WATER POLLUTION CONTROL RESEARCH SERIES •16060 GHR 11/71
   INVESTIGATIONS  CONCERNING  PROE ABLE
     IMPACT OF NITRILOTRIACETIC  ACID
            ON GROUND WATER
     ENVIRONMENTAL PROTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development, and demonstration
activities in the Office of Research and Monitoring, Envi-
ronmental Protection Agency, through in-house research and
grants and contracts with Federal, State, and local agencies,
research institutions, and industrial organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D. C. 20460

-------
          INVESTIGATIONS  CONCERNING PROBABLE IMPACT

          OF NITRILOTRIACETIC ACID ON GROUND WATERS
                               by
     William J. Dunlap,  Roger L. Cosby,  James F. McNabb,
             Bert E. Bledsoe, and Marion  R.  Scalf
               ENVIRONMENTAL PROTECTION AGENCY
              Office of Research and Monitoring
             Robert S. Kerr Water Research  Center
            National Ground Water Research  Program
                        Ada, Oklahoma
                      Project #16060 GHR


                        November 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 60 cents

-------
               EPA Notice
The mention of trade names or commercial
products in this report does not constitute
endorsement or recommendation for use.
                    11

-------
                                ABSTRACT
Laboratory studies were employed to investigate the fate and effect of
NTA both in ground waters and in soil profiles overlying ground waters.

Studies of the sorption of NTA by sand, loam, and clay-loam soils indicated
that sorption of NTA on soils could slow its movement into and through
ground waters.  Sorption will probably not be sufficient to prevent or
greatly reduce potential pollution of ground water by NTA used as a
detergent builder.

Soil column studies were employed to investigate the degradation and
effect on metals of NTA infiltrating through soils.  These studies
indicated:  NTA infiltrating through most unsaturated soils likely would
undergo rapid and complete degradation and contribute only inorganic
nitrogen compounds and carbonate to ground waters; NTA infiltrating
through saturated soils would probably experience only very limited
degradation, with a major portion entering ground water intact; any NTA
which escaped degradation during infiltration through soils could
transport such metals as iron, zinc, chromium, lead, cadmium, and
mercury from soils into ground waters.

Studies with model aquifers constructed from natural aquifer sand
indicated that NTA would likely undergo slow degradation in essentially
anaerobic ground-water environments, with production of CO^, Ctfy, and
possibly other organic compounds.
                                   111

-------
                                CONTENTS






Section                                                            Page




   I      Conclusions                                                1




   II     Recommendations                                            3




   III    Introduction                                               5




   IV     Materials and Methods                                      9




   V      Results and Discussion                                    21




   VI     Acknowledgements                                          49




   VII    References                                                51

-------
                                FIGURES


Number                                                             Page

  1       Nitrilotriacetic Acid                                      5

  2       Loam and Clay-Loam Soil Columns                           12

  3       Details of Simulated Aquifer Design                       17

  4       Simulated Aquifers                                        18

  5       NTA Sorption; Freundlich Isotherm at 20°C                 22

               14
  6       NTA,   C, and Metals in Effluent from Unsaturated         24
           Sand Column

               14
  7       NTA,   C, and Metals in Effluent from Unsaturated         25
           Loam Column

               14
  8       NTA,   C, and Metals in Effluent from Unsaturated         26
           Clay-Loam Column

               14
  9       NTA,   C, and Metals in Effluent from Saturated           29
           Sand Column

               14
 10       NTA,   C, and Metals in Effluent from Saturated           30
           Loam Column

               14
 11       NTA,   C, and Metals in Effluent from Saturated           31
           Clay-Loam Column

               14
 12       NTA,   C, and Metals in Effluent from Unsaturated         34
           Sand-Metals Sludge Column

               14
 13       NTA,   C, and Metals in Effluent from Saturated           35
           Sand-Metals Sludge Column
               14
 14       NTA,   C, and Metals in Effluent from Sand-Metal          33
           Ores Column

 15       Organic Carbon in Effluents from Saturated Sand-          40
           Metals Sludge Columns

 16       Visual Comparison of Effluents from Saturated Sand-       41
           Metals Sludge Columns Dosed with Feeds with and
           without NTA

 17       Degradation of NTA in Simulated Aquifers; Initial         43
           Charge and First Recharge
                                    vn

-------
                                FIGURES
                               (Continued)

Number                                                              Page

 18       Degradation of NTA in Simulated Aquifers; Second            44
           and Third Recharge

 19       Detection of Radioactivity in Gases from a Simulated        46
           Aquifer
                                  viii

-------
                                 TABLES


Number                                                              Page

  1       Weight and Volume Data for Soil Columns                    10

  2       R  Values of NTA and Some Postulated NTA                   13
            Degradation Products

  3       Summary of NTA Data for Simulated Aquifer                  48
            Studies

-------
             .   •                 SECTION.I

                                CONCLUSIONS

The  investigations  described  in this  report .permit  the  following
conclusions  concerning  the potential  impact of NTA  on ground water.

1.   Sorption of NTA by  some types of  soils may considerably slow
     the  initial movement of this compound into and  through ground
     waters.  However, sorption  of NTA by soils does not appear to be
     sufficient to prevent or  appreciably decrease the pollution of
     ground waters by this compound over the course  of its long-term
     usage as a detergent builder.

2.   NTA  infiltrating through  unsaturated soils, as  in properly
     operating septic tank percolation fields, will probably undergo
     rapid and complete  degradation.   Only inorganic nitrogen compounds
     and  carbonates  are  likely to enter ground water as the result of
     such degradation.   Introduction of these substances into ground
    waters is not desirable,  but the  quantities introduced as a result
     of NTA use in detergent are not likely to be sufficient to create
     critical pollution  problems.

3.   The degradation of  NTA infiltrating through unsaturated soils
    which contain high  levels of heavy metals, such as those receiving
    wastes from metal plating operations, may be partially inhibited
    as a result of toxicity of the metals to soil microorganisms; hence,
     some NTA may enter  ground water under such circumstances.

4.  NTA infiltrating through  saturated soil systems, as in flooded
    septic tank percolation fields and under some waste water lagoons,
    will likely undergo only  limited  degradation; hence, a significant
    portion will enter underlying ground water, possibly accompanied
    by small quantities of organic degradation products of NTA.

5.  NTA which reaches the water table will likely undergo slow degra-
    dation in the essentially anaerobic ground-water environment.  The
    final products of such degradation will probably include carbon
    dioxide and methane, and possibly other organic compounds of low
    molecular weight and relatively high volatility.  Unknown interme-
    diate products of NTA degradation may also be present in an aquifer
    when NTA is being actively degraded.

6.  NTA that infiltrates through overlying soil into ground water without
    being degraded may solubilize and transport into the ground water
    such metals as iron, zinc, chromium, lead, cadmium, and mercury
    which may be present in the soil.   Situations where NTA might
    infiltrate through soils receiving wastes containing high levels
    of toxic heavy metals,  such as metal plating wastes, would appear
    to entail consderable potential for metals contamination of underlying
    ground waters.  However,  comprehensive evaluation of the potential
    threat to ground-water quality posed by solubilization and transport
    of metals in soils by infiltrating NTA will require further elucidation
    of the factors controlling formation of NTA-metals chelates in soils.

-------
Even in the most favorable circumstances, use of NTA as a detergent builder
is likely to result in some increases in nitrogen and probably carbon in
ground waters.  Phosphates in waste waters do not readily move into or
through ground waters because of strong sorption in soils.  Hence, in
comparison to phosphates the use of NTA in detergents would likely result
in at least a limited adverse impact on ground water.

-------
                                 SECTION II

                              RECOMMENDATIONS

On  the basis  of  this  research,  it is  recommended  that  additional investi-
gations be  conducted  to  further elucidate specific questions concerning
the probable  impact of NTA oh ground  water before unlimited use of this
compound in detergent is begun.

Obviously,  the effect of NTA  which enters ground-waters, used as domestic
water supplies on the health  of consumers of such ground water must be
determined.

The potential for transport of  metals from soils into  ground waters by
NTA- should  be unequivocally established.  This will require thorough
investigation of the  complex  factors  affecting the formation of chelates
by NTA and  metals in soils, including both metals native to the soils and
those deposited to soils by waste disposal activities.  The possibility
of Initial  solubilization of  soil metals by infiltrating NTA, with the
metals remaining in solution  after degradation of the NTA, should also
receive attention.

Should limited use of NTA in  detergents coupled with strict environmental
monitoring  be approved for further evaluation of the acceptability of
this substance,, careful and comprehensive monitoring of NTA and metals
concentrations should be conducted at a number of selected sites where
transport, of metals Into ground water by infiltrating NTA appears most
likely to occur.

The pathway of anaerobic degradation  of NTA in ground-water environments
should be elucidated, with identification of Intermediate products of
degradation and determination of the  extent to which these intermediates
are likely  to be present during such  degradation.  Should the nature and
quantities  of degradation products merit sufficient concern, the mammalian
toxicities  of these compounds should  also be determined.  The extent of
NTA degradation in the presence of relatively large quantities of other
degradable  organic compounds  in essentially anaerobic ground-water
environments should also be the object of further study.

Additional  research should consider the possible occurrence during
successive  infiltration of NTA  through partially saturated and fully
saturated soils of partial, or  arrested aerobic degradation, with con-
commitant appearance in the water infiltrating to ground water of
intermediate products of NTA  degradation, such as iminodiacetic acid.

-------
                              SECTION III

                              INTRODUCTION

Modern synthetic detergents, which commonly contain 40-50 percent condensed
phosphates, are estimated to contribute more than two-thirds of the phospho-
rus load of waste waters in the United States.(1)  Since phosphorus in waste
waters is believed to be a major factor in the cultural eutrophication problem
now plaguing many of our surface waters, there is currently a considerable
expenditure of effort and resources by both industry and government directed
toward the reduction or elimination of the phosphate content of detergents.
However, the condensed phosphates  (known as builders) used in detergent
formulation contribute significantly to detergent performance by complexing
calcium, magnesium, and other metallic ions in wash waters which otherwise
would attenuate or destroy  the surfactant power of the detergent.  Therefore,
in order to maintain detergent performance at anywhere near presently
accepted levels and yet achieve the goal of reduction or elimination of
phosphates in detergents, it is likely that an environmentally acceptable
replacement for the condensed phosphate detergent builders must be developed.
The leading candidate thus far proposed for this purpose is nitrilotriacetic
acid  (Figure 1), a member of a group of amino-carboxylic acids long known
as exceptionally efficient  chelating agents.
                             0
                             II
                         HO-C
                                  ^        -*
                                    N-CH--C.

