Witeo Witco CorporationOne American LaneGreenwich, CT 06831-2559(203) 552-2000(203) 552-2010 Fax

17 June 1996

Mr. Eric NewfnanUSEPA (3HW42)841 Chestnut StreetPhiladelphia, PA 19107

Re: Treatability Study Summary Progress Report

Dear Mr. Newman:

In accordance with Section 8.7 of the Administrative Order for RemovalResponse Action (EPA Docket No. IH-95-55-DC) for the Halby ChemicalSite, Witco Corporation hereby submits a Treatability Study SummaryProgress Report for the bench scale work. This report is intended tosummarize the work and results to date and is not intended to fullyreport on all treatability activities. The final bench scale TreatabilityStudy Report will provide additional documentation on background,analytical procedures, quality assurance, etc. If you have any questionson this matter, please call me at 203-552-2476 or Richard J. Dulcey ofERM at 610-524-3610.

Sincerely,

Raj VyasProject Coordinator

cc: M. MausnerP. FahrentholdB. MercurioR. DulceyM.GoldJ.NortzV.HahnJ. Biggs-SangerB, RootJ. Mattox

A R U O I 2 0 I

REPORTORIGINAL

(Red)

WTTCO Corporation

Treatability Study SummaryProgress ReportOxidation of Carbon Disulfide inSoils — Halby Chemical Site,New Castle, Delaware

17 June 1996

Prepared by:Fahrenthold & Associates

1470 Enea Circle, Suite H-1740Concord, CA 94520

andEnvironmental Resources Management, Inc.

855 Springdale DriveExton, Pennsylvania 19341

TABLE OF CONTENTS

1.0 INTRODUCTION . 1-1

2.0 CHEMISTRY OF CARBON DISULFIDE OXIDATION 2-1

3.0 CHEMICALS AND SOILS FOR TESTING 3-1

4.0 EXPERIMENTAL CONDITIONS 4-1

5.0 EXPERIMENTAL PROCEDURE 5-1

5.1 SPIKED SOIL EXPERIMENTS 5-1

5.2 HALBY SITE SOILS — SMALL-SCALE STUDIES 5-2

5.3 HALBY SITE SOILS — LARGE-SCALE STUDIES 5-5

6.0 DISCUSSION OF RESULTS TO DATE 6-1

6.1 SPIKED SOIL RESULTS 6-1

6.2 HALBY SITE SOIL 6-1>

7.0 CONCLUSIONS TO DATE 7-1

ATTACHMENT

A CARBON DISULFIDE SAMPLE ANALYSIS

A R U O I 2 0 3

LIST OP FIGURES

1 Reaction Mix Temperature Increase with Sodium Percarbonate Addition2 Reaction Mix Temperature Increase with Sodium Perborate Addition3 Conversion of C$2 Versus Mole Ratio of Oxidant to C$2

ii

CWSffttl(Bcfl '

1 Non-Halby Spiked Soils Reacted at Varying Oxidant: C$2Molar Ratios

2 Small-Scale Halby Site Soils Experimental Matrix3 Small-Scale Halby Soil Samples Reacted Under Controlled Temperature

(<25°C) Conditions4 Large-Scale Halby Soil Study Molar Ratios5 Performance Data from Treatment of Halby Site Soils6 Calculations for Percent Reduction in CS2 Concentrations

111

1.0 INTRODUCTION {Red}

The purpose of this treatability study was to investigate and develop aneffective, in-situ treatment for the oxidation of carbon'disulfide (CS2)present in soils at the Halby Chemical Site. Treatment has been proposedto reduce the concentration of carbon disulfide in soils below that whichconstitutes an unacceptable risk. Because of concerns regardingexcavation, a method that could be applied in-situ was pursued in thisstudy.

The study was performed on.representative soils from the site; generally,those from the area of the former on-site ditch at depths known to containelevated levels of CS2 (averaging 34,000 mg/kg). Samples were collectedon 2 and 3 April 1996 by Langan Engineering and Environmental Services,Inc. (Langan) utilizing the sampling protocol outlined in Langan's letter toMr. Eric Newman of the U.S. Environmental Protection Agency (EPA)dated 25 March 1996.

The laboratory experiments were conducted in a manner as representativeof field conditions as possible. This included maintaining sealedcontainers and generally performing experiments at cooled temperature,with limited pre-experiment homogenization so as to obtainrepresentative analysis. Initial small-scale samples, as described further inthis document, were analyzed during the treatability studies usingscreening methods for short-turnaround analysis. Large-scale sampleswere analyzed using EPA CLP methods and quality assurance/qualitycontrol procedures to ensure that the data gathered in the study werereliable. Three sets of experiments have been essentially completed, withthe fourth in progress:

• studies using spiked non-Halby soils,• small-scale studies using Halby soils,

• large-scale studies using Halby soils, and

• focused studies using Halby soils for quantification of processparameters.

