCs. In most cases,
specific compound analyses are suggested, again due to enhanced infor-
mation obtained relative to surrogace measurements at comparable cost
and time. The potential does exist, however, for the use of POX as a
surrogate for VOCs. The analytical cost survey reported in Table XII
indicates average costs of $77 and $114 for POX and P&T/GC analyses,
respectively. POX may become the preferred tool with which to screen
for VOCs in certain situations if it is firmly established that dupli-
cate analyses on every sample are not required. However, a negative
but resolvable consideration must first be dealt with; namely,
sufficient data must be acquired upon which to adequately demonstrate
the relationship betwson POX and VOCs in a variety of surface and
ground waters.
For semi-volatile and non-volatile industrial synthetic organo-
halides, TOX, NPOX, LLE/GC, and OCT analyses such as HPLC and D/GC are
available. The relationship between these surrogate and nonsurrogate
parameters has not beer, adequately shown. Some waters contain OX
materials that are not accounted for by chromatographic techniques,
whereas the accuracy of OX by carbon adsorption-is uncertain.
Once again, specific chemicals should be measured if this ca;: be
accomplished at reasonable cost. However, the wide variety of organo-
halides which may exist at a given site, such as a landfills, can make
TOX or NPOX measurement cost effective as a screening tool for halogen
containing organics. If a "high" TOX or NPOX value were obtained
during screening, specific compound measurements would then be
attempted.
21
-------
SUMMARY AND CONCLUDING REMARKS i
We have seen i.hat both surrogate measurements and specific com-
pound determinations can be useful in assessing water quality. In some
cases, specific chemical analyses provide the required information.
However, in other situations, halogen containing chemicals either
cannot be easily measured or cannot be measured at all, tnd TOX or
NPOX can be useful as an indicator of chemical content.
For a survey of possible organohalide contamination, the organo-
halide measure.-lents may be entirely satisfactory for the information
they provide as a prelude to more detailed analysis, or simply as an
early warning indicator.
An important application of organhalide measurement is in unit
process design, control and monitoring. Haste water organohalide
measurement discerns compounds primarily of industrial origin.
Finished drinking water analysis primarily discerns disinfection by-
products.
In tha earlier section dealing with the measurement of disinfec-
tion byproducts, a discussion of THM and NPOX data generated under sets
of carefully controlled "treatment" conditions provides an excellent
example of this unit process application. Information derived from
these types of data can lead to treatment plant design which minimizes
the formation, and thus the concentration of disinfection byproducts in
the treated water. Byproducts can be monitored and to some extent
controlled in the operation of the plant.
One assumes that, from a health standpoint, it is desirable to
minimize these halogtn containing constituents. One must realize,
however, that in minimizing these constituents, an attempt is being
made to control chemicals that, in most cases, cannot be directly
measured. The difficulty in placing a value on this type of measure-
ment is in the uncertainty of whethar specific chemicals of health
concern have in fact been controlled. In addition, non-halogen con-
taining chemicals are not being considered. Controlling organohalide
disinfection byproducts may have little or no relationship to the
control of non-halogen containing compounds that are of health concern.
A dotted line connects disinfection byproducts to industrial
synthetic organics in Figure 4. The dotted line is there to draw
attention to the fact that many water samples have some potential for
containing both classes of compounds. One should be aware that organo-
halide methods, being non-exclusive, have the hidden potential for
biasing any attempted exclusive measure of either class of target
organohalides.
In spite of the shortcomings discussed in this paper, the TOX
method does have a unique advantage. It can demonstrate the absence of
organohalide contamination without the need to resort to a variety of
specific methods to eliminate all the possibilities. The cost savings
of monitoring any situation with one method where two or more might
otherwise be required, should be obvious. If one needs to demonstrate
the absence either of volatile organohalides or non-volatile organo-
halides, POX and NPOX can be used to similar advantage.
77
-------
REFERENCES
1. Kuhn, V., Sontheiraer, H., Several Investigations on Activated
Carbon for the Determination of Organic Chloro-Compounds. Vom
Wasser, 15:65 (197'j).
2. Kuhn, W., Sonthemier, H., Analytic Determination of Chlorinated
Organic Compounds with Temperature-Progra=2fid Pyrohydrolysis.
Von Wasser, 41:1 (1975).