                               ,0**'
                          o = c
                             I
                             OH
                     FIGURE 1.   NITRILOTRIACETIC ACID

-------
Nitrilotriacetic acid (hereafter called NTA) is an organic compound containing
no phosphorus; hence, its use in detergents as a replacement for a significant
portion of the condensed phosphate builders would be expected to proportionately
decrease the aquatic phosphorus problem.  By virtue of its ability as a chelating
agent, NTA fulfills the major functional requirement for a detergent builder
since, in a properly buffered detergent solution, it effectively complexes
the troublesome metallic ions capable of adversely affecting detergent perform-
ance.  Also, a process is available which permits the production of NTA in
highly pure form (as the sodium salt) at a cost and in quantities required
for the large scale use of this compound in detergents.

There are, however, considerations other than the phosphorus problem regarding
the environmental acceptability of NTA.  Although microorganisms found in sewage
treatment systems have been shown capable, under proper conditions, of degrading
NTA to end products of water, carbon dioxide, and nitrate,(2>3,4) the extent
and pathways of NTA degradation in various aquatic environments, the toxicity
of the parent compound and possible organic products of its degradation, and
the effect of NTA on the solubility and transport of metals in aqueous
environments have not been clearly defined.  Hence, valid questions exist
concerning the impact on water resources of large scale employment of NTA
in detergents, both from the standpoint of ecological effects and human
health.  The magnitude of the potential input of NTA to waste waters, which
would amount to over a billion pounds annually were NTA to replace only half
of the condensed phosphates currently utilized in detergents in the United
States, requires that these questions be carefully evaluated before final
designation of NTA as a suitable replacement for the phosphate builders in
detergents.

Among the major questions regarding the environmental acceptability of NTA
as a detergent builder are those concerning the potential effects of NTA on
ground-water resources.  Ground waters were estimated to provide approximately
23 percent of all fresh water and about 33 percent of public water supplies
in the United States in 1965.(5)  Also, in many areas ground waters constitute
valuable reserves of fresh water which are now used only in emergencies or
are as yet relatively untapped.  Protection of these valuable resources from
pollution is particularly important because pollution effects almost invariably
persist much longer and are more difficult to eradicate in ground waters than
in surface waters.

In considering the potential effects of NTA on ground-water resources, it is
necessary first to examine the pathways by which NTA might enter ground water.
These pathways are restricted essentially to two possibilities:  infiltration
through the soil mantle above the water table and discharge directly into the
ground water.  The first pathway would be operative in cases where waters
containing NTA are discharged to permeable soils appreciably above the water
table, such as in properly designed and operating septic systems, waste
lagoons, and related sewage spreading systems.  The second pathway would be
followed in the all too frequent situations where septic systems, lagoons,
and similar waste disposal systems are connected directly to ground waters,
and also in cases of direct artificial recharge of aquifers.

-------
NTA infiltrating through the soil mantle above the water table would be
subject to possible sorption on soil particles, degradation, and reaction
with metal ions to form soluble metal chelates.  NTA reaching the water
table, whether by direct introduction to ground water or after infiltrating
through the overlying soil, would be subject to analogous phenomena within
the ground-water environment, where conditions of low oxygen tension
ordinarily prevail.

These environmental possibilities necessitate that the following specific
questions be  answered in order to assess the probable impact on ground
water of large scale use of NTA in detergents.

1.  How will  sorption affect the infiltration of NTA through soil
    profiles  and its movement within ground-water aquifers?

2.  To what extent will NTA be degraded during infiltration through
    unsaturated and saturated soils and within essentially anaerobic
    ground-water environments ?

3.  If NTA is degraded during infiltration through soil profiles or
    within ground-water environments, what will be the nature and fate
    of the degradation products?

4.  To what extent will NTA affect the transport of metals into and
    through ground water?

This report describes a research project undertaken in an effort to provide
answers to these questions.  Time limitations dictated that the project be
relatively short term, and research was therefore limited to a rather
intensive laboratory investigation consisting of three phases which were
conducted more or less concurrently.

The first phase of the investigation was concerned with the effect of
sorption on the infiltration of NTA through the soil mantle and its movement
through ground water.  This involved determination of sorption isotherms for
NTA on typical sand, loam, and clay-loam soils.

In the second phase of the research effort, soil columns were employed for
studies pertaining to the fate of NTA infiltrating through the soil mantle
and effects of NTA on metals transport in ground water.  Initially, columns
containing typical sand, loam, and clay-loam soils were utilized to investi-
gate the possible degradation of NTA during infiltration through both
unsaturated and saturated soils and the effects of NTA on the solubility
and transport of metals native to the soils.  Subsequently, columns containing
sand enriched with heavy metals were utilized for the principal purpose of
examining further the question of possible transport of metals into ground
water by NTA.

The third phase of the project involved the use of simulated aquifers
constructed in the laboratory from natural aquifer sand for investigation
of the possible degradation of NTA in an essentially anaerobic ground-
water environment.

-------
                               SECTION IV

                         MATERIALS AND METHODS

                            Sorption Studies

Sorption isotherms were determined for NTA on a sand, a loam, and a clay-
loam soil in order to broadly span the various soil types which might be
encountered in natural situations.  The soils used, all of which were
obtained in the vicinity of Ada, Oklahoma, were identified by United
States Department of Agriculture Soil Conservation Service soil surveys
as Konawa loamy fine sand, Claremore loam, and Burleson clay loam.  Soils
were air dried and screened to remove pebbles, seeds, and similar extraneous
matter before use.

Solutions of the trisodium salt of NTA (NTA Na-j) uniformly labeled with
carbon- 14 were utilized in sorption studies as well as the other phases
of this investigation.  These solutions were prepared from l^C-NTA,
uniformly labeled, specific activity 2.5 mCi/mmole (New England Nuclear
Corporation, 575 Albany Street, Boston, Massachusetts) mixed with proper
proportions of unlabeled NTA Na3-H20 (NTA Batch No. 1, supplied by the
Soap and Detergent Association, 485 Madison Avenue, New York, New York)
to give the desired activity and NTA Na^ concentration.  Stock solutions
of NTA Na3 were sterilized by filtering through 0.45 y Millipore filters
and aseptic techniques were employed throughout the sorption studies to
eliminate the possibility of microbial degradation of NTA.

For determination of quantities of NTA sorbed by the soils in contact with
NTA solutions, 2 g portions of soil were carefully weighed into individual
500 ml Erlynmeyer flasks stoppered with polyurethane plugs.  Flasks and
contents were sterilized by autoclaving for 15 minutes at 121°C.  A 100 ml
aliquot of sterile aqueous l^C-NTA Na3 solution containing 2, 10, or 40 mg/1
NTA Na3 was added aseptically to each flask.  Blank flasks containing
aliquots of the same NTA Na3 solutions but no soil were similarly prepared.

Both blank flasks and those containing soil were agitated on a rotary shaker
at 20°C.  Periodically, flasks were removed from the shaker for approximately
15 minutes to allow soil solids to settle, and a 2 ml sample of the aqueous
phase was removed aseptically from each flask.  These samples were centrifuged
to remove suspended matter.  Levels of radioactivity, and hence NTA Na3, in
the centrifugates were determined by means of a Beckman LS-150 liquid
scintillation spectrometer, employing 1.5 ml of centrifugate in 16 ml of
toluene scintillation cocktail containing 8.0 g/1 Butyl PBD and 0.5 g/1 PBBQ.
When maximum scrption of NTA Na3 on soil particles had occurred as indicated
by attainment of constant NTA Na3 concentrations in the aqueous phases of
the soil flasks, the quantities of NTA Na3 sorbed per gram of soil at the
observed equilibrium concentrations were determined from the following
relationship .

-------
                        Total NTA Na3
                        in 100 ml of
NTA Na» Sorbed/g Soil = solution in
      3
                        blank flask
Total NTA Na3      Total NTA Na3
in liquid phase    removed from
of soil flask   +  soil flask
at equilibrium     during sampling
Values thus obtained were utilized for construction of sorption isotherms.
                         Soil Column Studies

                   Studies Employing Natural Soils

Initial soil column studies were concerned with the possible degradation of
NTA during infiltration through unsaturated and saturated sand, loam, and
clay-loam soils and the effect of NTA on metals native to these soils.

The soils employed were those used in the previously described sorption
studies, namely Konawa loamy fine sand, Claremore loam, and Burleson clay
loam.  These soils were packed into 10.2 cm O.D. x 117 cm borosilicate
glass columns equipped with 1 cm O.D. outlets.   Plugs of glass wool topped
by 2.5-4 cm of pea gravel served as supports for the soils in the columns.

For preparation of sand and loam columns, soil was added to the column in
small increments sufficient to increase the column height 5-8 cm, and the
column contents were prodded with a metal rod after each addition to insure
uniformity of packing.  However, preparation of clay loam columns by this
procedure was unsatisfactory because of very low rates of liquid flow
through the columns obtained.  Clay-loam columns were therefore prepared
by first screening the soil to obtain an 8-70 mesh cut of soil particles
and then adding this material incrementally to the columns, employing
vibration to obtain relatively uniform packing.  Two columns of each soil
type were prepared, with exactly equal weights of soil being utilized for
preparation of both columns of a given type.  Columns of soil were 9.5 cm
in diameter and averaged 99 cm in height.  Additional descriptive data for
the columns are given in Table 1.
                                Table 1

                 Weight and Volume Data for Soil Columns

                                          Water Required
                                         To Fill Column To  Water Drained
  Column                Weight Soil (g)  Top of Soil (ml)   by Gravity (ml)

Konawa Loamy Fine Sand      11,460            2,170             1,300
Konawa Loamy Fine Sand      11,460            2,110             1,280
Claremore Loam               8,780            3,260             1,285
Claremore Loam               8,780            3,045             1,165
Burleson Clay Loam           9,630            3,490                *
Burleson Clay Loam           9,630            3,540                *
  *Data not obtained.
                                    10

-------
One column of each soil type was operated under unsaturated, or aerobic,
conditions with the water table near the bottom of the column, while the
other column of the pair was operated under saturated, or essentially
anaerobic, conditions by maintaining the water table near the surface of
the soil.  Columns were operated at room temperature (usually 22-25°C) in
a darkened fume hood.  The loam and clay-loam columns used in these studies
are pictured in Figure 2.

Columns were dosed daily for 58 days at a hydraulic loading of 2 gal/ft2
(580 ml/day for each column) with a weak synthetic sewage prepared from
the liquid diet food Sego (Pet, Incorporated, Saint "Louis, Missouri;
BOD approximately 163,000 mg/1), l^C labeled NTA Na3, and distilled water.
This synthetic sewage, which contained approximately 50 mg/1 both of BOD
and NTA Na3 and had a specific activity of 900 pCi/ml, was fed to the
columns at a rate of 2 ml/min during a daily dosing period of approximately
5 hours by means of a Technicon proportioning pump (Technicon Corporation,
Tarrytown, New York).  Sterile synthetic sewage was prepared fresh daily
and feed lines were washed thoroughly with detergent and flushed with
95 percent ethanol after each dosing in order to minimize the possibility
of NTA degradation occurring before the synthetic sewage entered the columns.

Upon completion of the initial 58 days of operation, dosing of the soil
columns was continued by essentially the same procedure for 19 additional
days, except that no NTA was added to the synthetic sewage.

Column effluents were analyzed daily for NTA Na^, radioactive carbon and
pH.

NTA Na~ was determined by an automated zinc-zincon procedure which was
essentially that of Thompson and Duthie,(2) except that a 6 mm O.D. x
56 cm tube packed with 20-50 mesh Amberlite IR-120 ion exchange resin
(HT1" form) was inserted in the system between the sampling and mixing
stages to remove cations likely to interfere with the analysis.  Radio-
active carbon was determined by the previously described liquid scintil-
lation spectrometry procedure employed in the sorption studies.

Selected effluents were also subjected to thin layer chromatography and
radioautography in an effort to detect intermediate organic degradation
products of NTA.  Ten milliliter effluent samples were concentrated under
vacuum in rotary evaporators, and 50 ul aliquots of the concentrate were
chromatographed with standard samples of NTA and some possible NTA
degradation products including:  N-methyliminodiacetic acid; iminodiacetic
acid; N,N-dimethylglycine; glycine; and trimethylamine.  Thin layer
chromatography systems employed were:  Avicel SF microcrystalline cellulose
as stationary phase and either tertiary butyl alcohol-formic acid-water
(40:10:16, called tBFW) or phenol-water (3:1) as mobile phase; and, a
mixture of 20 percent Amberlite CG-400 anion exchange resin-80 percent
Avicel SF as stationary phase with either tBFW or 0.15 N sodium chloride
as mobile phase.  R,. values obtained with these systems are presented in
Table 2.
                                   11

-------
JMft I  .+.-mL
            FIGURE 2,  LOAM AND CLAY LOAM
                       SOIL COLUMNS
                           12

-------
                                Table 2

                  Rf Values of NTA and Some Postulated

                        NTA Degradation Products

                             	Stationary Phase
                                                 20% Amberlite CG-400
                              Avicel SF          80% Avicel SF
  Compound                   MP 1   MP 2         MP 1            MP 3

Nitrilotriacetic Acid        0.41   0.31         0.33            0.38
N-Methyliminodiacetic Acid   0.44   0.58         0.42            0.90
Iminodiacetic Acid           0.37   0.31         0.35            0.82
N,N-DimethyIglycine          0.54   0.91         0.48            0.92
Trimethylamine               0.55   0.88         0.57            0.93
Glycine                      0.44   0.37         0.38            0.87

   Mobile Phases:
     MP 1    t-butyl alcohol-formic acid-water  (40:10:16)
     MP 2    phenol-water  (3:1)
     MP 3    0.15 N sodium chloride


The system of 20 percent  CG-400 resin-80 percent Avicel SF with 0.15 N NaCl
was most used in this work because it was less affected than the other systems
by substance in the effluent concentrates which tended to interfere with
migration of NTA and NTA  degradation products  on the chromatograms and also
because it gave significant separation between NTA and the possible degradation
products studied, although the latter compounds were not individually differ-
entiated.

Developed chromatograms were placed in contact with Kodak No-Screen Medical
X-Ray film for 22-24 days to obtain radioautograms.  After radioautography
chromatograms were sprayed with chromogenic reagents to visualize the standard
compounds in order that radioautograms and chromatograms could be compared.
The visualization process involved spraying with 0.01 percent zinc sulfate
in methanol, drying the plate while exposing it to ammonia fumes, spraying
with 0.1 percent zincon (2-carboxy-2'-hydroxy-5'-sulfoformazylbenzene) in
ethanol, and again drying the plate in ammonia fumes.  By this procedure,
NTA and the  possible NTA  degradation products  studied were visualized as
rosy pink spots on a blue background.  As little as 0.5 yg NTA produced
visible spots on chromatograms, but much larger quantities of N,N-dimethylglycine
and trimethylamine were required for detection.

In order to  determine the extent of carbon dioxide production resulting from
degradation  in the columns of NTA labeled with l^C,   CC^  was determined in
selected effluents.  For  this determination, aliquots of effluent samples
were acidified with 0.1 N hydrochloric acid and the acidified samples were
then purged with nitrogen gas.  The purge gas  from each sample was passed
through a CC>2 scavenger solution consisting of 2-ethoxyethanol and 2-ethanolamine
(8:1), and the activity of the scavenger solution was then determined by liquid
scintillation spectrometry.

                                   13

-------
Column effluents were also analyzed for selected metals in an effort to
discern effects of NTA on solubility and transport of metals native to the
soils used in the columns.

Emission spectrograms were first obtained from effluents produced at the
beginning of the study, when no NTA was passing through the columns, and
from effluents produced after NTA or its degradation products began, to elute
from the columns, as indicated by radioactivity of the effluents.  This
involved ten-fold evaporative concentration of the effluent samples,
followed by analysis of the concentrates by means of a Jarrell-Ash Mark IV
Ebert convertible plane grating spectrograph, employing the rotating disc
attachment for sampling and a high voltage spark for excitation.  The
spectrograms thus obtained were compared semi-quantitatively to detect those
metals appearing to increase significantly in concentration as NTA concen-
trations in the effluents increased.  These metals (iron, manganese, zinc,
nickel, and cobalt) were then quantitatively determined by atomic absorption
spectrophotometry in effluents obtained throughout the study, employing a
Perkin Elmer Model 403 atomic absorption spectrophotometer and standardized
analytical procedure.(6)


              Studies Employing Sand Enriched with Metals

Upon completion of the initial soil column studies, additional studies of a
similar nature were begun with the principal objective of gaining further
information concerning the possible transport of metals into ground water by
NTA.  For these studies soil columns 9.5 cm in diameter and approximately
99 cm in height were prepared from Konawa loamy fine sand enriched with lead,
zinc, chromium, cadmium, and mercury.

All columns were initially packed to a depth of approximately 35 cm with
loamy fine sand containing no additive, employing the same packing procedures
as used for preparation of the sand and loam columns employed in the initial
soil column studies.  Enriching materials were then thoroughly mixed with
sufficient additional saad to provide a total weight of 11,460 g of soil in
each column, and packing of the columns was completed with these enriched
soils.

Two columns were prepared from sand enriched with the following natural ores
(obtained from Burminco Rocks and Minerals, 128 Encinitas, Monrovia, California):
galena (PbS); sphalerite (ZnS); chromite (Fe(Cr02)2)5 greenockite (CdS); and
cinnabar (HgS).  These ores were ground to insure homogenity, analyzed to
determine actual metal content, and then added to the sand matrix in sufficient
quantities that the individual packed column would contain 5 g of each of the
five metals except chromium.   Available quantities of chromite, which contained
only 0.1 percent chromium, were sufficient to provide only 995 mg of this metal
per column.

The other four columns were packed from sand enriched principally with a
synthetic sludge prepared from a solution of soluble salts of lead, cadmium,
chromium, and zinc.  The metals were precipitated by adjusting the solution
to pH 10 with sodium hydroxide, and the resulting sludge was collected by


                                    14

-------
centrifugation, dried, and ground to a powder.  This material, supplemented
with, mercuric sulfide  (cinnabar) as a source of mercury, was mixed with
sand in such quantity  that the individual packed column would contain 5 g
of each of the five metals.

Two of the columns containing synthetic sludge were operated under unsatttrated,
or aerobic, conditions, while the other pair of sludge-containing columns and
the two columns containing the ores were operated under saturated, or
essentially anaerobic, conditions.  In order to further eliminate possible
aerobic zones in the tops of saturated columns, a blanket of nitrogen was
maintained over these  columns at all times.
                                                               r\
The columns were dosed daily at a hydraulic loading of 2 gal/ft  (580 ml/day
for each column).  One column of each pair (unsaturated sludge, saturated
sludge, and saturated  ore) was dosed with 67 mg/1 of ^C labeled NTA Na~ in
distilled water (specific activity 920 pCi/ml).  The second column of each
pair served as a control and was dosed with distilled water only.  The same
precautions as used in the initial soil column studies were employed in
this work to prevent significant microbial activity in the feeds until they
had entered the column.  In order to assess the effects of an additional
carbon source on the parameters of interest, sufficient Sego to impart a
BOD of 50 mg/1 was added to the daily feed solutions starting with the
56th and 53rd day of operation, respectively, for the columns dosed with
NTA and the control columns.  The columns were operated for a total of
92 days.

Effluents from columns dosed with NTA Na-j were analyzed for NTA Na^ by
the automated zinc-zincon method, for radioactive carbon by liquid
scintillation spectrometry, and for pH in the same manner as in the initial
soil column studies.   Iron, manganese, lead, zinc, chromium, and cadmium
were determined by atomic absorption spectrophotometry, employing standard
procedures. '°'  Mercury in selected effluents was determined by the FWQA
Provisional Method for mercury  (Flameless AA Procedure)(?) employing a
Perkin Elmer Coleman Model 50 mercury analyzer system.

Selected effluents were also analyzed for total organic carbon, employing
the Beckman Model 915  Total Carbon Analyzer,("' and redox potentials of
effluents from saturated columns were intermittently determined during the
latter half of the studies by means of Orion Model 96-78 platinum redox
electrodes.  These electrodes were placed in effluent delivery tubes at
points considerably removed from outlets to preclude possible interference
resulting:from back diffusion of atmospheric oxygen.

Control column effluents were subjected to all the analyses listed above
except NTA Na3 and 14C.


                       Simulated Aquifer Studies

In order to investigate the possible degradation of NTA in an essentially
anaerobic ground-water environment, model aquifers simulating as nearly
                                   15

-------
as possible conditions prevailing in a natural aquifer were constructed
in the laboratory.  Figure 3 shows details of the simulated aquifer design.

Each laboratory aquifer was constructed from a 4 1 aspirator bottle
containing 3 1 of natural aquifer sand.  A 6 mm O.D. borosilicate glass
tube which extended through the rubber stopper sealing the aspirator
bottle and into the aquifer sand almost to the bottom of the aquifer
container served as a well shaft.  This tube was perforated below the
surface of the sand, with the perforations placed in a spiral arrangement.
The tube was packed loosely with glass wool to prevent entry of aquifer
sand into the water pumped from the aquifer and was topped with a 1/4 inch
Swagelok Zytel tee which served as a well head.  The side arm of this tee
was connected to a closed loop of Tygon tubing which passed through a
peristaltic pump and back to the aquifer unit where it was attached to a
short piece of tubing which penetrated the sealing stopper.  This system
permitted water to be pumped from the saturated depths of the aquifer sand
through the well head and returned via the closed tubing system to the
aquifer unit.  The upper arm of the well head tee was sealed with a Teflon-
lined rubber septum through which the pumped stream could be sampled.