3. Kolle, W., Sontheiraer, H., Stieglitz, L., Cualification Test of
Activated Carbon for Water Works Based on Irs Adsorption Properties
for Organic Chlorine Compounds. Von Wasser, 44:203 (1975).
A. Kuhn, W., Fuchs, F., Sontheimer, H., Z. Vasser Abwasser Forsch.,
10:6:162 (1977).
5. Dressaan, R., McFarren, E., and Synons, J., An Evaluation of the
Determination of Total Organic Chlorine (TOOL) in Water by Adsorp-
tion onto Ground Granular Activated Carbon, Pyrohydrolysis and
Chloride-Ion Measurement. Proc. 5th AVWA VQTTC, Kansas City, MO;
(1977).
6. Takahashi, Y., Analysis Techniques for Organic Carbon and Organic
Halogen. Proc. EPA/NATO-CCMS-Conference oa Adsorption Techniques,
Reseen, VA (1979).
7. Jekel, M., Roberts, P., Total Organic Ralogea Measurements Cor the
Characterization of Reclaimed Waters: Occurrence Formation and
Removal. Environ. Set. & Techr.ol., 14:970 (1980).
8. Glaze, V., Peyton, G., Rawley, R., Total Organic Halogen as Water
Quality Parameters: Adsorption/Microcoulosecric Method. Envir.
Sci. & Technol., 11:685 (1977).
9. Gulf South Research Institute, Total Organic Gilorine (TOQ)
Analysis Instrumentation for Direct Determination in Water Without
Preconcentration. USEPA, EMSL, Cincinnati, C2; Grant No. 805340.
10. Lunde, C., Gether J., Josefsson, B., The Sua of Chlorinated and of'
Brominated, Non-Polar Hydrocarbons in Water. Bull, of Envir.
Contan, and Tox. 13:656 (1975).
11. Wegtsan, R., Greve, P., The Microcoulometric Determination of
Extrsctable Organic Halogen in Surface Water; Application to
Surface "ater; Application to Surface Waters of the Netherlands.
The Set, of the Total Envir., 7:235-245 (1977).
12. McCihlll, M., Cbnroy, L., Maier, W., Oetenxiaation of Organically
Combined Oilorine in High Molecular Weight Aqoatic Organics.
Envir. Sci. & Technol., 14: 201 (1980).
23
-------
13. Drecsraan, R., Procedure for Che Application of the CAOX-as-Q
Method to Disinfected Water. Addendum ibid 5.
14. Dressman, R. and Stevens, A., The Analysis of Organoh.-lides in
Water-an Evaluation Updats. Jour. AWWA, 75:8:431 (Aug. 1983).
15. Bunn, W., Haas, B. Deane E., Kleopfer, R., Formation of Trihalo-
methanes by Chlorination of Surface Water. Envir. Letters, 10:205
(1975).
16. Meier, J., Lingg, R., Buil, R., Formation of Mutagens Following
Chlorination of humlc Acid - A Model for Mutagen Formation During
Drinking Uater Treatment. Mutation Research, 118:25 (1983).
17. Cblenan, E., et. al., GC/MS Analysis of Katagenic Extracts of
Aqueous Chlorinated Humlc Acids - A Comparison of the Byproducts
to Drinking Water Contaminants. Submitted to Envlr. Scl. &
Technol.
18. Stevens, A., Moore, L., Dressman, R., Seeger, D., Disinfectant
Chemistry in Drinking Water - Overview of Impr.cts on Drinking
Water Quality. Presented Before the Division of Environmental
Chemistry, ACS, Washington, DC; (Sept. 1983).
19. Fleischacker, S., Randke, J., Formation of Organic Chlorine in
Public Water Supplies. Jour. AWWA 75:132 (March 1983).
20. USEPA, Standards and Interim Status Standards for Owners and
Operators of Hazardous Waste Treatment, Storage, and Disposal
Facilities. Fed. Re^., 45:98:33240 (May 19, 1980).
21. USEPA, National Revised Primary Drinking Water Regulations,
Volatile Synthetic Organic Chemicals in Drinking Water; Advanced
Notice of Proposed Rulemaking. Fed. Reg., 47:43:9352 (Mar. 4,
1982).