The stopper of each aquifer unit was also penetrated by a glass tee, one
arm of which served as a port for evacuation and gassing of the simulated
aquifer while the other arm was sealed by a rubber sleeve-type septum
through which the gases above the sand and water could be sampled.  In
addition, each aquifer unit was provided with still another 6 mm O.D. tube
penetrating to the gas space above the aquifer sand to serve as an outlet
for pressure checks and venting of gases if needed.

Natural aquifer sand for construction of the simulated aquifers was obtained
near Byars, Oklahoma, from a shallow aquifer underlying the floodplain of the
South Canadian River.  The sand was obtained by carefully excavating to a
point slightly below the water table, then scooping the saturated sand into
a sterile stainless steel bucket with a sterile scoop.  The sand was then
taken immediately to the laboratory and 3 1 portions were loaded under
aseptic conditions into each of three presterilized simulated aquifer units.
After completion of sand loading one of these units was sterilized again at
121°C for 30 minutes to provide a control.

Most of the water native to the aquifer sand was removed from all three units
by applying a vacuum of approximately 29 inches of mercury at the bottom
outlets of the aspirator bottles.  Sterile solutions of NTA Na3 in distilled
water were then allowed to flow slowly into the evacuated units through the
lower outlets until the liquid levels were approximately 25 mm above the
sand levels.  Approximately 700 ml of sterile NTA Na., solution were required
for each simulated aquifer unit.  The solution charged to all three aquifers
contained 50 mg/1 NTA Na3.  The feed solutions for the sterile control
aquifer and one of the other units, designated Simulated Aquifer Number 2,
contained 900 pCi/ml ^C-NTA (uniformly labeled).  The feed for the third
aquifer unit, designated Simulated Aquifer Number 1, contained no isotopically
labeled NTA.  Initial concentrations of NTA Na3 and l^C in the aqueous
                                    16

-------
                       Zytel Tee
                       Well Head
rTeflon Lined  Septum .


          Flow -*-
To Vacuum
Pump  8 N-
Source
                                           Flow
        ) Rubber  Stopper
                                    Gloss Tubing
                                    Well Shaft
                                                    Closed Loop of  Tygon Tubing

                                                    Passing Through Peristaltic Pump
                                                        4  Liter
                                                        Aspirator  Bottle
                                                           Screw  Clamp
                                                      Glass Wool
                      FIGURE 3,   DETAILS OF  SIMULATED
                                   AQUIFER DESIGN
                                         17

-------
phases of the aquifers were actually only 60-70 percent of concentrations
in the feed solutions because of sorption of NTA on the aquifer sand and
dilution of the feed solutions by native water remaining in this sand.

After being charged with NTA Na-j solution, the simulated aquifers were
twice thoroughly vacuum evacuated and refilled with nitrogen gas through
the top gas ports in order to establish conditions of very low oxygen
tension within these units.  The three aquifers were then placed in a
constant temperature room where they were maintained in the dark at 20°C
throughout the period of investigation.  Figure 4 pictures the simulated
aquifers and peristaltic pump as arranged during operation.
                     FIGURE 4,   SIMULATED AQUIFERS
                                    18

-------
The simulated aquifers were operated for a period of 265 days.  They
were recharged with solutions containing additional NTA after 52, 143,
and 209 days of operation.

The first recharge was accomplished by aseptically injecting through the
well head septums of both Aquifers Number 1 and 2 ten ml of sterile
solution consisting of 60 mg of NTA Na« in distilled water.  This solution
was expected to increase NTA Na3 concentrations in the water in the
aquifers by 35-45 mg/1, but uncertainties regarding the actual volume
of water in the aquifers and the quantity of NTA Na3 that would be sorbed
on aquifer sand precluded calculation of an exact quantity of NTA Nag
needed to give a definite concentration in a recharged" aquifer.  The
recharge solution for Aquifer Number 2 also included 13.2 yc of uniformly
labeled l^C-NTA which was expected to increase radioactivity 10-12 fold
and hence facilitate radioautography studies.  The control aquifer was
not recharged at this time.

For the second recharge sterile solutions consisting of 60 mg of NTA Na^
in 200 ml of distilled water were pipetted aseptically into Aquifers
Number 1 and 2 through the gas sampling ports, after removal of the sleeve-
type septums, in order to replenish both NTA and water levels in the
units.  A 200 ml solution containing 30 mg of NTA Nao was similarly added
to the control aquifer.  The recharge solutions for Aquifer Number 2 and
the control also contained 39.6 yc each of l^C-NTA.  All septums and worn
tubing were changed at this time, and the units were re-evacuated and re-
gassed with nitrogen.  Also, to further preclude the possibility of oxygen
entering the aquifer units, all three units and the peristaltic pump system
were at this time placed in a specially constructed plexiglass box filled
with nitrogen and were maintained there for the remainder of the investi-
gation (Figure 4).

The third recharging of Aquifers Number 1 and 2 was conducted by essentially
the same procedure as that employed in the second recharging.  However, the
200 ml of recharge solution also contained 0.5 ml Sego in order that the
effect of an extraneous carbon source on NTA degradation in the aquifers
might be observed.  On the basis of the BOD of Sego, these solutions should
have imparted a BOD in the range of 60-90 mg/1 to the water in the aquifer.
The control aquifer was recharged with 100 ml of sterile solution consisting
of 0.5 ml Sego in distilled water with no NTA.

The aquifer units were pumped occasionally for a few hours to simulate
movement of ground water and to maintain as much uniformity as possible
within the units, particularly after recharging.  Units were always pumped
for a short period immediately before sampling in order to obtain repre-
sentative samples from the depths of the aquifers.  Sampling, which was
generally conducted at 3-7 day intervals throughout these studies, was
accomplished by withdrawing required quantities of liquid through the well
head septums by means of sterile syringes while the units were being pumped.
                                    19

-------
Samples taken from the simulated aquifers were analyzed for NTA Na., by the
automated zinc-zincon method and for l^C by liquid scintillation spectrometry
as previously described.  Selected samples were also subjected to radioautography,
employing essentially the same procedures as used for soil column effluents,
except that 10-50 yl aliquots of aquifer water samples were usually chromat-
ographed with no prior concentration.  This chromatography of dilute samples
was an effort to avoid chromatographic migration problems encountered in the
soil column work resulting from concentration of salts and possibly other
substances present in the samples.  Use of the dilute samples for radio-
autography was possible because of the relatively high activity of the
aquifer samples after the first recharge.  Autograms were also exposed for
as long as 96 days in an effort to detect substances present only in very
low quantity on the chromatograms.

The composition of the gas above the water and sand in the simulated aquifer
units was determined each time a liquid sample was obtained.  A Varian-
Aerograph 90 P gas chromatograph equipped with an internal and external
column in series was employed for these analyses.  The internal (first)
column was 80/100 mesh Porapak Q (6.35 mm x 183 cm) operated at 55°C while
the external (second) column was 60/80 mesh Molecular Sieve 5A (6.35 mm x
92 cm) operated at room termperature (23-25°C).  Carrier gas (He)  flow
rate was 50 ml/min.  This system permitted analyses for methane,  carbon
dioxide, oxygen, and nitrogen with a single injection of sample.
                                    20

-------
                               SECTION V

                         RESULTS AND DISCUSSION

                            Sorption Studies

The data obtained for sorption of NTA on Burleson clay loam, Claremore
loam, and Konawa loamy fine sand soils in this investigation are best
described by means of the Freundlich adsorption isotherm, as shown in
Figure 5.  Greatest sorption was evidenced by the loam soil, with the
sand being least effective as a sorbant.  At an equilibrium concentration
of 50 mg/1 of NTA Na-j in water sorption values for the loam, sand, and
clay-loam soils were 64, 28, 8.7 yg NTA Nao per gram of soil, respectively.
Analogous data at 5 mg/1 NTA Na3 were 10.2 yg/g, 3.5 yg/g, and 0.98 yg/g.

These data show that NTA is not strongly sorbed by any of the three soils
studied.  The relatively weak sorptive forces exhibited by these soils
toward NTA cannot, however, be completely discounted in considering the
movement of NTA into and through ground water because of the high ratio of
soil encountered by water infiltrating through the soil mantle and moving
within the water table.  This is illustrated by considering the movement
of NTA through a hypothetical 10 ft  column of soil of 1 ft^ cross sectional
area and 10 ft depth.

Data obtained in column preparation for the soil column studies (see Table 1)
indicate that a 10 ft^ column of dry Claremore loam would weigh about 352 Kg
and would hold approximately 126.5 1 of water when saturated.  The sorption
data obtained in this work (y/m = 62 yg/g at C = 50 mg/1) (Figure 5) show
that in order for the water in the column of loam to contain 50 mg NTA Na3/l
under equilibrium conditions there would necessarily be a total of about
22.5 g NTA Nao sorbed on the soil.  Assuming plug flow and absence of non-
equilibrium conditions, it can be calculated that in order for the NTA Nao
concentration of the effluent from this column of soil to equal the influent
concentration the column would have to be dosed for approximately 76 days
at the daily rate of 2 gal/ft^ if the influent concentration were 50 mg
NTA Na3/l.  If there were no sorption of NTA by the soil, theoretically
only 17 days of such dosing would be required for influent and effluent
NTA concentrations to be identical.  Hence, the time required for an NTA Na~
solution of 50 mg/1 to move through the Claremore loam would appear to be
increased by 4-5 fold as the result of sorption of the NTA on the soil
particles.  Similar calculations indicate a 6-7 fold increase in migration
time through the loam soil for a solution containing 5 mg NTA Na3/l
because of the relatively greater proportion of NTA sorbed at lower
concentrations.  For the Konawa loamy fine sand migration times for 5 to
50 mg/1 solutions of NTA Nao are indicated by such calculations to be
increased only about two-fold as the result of sorption effects.

These calculations, while admittedly crude, do indicate that under some
circumstances sorption on soil particles could considerably retard the rate
of infiltration of NTA into ground water and the movement of this compound
                                    21

-------
0.2
                                                    Clay Loam
                                                   K = 0.80
                                                   n = I.10
  Loam
K = 2.80
n = I .25
                                        Sand
                                       K = 0.22
                                       n = I.07
                       Where:

                 = Unit Adsorption -
                                                           NTA
                                  m                     g soi I
                                  C - Equilibrium Concentration - mg/ I

                                  K and n are empirical constants
6  8  10
                                  20
                       40  60  80 100
                       C (mg/l  NTANa3)


               FIGURE 5,   i(TA SORPTION; FREUNDLICH
                          ISOTHERM AT 20° C
                           22

-------
within aquifers.  Nevertheless, as  an individual mechanism, sorption of
NTA by soils does not appear  to be  sufficient  to greatly reduce the
potential pollution of ground waters by  this compound  over tne course
of its long-term usage as  a detergent builder.  However, by retarding
the movement of NTA in soils  and hence effectively increasing its
residence time in regions  of  high microbial activity sorption may play
a significant secondary  role  in the removal of NTA from infiltrating
waters by microbial degradation.
                           Soil  Column  Studies

                     Studies Employing  Natural  Soils

The results  obtained in  the soil  column experiments  concerned with the
effect of  infiltration of  NTA through  unsaturated  (aerobic) sand, loam,
and clay-loam soils  are  reflected in Figures 6-8, which present graphically
the concentrations of NTA, radioactive carbon,  and selected metals occur-
ring in  the  unsaturated  soil  column effluents  during the 77-day period of
inves tigation.

As shown in  Figure 6, NTA  was present  in  the effluent from the unsaturated
sand column  from the second through the ninth  day of operation, with a
peak level near  feed concentration occurring on the  fifth day.  By the
tenth day, however,  NTA  had virtually  disappeared from this effluent and
was not  again detected as  an  effluent  constituent during the remaining
67 days  of operation. Essentially no  NTA appeared in effluents from the
unsaturated  loam and clay  loam  columns at any  time during the 77 days of
operation, even  though approximately 29 mg of  NTA Nao were applied daily
to each  column (Figures  7  and 8).  The early presence and subsequent
disappearance of NTA in  the sand  column effluent undoubtedly reflected a
short acclimation period required for  establishment  within the soil of a
microbial  population capable  of degrading NTA.   Because of the greater
sorptive power of the loam and  clay-loam  soils,  the  movement of NTA in
these columns was probably retarded sufficiently that acclimation and
subsequent degradation occurred before any NTA traversed the lengths
of these columns.

During most  of the period  of  operation, effluents from all three unsaturated
soil columns  contained significant quantities  of -^C which were not accountable
as NTA,  thus  indicating  the presence in these  effluents of products of NTA
degradation.   Thin-layer chromatography and radioautography studies of
selected effluent samples  collected throughout  the operating period did not
reveal the presence  of any organic degradation  products of NTA in these
effluents.  Analysis  of  selected  effluent samples from the three columns
for  l^COo  indicated that  virtually all of the  l^C present in these
effluents  was  incorporated in CC^  except  that  appearing in l^C-NTA in the
early effluent samples from the sand column.
                                   23

-------
                                                  14,
     --50	
  800-
      -40
  600-
U
  400-
  200-
      -IO
                           Feed Level, NTANo3   8    C
                  10      20       30      40
                         Days of Operation
50       60
70
           FIGURE 6.   i(TA, I4C,  AND ftiALs  IN  EFFLUENT
                       FROM UNSATURATED SAND COLUNN
                               24

-------
   10
   6
0)
    + 50
 800-
     -40
 600-
     -30
<0.400-
          to
         o
         z
     -20
 200-
     -10
                                  Level, NTA No,  8 I4C
                                  .14,
                                  NTA

                 10       20      30       40      50


                            Days  of  Operation
                                                           60      70
             FIGURE 7,   NTA/ -"C/ AND METALS IN EFFLUENT

                         FROM UNSATURATED LOAM COLUMN
                                    25

-------
     16




     14




     12




     10
jo
a
                                                     I	1	I	I    l	I
                                                        14,
      --50
                                 /-Feed Level, NTANa3  a '^C
   800-
       -40
   600-
       -30
U
40O-
       -20
            1C
            o
            z
   200-

                                •NTA
                       i	i f   i	i	i

                                                                  \
                   10       2O      30       40

                           Days  of  Operation
50       60
                                                                  70
              FIGURE 8,   I(FA,   C, AND  ftiALs IN EFFLUENT

                          FROM UNSATURATBD CLAY LOAM COLUMN
                               26

-------
As  Figures  6,  7,  and 8 show,  a major portion of the radioactive  carbon
applied  to  the unsaturated soil columns never appeared in the column
effluents.   It seems likely that most of this l^C was  lost to the
atmosphere  as  I4cc>2,  with perhaps  a small quantity being incorporated
into  cellular  material in the columns.   This is consistent both with the
analytical  data for the column effluents and the report by Thompson  and
Duthie which shows  that aerobic microorganisms can  degrade NTA completely
to  C02 and  nitrate. (2)

These experiments with unsaturated  soil columns indicate  that NTA  infil-
trating  through unsaturated soil systems probably will undergo rapid and
complete degradation,  with the only products likely to enter  ground  waters
as  a  result of such degradation being C02 and probably inorganic nitrogen
compounds .   The magnitude of  additional inorganic nitrogen pollution of
ground water which  is  likely  to result  from the projected use of NTA in
detergents  is  probably not sufficient for critical  concern.   For example,
complete degradation in unflooded septic tank percolation fields of
35  mg NTA Na3/l,  a  concentration which  could occur  in  septic  tank  effluents
if  detergents  containing 25 percent NTA Na3 were used, (&)  would result in
a maximum increase  of  only 1.91 mg  of nitrogen irt each  liter  of percolate
entering underlying ground waters.   Nevertheless, introduction into  ground
waters of additional nitrogen in the quantities likely  to result from NTA
usage in detergents is  certainly less than desirable,  and could result in
problems in aquifers where wide fluctuations of water  tables  may result
in  salt  concentration  through sequential deposition and resolubilization.

Concentrations of zinc and iron in  the  effluent from the  unsaturated sand
column attained peak levels when NTA was present in highest concentration
in  this  effluent  and declined significantly when NTA disappeared from the
effluent (Figure  6).   Hence,  NTA would  appear to be implicated in  the
leaching of these metals from the soil.   Manganese  concentrations  also
peaked in conjunction with NTA concentration,  but remained relatively high
after NTA had  disappeared from the  effluent.

Relatively  high concentrations of manganese and iron appeared in the
effluents from the  unsaturated loam and clay-loam soil  columns as  dosing
of  the columns with  NTA progressed  (Figures 7 and 8) . ,  In the absence of
control  columns not  dosed with NTA,  however, it cannot  be ascertained if
leaching of these soils  metals was  in any way related  to  infiltration of
NTA into the soil.   The  fact  that iron  and manganese concentrations  in
the effluents  did not decline significantly when NTA dosing of the columns
was terminated casts doubt on existence of such a relationship.  The
possibility was not  eliminated,  however,  that  increased quantities of
metals might enter ground waters as  a result of a mechanism involving
initial  solubilization  of soil metals through formation of NTA chelates,
with  the metals remaining in  solution and continuing to move  through the
soil  after  degradation  of the NTA.
Figures 9, 10, and 11 show graphically  the concentrations of NTA Na^,
and metals appearing in the effluents from the saturated sand, loam, and
clay- loam soil columns during the initial experiments concerning the effect
of infiltration of NTA through saturated (anaerobic) soils systems.
                                   27

-------
As shown by these figures, NTA first appeared in column effluents  on  the
fourth, eleventh, and seventh day of operation respectively  for  the sand,
loam, and clay-loam columns.  NTA Nag in the saturated sand  column effluent
attained peak concentrations of about 46 mg/1 during the period  of the
ninth through eleventh days of operation, with -^C showing a corresponding
peak of about 935 pCi/ml, which was somewhat above the l^C feed  level of
900 pCi/ml (Figure 9).  In the saturated loam column effluent a  peak  NTA
concentration of nearly 59 mg/1, which was considerably above the  feed
level of 50 mg/l, and a corresponding l^C peak in the neighborhood of
1160 pCi/mlwere exhibited on the sixteenth day of operation  (Figure 10).
After attaining these peaks, concentrations of NTA Nag in effluents from
both the saturated sand and loam columns decreased and remained  within
the range of 65-85 percent of feed levels until dosing of the columns with
NTA Nag was terminated.  Following the initial peaks, l^C levels in both
effluents also decreased and were generally 80-95 percent of feed  concen-
tration for the remainder of the dosing period.

The effect of dosing NTA Nag to the saturated clay-loam column was quite
different than the effect of similar operation of the sand and loam columns
during the first 30 days of study.  As Figure 11 indicates, both NTA  Nag
and l^C effluent concentrations attained relatively low peaks on the  14th
day of operation and then declined to minima on about the 29th day.   The
decline in NTA Nag and l^C concentrations from the 14th to 29th  days  of
operation was attributed to aerobic degradation of NTA resulting from the
presence of air in the capillary fringe of the column and the entrapment
of air pockets in the soil interstices throughout the column.  After  about
the 30th day of operation, the saturated clay-loam column began  performing
in a manner similar to the other saturated columns, probably at  least
partially as a result of intentional flooding of the capillary fringe of
the column on the 29th day.  Depletion of oxygen in trapped intersticial
pockets of air, leading to a more truly saturated state within the column,
was also probably significantly involved in this change in column  performance.

There appears to be no ready explanation for the peak concentrations  exceeding
feed levels of l^C in effluents from the saturated sand and loam columns and
of NTA Nag in the effluent from the loam column.  However, the fact that both
NTA Nag and l^C effluent concentrations were less than feed levels and that
l^C effluent levels were equivalent to 2-10 mg/1 more NTA Nag than actually
appeared in column effluents during most of the operating period for both
the saturated sand and loam columns and for the clay-loam column after the
30th day, indicate that some degradation of NTA was occurring within  these
columns.  Analysis of column effluents for  -^CC^  revealed that  a major
portion of the l^C present in form other than NTA was incorporated in C02-
However, thin-layer chromatography and radioautography of selected effluent
samples obtained throughout the study showed clearly that some of  this l^C
was present as a component of at least one organic degradation product of
NTA.

Unknown substances in the effluent samples which caused migration  anomalies
on thin-layer chromatograms permitted thin-layer chromatographic analysis
of the effluents only by means of the system employing Amberlite CG-400
                                    28

-------
     10



     8
to

~a
2    2
          fr   ^v-^^=
          //  Mn -^
  IOOO-
    800-
0
o.
   600-
   400-
   20O-
       -60
                                              14,
                            Feed Level, NTANo3 S IHC
               10       20       30      40

                          Days of Operation
50      60
70
            FIGURE 9,   NTA, 14C/ AND TOTALS IN EFFLUENT

                       FROM SATURATED SAND COLUMN
                                   29

-------
12


10
                                                                 Fe
J2   6
o
4)
^   4
  1200-
                     A
                                                    I4
                                     Level, NTANo3  a  C
                                                                   70
                            Days  of  Operation

            FIGURE 10,  IJTA/  ^ C, AND ftiALs  IN EFFLUENT
                        FROM  SATURATED LOAM COLUMN
                              30

-------
      12
      10
 o»
 E
 
-------
resin-Avicel SF as stationary phase with 0.15 N NaCl as mobile phase.  In
this system the corrected Rj value of the NTA degradation product occurring
in the saturated column effluents appeared to most nearly correspond with
that for iminodiacetic acid (see Table 2, page 13).  However, iminodiacetic
acid and several of the other postulated NTA degradation products have very
similar Rf values in the CG-400 resin-Avicel SF-0.15 N NaCl system, and hence
their differentiation is questionable when migrational problems such as those
occurring in chromatography of the effluent samples are encountered.