22. Dressman, R., Najar, B., and Redzikowski, R., The Analysis of
Organohalldes (OX) in Water as a Group Parameter. Proc. 7th AWWA
WQTC, Philadelphia, PA; (1979).
23. Takahashi, Y. and Moore, R., Measurement of Total Organic Halides
(T0X> in Water by Carbon Adsorptior./Microcoulometric Determination.
Presented at the 177th National Meeting of the American Chemical
Society, Div. of Envir. Chem., Honolulu, HI; (1979).
24. Takahashi, Y., Moore, R., and Joyce, R., Measurement of Total
Organic Halides (POX) in Water Using Carbon Adsorption and Micro- •
coulometric Determination. Presented at the 179th American
Chemical Society, Div. of Envir. Oem., Houston, TX; (1980).
25. Riggin, R., et.al., Development and Evaluation of Methods for
Total Organic Ilalide and Purgeable Organic Halld*; in Wastewater.
USEPA, EMSL, Cincinnati, OH; EPA-600/54-84-008 (Jan. 1984).
' 24
-------
26. Sorrell, R. K., DaJy, E., Boyer, t., and Brass, Hi J., Monitor-
ing for Volatile Organohalides Using Purgeable and Total Organic
Hallde Analyses as Surrogates. Presented at 5th Conference on
Water Cilorination, Environmental Impact and Health Effect;,
Williamsburg, VA June 3-8, 1984.
27. Sorrell, R., Cbntner, C., and Dressman, R., USEPA, TSD/MERL;
Unpublished Data.
28. Preliminary Report - Community Water Supply Survey-Resaaple.
USEPA, Drinking Water Quality Assessment Branch, TSD, Cincinnati,
OH; (1981).
29. Sorrell, R., Total Organic Halide: Occurrence, Stability and
Process Control in Drinking Waters. Presented at the 181st
National Meeting of the American Chemical Society, Div. of Envir.
Chen., Atlanta, GA; (March 1981).
30. Sorrell, R., Stability of Total Organic Hallde in Environmental
Water Samples. Submitted to Envir. Scl. & Technol.
31. Hopkins, C., Vnviron. Eng. and Sci., Stanford U., Stanford, CA;
Personal Communication.
32. Banovlc, J., Water Research Laboratory, City of Cblumbus, Columbus,
OH; Personal Communication.
33. USEPA, Hazardous Waste Management System, Part VII; Standards and
Interim Status Standards for Owners' and Operators of Hazardous
Waste Treatment, Storage, and Disposal Facilities. Fed. Reg.,
45:98:33239 (1980).
34. Dohnnann Division, Xertex Corporation, Santa Qara, CA; Persoual
Communication.
35. Christian, R., Norwood, D., Millington, D., Johnson, D., Stevens,
A., Identity and Yields of Major Halogenated Products of Aquatic
Fulvic Acid Chlorination. Envir. Sc. & Technol., 17:10:625 (1983).
25
-------
600
500
400
300
200
100
_ 0
100
200
' (°C( 4
pH:
r» *» m
» K -
20
h » i»
w K •-
36
20
r- 9 in
9 r* '-
9 tn in
u» r» ~-
rt Jl
iri r.
36
^ n ifl
ib f* •*
Reaction Time
Scale (Hrs)
20 mg CI. 5 mg HA/L
8 mg CI, 5 mg HA/L
FIG.1 Reaction of Humic Acid at Two Chlorine Doses
26
-------
600
Chlorine
20 ng/L
500
400
Reaction Time
Scale (hrsl:
0
2
72
144
300
x
O
a.
200
100
Chlorammes
22 9 my/1.
Chlorine Dionde
20 7 mg L
Ha
100
5
x
PH
-a — —
u v> w u u ir
200 ¦
FIG. 2 A Comparison of the Formation of NPOX and THMs |CHCI3) at 20°C m Distilled
Water Solutions of 5 mg Humic Acid L Dosed with Various Disinfectants Note
thai NPOX ~ THMs = TOX (the enure bar).
27
-------
TO*
aa
lagCI/L
280-
200-
1#0~i
100-
• 0 —
V.
o-l
I I
a r
-T"
18
-T-
24
*0* REDUCCO
«- soiritt
8 1
"I—
10?
OATS
PIG, 3 Tim* Storage of Cincinnati Tap Water at 20*C
------- |