The quantities of organic product of NTA degradation present in the
saturated soil column effluents were apparently quite low, and further
efforts to characterize this substance were not undertaken during these
studies.

It cannot be ascertained conclusively from these experiments that the
observed NTA degradation was truly anaerobic in nature, or if it was
actually aerobic degradation which occurred in the capillary frirge
regions of the columns and which was arrested by lack of oxygen as the
feed solution percolated deeper into the columns.  The possibility of
aerobic degradation is supported by the following facts:  feed solutions
contained some oxygen when they entered the tops of the soil columns
because of the dosing procedure employed; and, intentional flooding of the
capillary fringe of the saturated sand column on the 35th day of operation
resulted in a decrease in the amount of NTA degradation occurring in this
column (Figure 9).  However, more than twenty percent of the NTA Nag
entering the saturated sand column was apparently degraded even after
flooding of the capillary fringe.  This observation, combined with confirmed
anaerobic degradation of NTA in simulated aquifer studies described later
in this report, provides support for the occurrence of at least some true
anaerobic degradation of NTA in the saturated soil columns.  Initial limited
aerobic degradation followed by slow anaerobic decomposition would appear
a distinct possibility.  The possibility of such a sequence gives reason
to speculate that the organic degradation product of NTA observed in these
studies might have resulted from arrested aerobic degradation.

Overall, these studies with saturated sand, loam, and clay-loam soil
columns indicate that NTA infiltrating through saturated soil systems,
such as flooded septic tank percolation fields or under waste water
lagoons, will likely undergo at most only slow degradation, and hence
high proportions will probably reach ground water essentially unchanged.
The possibility is suggested that NTA degradation products of possible
health concern(9) might be produced in saturated soils, but the evidence
for this possibility is tenuous and if such compounds should indeed be
produced, the quantities entering ground water would likely be extremely
small.
                                    32

-------
As indicated in Figures 9,  10, and  11,  there were generally  good  correl-
ations between concentrations of  certain metals and concentrations of
NTA Na^ in effluents from the three saturated  soil columns.  These metals
were:  iron, zinc, and manganese  for  the sand  column; iron and nickel
for the loam column; and, iron, nickel, and cobalt (although the  concen-
tration of the latter element was' never very high) for  the clay-loam
column.  The decreases in metals  concentrations in the  sand  and loam
column effluents which accompanied  decrease in NTA Na3  concentrations
following termination of NTA Na3  dosing on the 58th day of operation
appear particularly significant.  Observation  of the effects of termina-
tion of NTA dosing on metals concentrations in the effluent  from  the
saturated clay-loam column  was not  possible because clogging and
resulting poor flow performance precluded the  collection of  further
meaningful data from this column  after  about the 60th day.  However, the
minima in effluent metals concentrations corresponding  to the minimum in
NTA Naq concentration occurring on  the  29th day of operation provides
significant additional information  concerning  the effect of  infiltrating
NTA on the metals in the clay-loam  soil.

Reasonable caution must be  exercised  in interpreting the metals data
obtained in these experiments with  saturated natural soils because of
the absence of data from control  columns.  Nevertheless, the studies
do appear to indicate strongly that NTA solutions infiltrating through
saturated soil systems are  likely to  solubilize and transport into ground
water some metals native to the soils.
               Studies Employing Sand Enriched with Metals

In Figures  12  and  13 are presented results obtained when solutions con-
taining  l^C labeled NTA Na-j were infiltrated through both unsaturated
and saturated  columns of Konawa loamy fine sand enriched with metals
sludge containing  lead, cadmium, chromium, zinc, and mercury.

As Figure 12 shows, NTA first  appeared in the effluent from the unsatu-
rated column on  the 19th day of operation and attained a maximum concen-
tration  of  49  mg NTA Na3/l  (73 percent of the 67 mg NTA Na3/l feed) on
the 39th day.  Thereafter, NTA concentrations declined slowly and became
relatively  stable  at about 25  percent of feed level after the 70th day.
Radioactive carbon levels generally were very close to equivalent NTA Na3
concentrations during the first 54 days of operation, but then remained
at higher levels  (40-55 percent of feed) than NTA for the remaining period
of study.

Cadmium  levels in  the effluent from this column correlated well with NTA Na3
reaching peak  concentrations of about 14 mg/1 at 40-44 days of operation.
Iron, which was not added in the metals sludge, first appeared in the
effluent on the 4th day of operation and ranged between 2 and 3 mg/1 from
the 17th day to the 55th day,  while very low levels (maximum about 0.5 mg/1)
                                    33

-------
    32




    28




    24
 o«  20

J:



 w  16

 o


 *  12
  1080-T 80




   960- • 70




   340-
   720-
_ 60O-
O

 Q.
   ^BO-
   SS 0-
   24O-
    120-
                    •Feed Level, NTANo3 a    C
       - 60
        50
       -40
       - 30
       h 20
       - 10
 ro
o
10      20     30     40      5O      60     70



                Days  of  Operation
                                                                  80
                                                          90
                FIGURE J2.  IJTA,  14C, AND METALS IN EFFLUFNT

                            FROM UNSATURATED SAND-f'tTALs  SLUDGE

                            -COLUMN
                                34

-------
    36



    32



^  28





I  ^


 «  20
N   '6
 .   12
•o
o
       216



       192



       168



       144



       .20  ^

             o>

       96  ^


             o>
       72   I



       48



       24
   1080-



   960



   840-



   720-



^ 600



tT 48O-
o.


O 360-



   240-



   120-
               .Feed Level, NTA No,   a
        I ^j   ^	,_.j--  - — _   _.   _         _
        30
        20
        10
                 10
20     30     40      50


      Days  of  Operation
                                                    60
                                                            70
             FIGURE 13,  [JTA,  -0, AND METALS IN EFFLUENT

                          FROM  SATURATED SAND-f-trALs  SLUDGE

                          COLUW
80
90
                                     35

-------
of zinc were present in the effluent when maximum concentrations of NTA
occurred.  The effluent from the control column of sand enriched with
metals sludge, which was dosed only with water or water and Sego, contained
some cadmium during the first 4 days of operation but none thereafter.  The
quantity of cadmium eluted from the experimental column dosed with NTA was
at least ten-fold greater than that eluted from the control column.  The
control effluent contained only very low levels of iron (maximum 0,3 mg/1)
from the 27th day through the 53rd day and did not contain zinc at detectable
levels (0.05 mg/1) at any time during the study.

Figure 13 shows that NTA first appeared in the effluent from the saturated
column of sand enriched with metals sludge on the 22nd day, attained a
maximum concentration of about 80 mg NTA Na3/l (119 percent of the 67 mg/1
feed level) on the 57th day, and then slowly declined to about 60 mg NTA Na3/l
at termination of the experiment.  Radioactive carbon effluent levels paral-
leled NTA Na3 concentrations throughout the period of operation, but during
much of this period were equivalent to as much as 15 percent less NTA Nao
than indicated to be present in the effluent by the zinc-zincon analytical
method.  This discrepancy may have resulted from presence in the effluent
of some unknown substance causing a positive error in the zinc-zincon
analytical procedure.  Limited studies involving purification of NTA in
effluents by thin-layer chromatography followed by analysis by the zinc-
zincon method gave some credence to this postulate.  Color in some of the
samples may have produced a quenching effect, contributing to some reduction
in apparent   C levels also, although experimental evaluation of this
possibility indicated that any such quenching which occurred was very
limited in effect.  At any rate, the NTA Nao and l^C analytical data
obtained appeared adequate to suffice for the purposes of this study.

Cadmium levels in the effluent from the saturated sand-metals sludge column
generally correlated well with NTA Na^ concentrations, attaining a maximum
of about 36 mg/1 on the 56th day of operation when the NTA Na3  concentration
also was highest.  Iron also appeared in the effluent at a maximum concentra-
tion of 38 mg/1 coinciding in time with the peak NTA Na3 level.  However, iron
concentrations were erratic, showing several peaks and valleys.  Considerably
less total iron than total cadmium was eluted from the column over the period
of operation.  In addition, lesser quantities of zinc, chromium, lead, and
mercury appeared in the effluent during the period that NTA was present, with
peak concentrations generally appearing in synchrony with iron peaks.  Maximum
observed concentrations for these metals were:  Zn, 7.1 mg/1; Cr, 5.8 mg/1;
Pb, 3.8 mg/1; Hg, 122 yg/1.  The effluent from the control column of saturated
sand enriched with metals sludge contained relatively very low levels of
cadmium (maximum 5 mg/1), mercury (maximum 2 yg/1), and iron (maximum 2 mg/1)
on a few isolated occasions during the study.  For the most part, however,
levels of the metals determined in the control effluent were below detectable
limits (less than:  0.05 mg/1 for Fe, Zn, and Cd; 0.2 mg/1 for Cr; 0.3 mg/1
for Pb; and 0.05 yg/1 for Hg) during the study.

Results obtained when NTA was infiltrated through a saturated column of
Konawa loamy fine sand enriched with ores of lead, cadmium, chromium, zinc,
                                    36

-------
and mercury are presented graphically in Figure 14.  As this figure
shows, NTA Na- first appeared in the effluent from this column on the
third day of operation, attained a concentration near the feed level
(67 mg/1) on the seventh day, and remained near feed level throughout
the remaining 85 days of the study.  Radioactive carbon levels in the
effluent followed NTA Nag concentrations very closely except during the
period of the 5th through about the 20th day of operation when, as in
the case of the saturated sand-metals sludge column, l^C levels were
significantly below equivalent NTA Na, concentrations.

Iron was present in some concentration in the effluent from the saturated
sand-ore column from the seventh day of operation to the end of the study.
As with the saturated sand-metals sludge column these effluent concentra-
tions of iron were erratic, showing major peaks of about 13 mg/1 and
37 mg/1 on the 32nd and 45th day of operation, respectively.  Zinc first
appeared in the effluent on the third day of operation, attained a peak
concentration of 3.2 mg/1 on the 9th day, and was present at concentrations
below 0.5 mg/1 from the 20th day until termination of operation except for
low peaks occurring when maximum iron concentrations were noted.  Low
concentrations of chromium  (maximum 2.1 mg/1), cadmium (maximum 0.5 mg/1),
lead  (0.4 mg/1), and mercury (maximum 65 yg/1) also appeared in the
effluent when maximum iron  concentrations were present.  Concentrations
of these metals in. the effluent from the control sand-ore column were
generally below detectable  limits.  A low concentration of iron, peaking
at 4.2 mg/1 and exceeding 0.2 mg/1 only during the 33rd through 45th day
of operation, and several concentrations of mercury not exceeding 4 yg/1
occurring from the 29th through the 36th day, were  the only exceptions.
The total quantities of these metals eluted from the control column were
many  times less than the total quantities eluted from the experimental
column.

Comparison of the NTA and ^C data obtained in this second series of soil
column studies with similar data obtained from the. unsaturated and saturated
sand  columns in the initial soil column studies (Figures 6 and 9) reveals
that under both unsaturated and saturated conditions the movement of NTA
through sand enriched with metals sludge (Figures 12 and 13) was very much
slower than through sand alone.  On the other hand, movement of NTA through
saturated sand enriched with metal ores (Figure 14) was little different
than movement of this compound through sand alone  (Figure 9).  Obviously
the metal hydroxides that principally constituted the metals sludge exerted
some retarding effect on infiltrating NTA not manifested by the metal ores,
which were mostly sulfides.  The mechanism of this retarding effect, however,
remains a matter of conjecture.

Comparison of Figures 6 and 12 shows that although NTA infiltrating through
the unsaturated sand column was completely degraded after a short acclimation
period, a maximum of only about 75 percent degradation was achieved in the
column containing sand enriched with metals sludge operated under similar
conditions.  It seems most  likely that this effect was due to toxic effects
of the heavy metals in the sludge which resulted in partial inhibition of
                                    37

-------
     36T2I6
     32--192
     28- 168
o>
_£
«
U-
e~
.
O
•o"
O

24-

20-

16

12


8
144
*"•
120 a>
3
96 01
X
72


48
     4  24
   I08CH

   96O

   840-

   720-

   600-

\ 480-
6"
 0.
^ 360H
u
—  240-

   120-
        80
70
60
50
-10
              NTA
                        Feed Level NTANa3 8 I4C
                 10
                20
30
40
50
60
70
80
                              Days  of  Operation
              FIGURE 14,  NTA, ^C, AND [%TALS IN EFFLUENT FROM
                          SAND-f%TALbORES
                                                                         90
                                38

-------
microbial activity in the column.  Comparison of Figures 13 and 14 with
Figure 9 reveals that if any degradation of NTA occurred in the saturated
sand-metals sludge and sand-metal ores columns, it was certainly signifi-
cantly less than occurred in the saturated sand column during the initial
soil column studies.  This also suggests the possibility of inhibition
of NTA degradation due to metals toxicity, although more effective exclu-
sion of oxygen from the upper zones of the saturated columns in the second
series of soil column studies may have contributed to this apparently
decreased degradation.

The presence of NTA appears to be the only explanation for the significantly
greater metals concentrations in the effluents from the experimental columns
than in the effluents from the control columns.  The experimental and control
column pairs were prepared from identical quantities of the same enriched
soils and were operated by identical procedures except for the addition of
NTA to the feed for the experimental columns.  The pH values, which ranged
between 7 and 8 for all the columns during most of the period of operation,
were practically identical for the individual experimental and control
column pairs.  For the individual pairs of saturated columns there were
essentially no differences noted in redox potentials of effluents from
experimental and control columns.  E^ values were determined only during
the last half of the experimental period on an intermittent basis, but were
found to vary very little, ranging near -400 mv for the saturated sand-
metals sludge columns and -500 mv for the saturated sand-metal ores columns.

Differences in total organic carbon concentrations of effluents from
experimental and control columns appeared to mainly reflect the presence
of NTA in the experimental column effluent, as illustrated by Figure 15.
The effect oh TOG of adding sufficient Sego to impart a BOD of about
50 mg/1 to feeds for both experimental and control columns is obvious in
Figure 15.  However, this additional carbon source in feeds produced no
discernable effect on metals elution or NTA degradation in any of the
columns under study.

There is no ready explanation for the "surges" of metals, particularly iron,
occurring in the effluents from the saturated sand-metals sludge and sand-
metal ores columns.  It should be noted, however, that these effluents,
which were generally somewhat more cloudy than control effluents, became
extremely turbid and highly colored during periods of maximum metal elution
This is illustrated by Figure 16, which pictures effluents from saturated
sand-metals sludge experimental and control columns.

These experiments with sand columns enriched with selected metals leave
little doubt that infiltration of water containing NTA through saturated
soils containing heavy metals could result in solubilization and subsequent
transport of such metals into ground waters.  The extent and nature of such
solubilization and transport in a specific situation will be dependent on
a number of variables.  This is illustrated by the qualitative and quanti-
tative differences in the metals eluted from the sand-metals sludge and
sand-metal ores columns in these studies.  These variables will undoubtedly
                                    39

-------
    60
    50
_   4O
v,
o»
E
 c
 ©
.o
 fc»
 o
o
 o
 c
 o

 w
O
30
    20
    10
                                                 •TOC- Column Dosed

                                                     With  NTA
                                                     TOC-Control Column
                                                 C  Equivolent to NTA

                                                 Concentration— Column

                                                 Dosed With NTA
10    20     30     40     50     60

                Doys  of  Operation
                                                   70
                                                      80
90
100
             FIGURE  15.  ORGANIC  CARBON IN EFFLUENTS FROM
                          SATURATED SAND-f%TAi_s SLUDGE COLUMNS
                              40

-------
            FIGURE 16,  VISUAL COMPARISON OF EFFLUENTS FROM
                        SATURATED SAND-ftrALS SLUDGE COLUMNS
                        DOSED WITH FEEDS WITH AND WITHOUT NTA
include the quantities of various metals  in  the soil, the anionic species
with thich the metals are associated  in  the  soil matrix, the  concentration
of NTA in the infiltrating solution and  the  relative stabilities of  the
various metal-NTA  complexes under conditions prevailing in  the particular
saturated soil system of concern.

For example, cadmium was eluted in considerable quantity from the saturated
sand-metals sludge column, but appeared  only in very low quantity in the
effluent from the sand-metal ores column, apparently because  it was present
as the hydroxide in the sludge column and as the sulfide in the ores column.
Also, cadmium and iron apparently formed  the most stable complexes with NTA
of any of the metals determined under conditions prevailing in the saturated
                                   41

-------
sand-metals sludge column, and hence appeared- in greatest quantity in the
effluent from this column.  Zinc, chromium, lead, and mercury in this column
almost certainly were also chelated by NTA as indicated by their appearance
in low levels in the effluent with maximum concentrations generally corres-
ponding to maximum NTA levels.  However, the. stabilities of these ehelates
under conditions in the column were such that most of the NTA was involved
in formation of chelates with cadmium, iron, and possibly other metals not
determined in this study.  Had cadmium been present in lower concentration
or not present at all in this column, or had conditions prevailed in the
column under which the cadmium-NTA complex was less stable, such as a
different pH, one or more of the other metals likely would have appeared
in high concentration in the effluent.

One of the objectives of this second series of soil column studies was to
gain further information concerning, the possibility that metals in unsatu-
rated soil systems might be leached into ground waters as a result of initial
solubilization by infiltrating NTA., with the metals remaining in solution
and moving on into the underlying ground water even though the NTA portion
of the metal-NTA chelate was rapidly degraded subsequent to chelate formation.
No definitive information was obtained concerning this possibility because
only partial degradation of NTA occurred in the unsaturated column of santi
enriched with metals sludge.   However, these studies- did point out that
infiltration of water containing NTA through unsaturated" soils, containing
high levels of heavy metals,  could result in both NTA and metals entering
the underlying ground water because of metals toxicity and the resulting
inhibition of microbial degradation of NTA.

Overall, these studies appear to indicate a potential for pollution of ground
water by heavy metals in situations where NTA might infiltrate through soil
systems containing high levels of metals such as cadmium,, zinc, chromium,
lead, and mercury.  Introduction of NTA from detergents into waste lagoons
which have been receiving metals plating wastes would appear to be a
situation of particular concern.


                       Simulated Aquifer Studies

Figures 17 and 18 graphically depict the data from simulated aquifers employed
to investigate the possible degradation of NTA in an essentially anaerobic
ground-water environment.

As shown in Figure 17, concentrations of 30-32 mg NTA/Na3/l at the beginning
of aquifer operation decreased to essentially zero in a period of 52 days in
the aqueous phases of both Aquifer Number 2, which contained NTA uniformly
labeled with l^C, and Aquifer Number 1 which contained only nonlabeled NTA,
while the level of l^C in Aquifer Number 2 decreased approximately 40 percent
from about 560 pCi/ml to 330 pCi/ml.  During this same 52-day period of
operation NTA Na3 in the sterile control aquifer decreased from approximately
33 mg/1 to 29 mg/1, with l^C in this unit showing an even smaller decline.
                                   42

-------
   35-
   30-
                                    initiol  Chorge


                                   14C - Control
 10
o
Z

-------
                                                   Second  Recharge
          145    150    155
160    165    170    175    180    185    I9O    195
                      Third  Recharge
  40-
     -30
-3CH
o>
E
 (O
2 20H
   IO-
      10
         210    215    220   225   230    235    240   245    250    255
                            Days  of  Operation

           FIGURE 18,  DEGRADATION  OF [[FA IN SIMULATED AQUIFERS;
                       SECOND AND THIRD RECHARGES,
                                           260
                                44

-------
These data indicated beyond reasonable doubt that significant microbial
degradation of NTA had occurred in the experimental simulated aquifers
during this initial 52-day period of operation.  However, occasionally
during this period gas samples taken from the aquifer systems were shown
by gas chromatography to contain low levels (usually less than 0.2 percent)
of oxygen, probably as a result of problems with liquid sealed pressure
relief systems which were employed on the aquifers early in the study but
which were subsequently abandoned.  Hence, there was some doubt concerning
the role of oxygen in the observed degradation of NTA during the initial
period of aquifer operation.

Results obtained when the simulated aquifers were recharged with additional
NTA Na^ on the 52nd and 143rd days of operation served to further clarify
the nature of NTA degradation in these units.

In the period following the first recharging (Figure 17), degradation of
NTA was observed to occur at higher overall rates than in the initial
period of operation, with initial concentrations of 38 mg NTA Na~/l in
Aquifer Number 1 and 41 mg NTA Na-j/1 in Aquifer Number 2 declining to zero
in 47 and 33 days, respectively.  Following the second recharging, the
aquifers were operated under a nitrogen blanket to further preclude the
possibility of diffusion of oxygen into the units through tubing and septa.
An initial concentration of 46 mg NTA Na3/l in Aquifer Number 1 was
completely degraded in 55 days, while a concentration of 54 mg NTA Na3/l
initially present in Aquifer Number 2 was reduced to 5 mg/1 in the same
period  (Figure 18).

Gas samples obtained from the simulated aquifers during the periods of
operation following the first and second rechargings were shown by gas
chromatographic analyses to contain essentially no oxygen.  These same
analyses also revealed the production of methane and low levels of carbon
dioxide in both experimental aquifers during these periods, with no
production of either gas in the control aquifer.

Anaerobic degradation of NTA in the simulated aquifers was indicated beyond
reasonable doubt by confirmation of the methane produced in Simulated Aquifer
Number 2, which contained l^C labeled NTA, as a degradation product of NTA.
This was accomplished by fractionation of aquifer gas samples by chromatog-
raphy on a 6.35 mm x 183 cm Porapak Q column, with passage of successive
fractions of effluent gas from the chromatograph through 16 ml aliquots
of toluene scintillation cocktail and subsequent determination of the
radioactivity of these aliquots by liquid scintillation spectrometry.  A
typical analysis is presented in Figure 19.  As this figure shows, the
carbon dioxide in the aquifer gas was also an NTA degradation product, as
would be expected.  Since the toluene scintillation cocktail was undoubtedly
not efficient in scavenging the radioactive methane and carbon dioxide from
the gas chromatography effluent, the activities shown in Figure 19 probably
are indicative of much higher actual activities of the gases produced in
the aquifer.
                                    45

-------
Sample: 2 ml Gas From
       Aquifer  Number  2

Column: 6.35mm  x 183cm
       Porapak  Q,  80/100 mesh
       Fraction

          I
          2
          3
          4
          5
       Blank
Radioactivity  (CPM)

     38.8
    364.8
     39.8
    465. 9
     41.4
     38. 4
 FIGURE 19,  DETECTION OF RADIOACTIVITY IN GASES
             FROM A SIMULATED AQUIFER,
                   46

-------
Thin-layer chromatography and radioautography of water samples obtained
from Simulated Aquifer Number 2 during the periods of operation following
the first and second rechargings clearly showed the progressive disappearance
of NTA from this unit after recharging.  However, there was no evidence for
the presence of any of the postulated degradation products of NTA listed
in Table 2 at any time during these periods of operation.  Radioautograms
did provide some evidence for the presence in this aquifer during degra-
dation of l^C-labeled NTA of at least one unknown radioactive compound
with Rf values very similar to those of NTA in the chromatographic systems
employed.  This substance was most noticeable in radioautograms of samples
obtained during maximum degradation of NTA following -the second recharge
of Aquifer Number 2.  It appeared to have corrected Rf values of 0.43
and 0.34, respectively, when chromatographed using the CG-400 resin-Avicel
SF-0.15 N NaCl system and the Avicel SF-tBFW system, while the corresponding
Rf values for NTA using these systems were 0.38 and 0.41.  Time limitations
precluded further investigation and characterization of this possible
anaerobic degradation product of NTA.  The chromatographic properties of
this substance were so similar to those of NTA that the possibility of
its being a chromatographic anomally, produced by the effect on NTA of
salts or other substances in the sample solutions, cannot at present be
eliminated.

Radioautograms of water samples obtained from Aquifer Number 2 after all of
the NTA had disappeared (as indicated by the zinc-zincon procedure) revealed
no radioactive products other than very low quantities of substance at Rf
values similar to those of NTA.  This material was sufficient to account
for only a small portion of the carbon-14 still present in the aqueous
phase of the aquifer at this time (Figure 17) .  Although a portion of this
l^C must have been present in the form of carbon dioxide and methane, it
would appear likely that other volatile degradation products, which would
have been lost during sampling and chromatography and hence would not have
been detected by radioautography may have been present in the aqueous phase
of the aquifer after all of the NTA had been degraded.

Figure 18 shows that degradation of NTA proceeded readily in both of the
experimental aquifers after the third recharging, which included Sego as
a carbon source in addition to NTA.  As shown by both this figure and
Table 3, overall rates of NTA degradation in the two simulated aquifers
following this recharge were somewhat less than the rates following the
first two rechargings.  The magnitudes of these rate decreases were
relatively small, however, and the extent to which they were attributable
to the extraneous carbon source is open to question.  Products of microbial
metabolism, accumulating in the aquifers during the 210 days of operation
prior to the third recharge, could also have caused some decrease in NTA
degradation.  It should be noted that addition of an extraneous carbon
source to an aquifer with the initial charge of NTA, before the microbes
present in the aquifer have adapted to degrade NTA, could produce a quite
different effect than observed here.  There was, however, no opportunity
to conduct such an experiment during these studies.
                                    47

-------
                                Table 3

           Summary of NTA Data for Simulated Aquifer Studies


                                                                 Mean Rate
Operational                     NTA Na-^ (mg/1)        Days of   of Degradation
  Period         Aquifer   Start  Finish  Decrease  Operation    (ing/I/day)

Initial Charge      1        32      0       32        52          0.615
Initial Charge      2        30      0       30        52          0.577
Initial Charge   Control     33     29        4        52          0.077

First Recharge      1        38      0       38        47          0.809
First Recharge      2        41      0       41        33          1.241
First Recharge   Control     29     27        2        53          0.038

Second Recharge     1        46      0       46        55          0.836
Second Recharge     2        54      5       49        55          0.891
Second Recharge  Control     48     43        5        55          0.091
Third Recharge      1        38      5       33        55          0.600
Third Recharge      2        41      2       39        55          0.710
Third Recharge   Control     42     38        4        55          0.073
Overall, the data obtained in the simulated aquifer studies indicated that
NTA entering an essentially anaerobic ground-water environment will probably
undergo degradation.  This phenomenon could be quite important in retarding
the development of high levels of NTA in aquifers receiving waste waters
containing this substance and in decontamination of aquifers which might
become polluted by NTA.  These studies also indicate that the final products
of NTA degradation within an aquifer will likely consist mostly of methane,
carbon dioxide, and probably other volatile materials.  Further investigation
is needed to determine if possibly deleterious organic products might be
generated during such degradation.

The sand employed for construction of the simulated aquifers in these studies
was obtained from a very shallow natural aquifer which probably contained a
numerous and active population of anaerobic microorganisms because of periodic
infusion with water of relatively high organic content.  This, as well as
retention time, probably explains the differences in the extent of NTA
degradation in the simulated aquifers and the previously described saturated
soil columns.  The soil columns were constructed from well aerated soils
that were necessarily processed in preparation for packing in a manner likely
to result in destruction of any anaerobic microorganisms that might be present.
It is also likely that many natural aquifers would have anaerobic microbial
populations much less numerous and active than the population present in the
aquifer providing the sand used in these studies.  Hence, the degradation of
NTA, or at least the initiation of such degradation, would be likely to occur
much more slowly in many natural aquifers, particularly deep aquifers, than
was observed in these studies.
                                    48

-------
                               SECTION VI

                            ACKNOWLEDGMENTS

A project of this nature necessarily involves the energy and talent of a
number of people to whom proper credit should be given.

Special acknowledgment is made of Bobby Newport who contributed greatly
in the design and operation of pertinent column studies.

Linda Harmon is acknowledged with thanks for designing the computer
program used in analyzing project data and for somehow managing to
type various drafts of the report while meeting all deadlines.

The assistance of Richard E. Thomas, who provided invaluable advice in
regard to soil column preparation and operation, and of Jack W. Keeley,
who was responsible for a significant portion of the initial project
planning and provided assistance and encouragement throughout the
investigations, is gratefully acknowledged.
                                     49

-------
                              SECTION VII

                               REFERENCES

1.  Anonymous, "Rx for Ailing Lakes - A Low Phosphate Diet," Environmental
    Science and Technology, _3, 1243 (1969).

2.  Thompson, J. E. and Duthie, J. R. , "The Biodegradability and Treatability
    of NTA,"  Journal WPCF. 40, 307 (1968).

3.  Shumate, K. S., Thompson, J. E., Brookbait, J. D., and Dean, C. L.,
    "NTA Removal by Activated Sludge - Field Study," Journal WPCF, 42,
    631 (1970).

4.  Pfeil, B. H., and Lee, G. F., "Biodegradation of Nitrilotriacetic Acid
    in Aerobic Systems," Environmental Science and Technology, 2^, 543 (1968).

5.  Murray, C. Richard, "Estimated Use of Water in the United States, 1965,"
    U. S. Geological Survey Circular 556 (19650-

6.  U. S. Federal Water Pollution Control Administration, "FWPCA Methods
    for Chemical Analysis of Water and Wastes," FWQA Analytical Quality
    Control Laboratory, Cincinnati, Ohio (1970).

7-  U. S. Federal Water Quality Administration, "FWQA Provisional Methods
    for Mercury (Flameless AA Procedure)," FWQA Analytical Quality Control
    Laboratory, Cincinnati, Ohio  (1970).

8.  Duthie, J. R., "Trisodium Nitrilotriacetate (NTA) - Environmental
    Safety Review," Sanitary Engineering Research Services Report, The
    Proctor and Gamble Company, Cincinnati, Ohio  (1970).

9.  Epstein, S. S., Testimony in Hearings before  the Subcommittee on Air
    and Water Pollution, of the U. S. Senate Committee on Public Works,
    May 6, 1970.
                                     51

-------
   .Accession Number
                           Subject Field & Group
                            05B
                                          SELECTED WATER  RESOURCES ABSTRACTS
                                                 INPUT TRANSACTION FORM
   Organization
          Robert S. Kerr Water Research Center, Ada, Oklahoma
          National Ground Water Research Program
          Investigations Concerning Probable Impact
          of Nitrilotriacetic Acid on Ground Waters
10
   Authors)
          Dunlap, William J.
          Cosby, Roger L.
          McNabb, James F.
          Bledsoe, Bert E.
          Scalf, Marion R.
                               16
Project Designation

 EPA Project No.  16060 GHR
                                21
                                   Note
22
    Citation
23
Descriptors (Starred First)

      *Ground Water,  *Path of  Pollutants,  *Biodegradation, Sorption, Metals, Detergents
25
Identifiers (Starred First)
      *Nitrilotriacetic Acid
27
    Abstract
   Laboratory studies were employed  to  investigate  the  fate and effect  of NTA both in
   ground waters and in soil profiles overlying  ground  waters.
   Studies of the sorption of NTA by sand,  loam,  and  clay-loam  soils  indicated that
   sorption of NTA on soils could slow  its  movement into  and through  ground waters.
   Sorption will probably not be sufficient to prevent  or greatly reduce potential
   pollution of ground water by NTA  used  as a detergent builder.
   Soil column studies were employed to investigate the degradation and effect en metals
   of NTA infiltrating through soils.   These studies  indicated:   NTA  infiltrating through
   most unsaturated soils likely would  undergo rapid  and  complete degradation and
   contribute only inorganic nitrogen compounds  and carbonate to ground waters; NTA
   infiltrating through saturated soils would probably  experience only  very limited
   degradation, with a major portion entering ground  water intact; any  NTA which escaped
   degradation during infiltration through  soils  could  transport such metals as iron,
   zinc, chromium, lead, cadmium, and mercury from  soils  into ground  waters.

   Studies with model aquifers constructed  from  natural aquifer sand  indicated that NTA
   would likely undergo slow degradation  in essentially anaerobic ground-water environ-
   ments, with production of CO , CH^,  and  possibly other organic compounds.
Abstractor
William
WR:102 (REV.
WRSIC
,T
JULY
Hi ml
1969)
at?
Institution
Robert

S.

Kerr

Water
SEND TO
Research Center


WATER RESOURCES SCIENTIFIC INFORMATION
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240

CENTER
                                                                              » SPO: 1969-359-339

-------