As is the case with most treatability studies, this work diverged from theWork Plan as more information was gathered. However, the workdescribed herein encompasses a greater scope than was anticipated in theoriginal Work Plan.

The first phase of testing was designed to determine overall oxidationcharacteristics for various oxidants on soil spiked with CS2. The second

1-1 _ _ . _ _ wrrco-6/17/w

A R l * O I 2 0 6

ffod}phase of testing determined whether the reaction between oxidants andcarbon disulfide occurred in site soils and the extent to which it occurredwith different oxidants. The third phase of testing was expected toprovide verification of the results of the initial phase and provide anevaluation of the significance of pH levels, rnoisture content, and time ofthe reaction. The fourth phase of testing is ongoing and is designed toanswer questions related to off-gas composition, hydrogen peroxide-caustic oxidation dynamics, and the fate of non-CS2 organics. The resultsof the experiments to date are reported in the following sections.

1-2 WTICO-6/17/96

ORSGINAl2.0 CHEMISTRY OF CARBON BISULFIDE OXIDATION

As reported in the literature, CS2 can be oxidized to innocuouscompounds through a process known as alkaline oxidation. The oxidationin alkaline solution has been studied by Adewuyi and Carmichael1. Thechemical equation is:

CS2 + 8H2Q2 + OH" ~>HCO3~ + 2HSO4'12H±±fiH2Q

The significant reagent for conversion of €82 is the peroxide (hydroxyl)radical. Four carriers of the hydroxyl radical were evaluated: hydrogenperoxide in aqueous solution, sodium persulfate (Na2S2O8)/ sodiumpercarbonate (Na2COs * 1.5H2O2)/ and sodium perborate (NaBO2 •H2O2 • 3H2O). Some other peroxide carriers and oxidizers had beenconsidered (sodium peroxide, ozone, oxygen) but they were not evaluatedin this study since they offered few advantages or several disadvantagesrelative to the four oxidants considered herein.

The oxidation process produces bisulfate ions, which are acidic, and act toretard further reaction of CS2 with the peroxide as the pH decreases. Inorder to counteract the decrease in pH, reactions were carried out in abuffered medium (in this case very moist soil, with added bufferingsolution when needed). The buffer concentration supplied adequatebasicity to maintain the reaction rate.

Adewuyi, U.G. and G.R. Carmichael. 1987. Kinetics of Hydrolysis and Oxidationof Carbon Bisulfide by Hydrogen Peroxide in Alkaline Medium and Application toCarbonylSulfide. Envir. Sci. & Tech. 21: 170-177.

2-1 . WnCO-6/17/96

3.0

Chemical additives were used for two purposes in the experimentalprogram: as oxidizing agents, and as pH adjustment compounds(buffers).

The oxidizing chemicals used in the study are available commercially.Hydrogen peroxide (50%), sodium percarbonate, sodium perborate, andsodium persulfate were obtained from Aldrich Chemical Company asproduct numbers 42,065-4; 37,143-2; 24,412-0; and.21,623-2, respectively.Sodium carbonate and sodium bicarbonate (both buffering compounds)were obtained from the same supplier as product numbers 22,353-0 and23,652-7, respectively. Sodium hydroxide was also used for pH control ininitial trials using hydrogen peroxide.

The non-Halby soil used for spiked sample evaluations was a silt-sandmixture not know to contain any carbon disulfide. This soil was obtainedfrom the Rio Grande river bed in Albuquerque, New Mexico, due toproximity to the treatability laboratory.

Two 5-gallon soil samples were obtained from the site according to theplan submitted to the EPA by Langan Environmental on 25 March 1996.The samples were analyzed at the site to determine the initialconcentration of CS2 present The two soil samples were immediatelyshipped under chain-of-custody to the ECD laboratory in Albuquerque,NM, where they were stored under refrigeration until needed. Threeanalyses were performed on the two samples obtained from the site, withthe following CS2 levels: 54,000; 3l;000; and 17,000 ppm, respectively.After receipt at the laboratory, one of the five-gallon samples washomogenized and further subdivided into five one-gallon samples andeach one-gallon subsample analyzed. The analysis results for thosesamples were: 27,000; 53,000; 17,000; 54,000; and 22,000 ppm respectively.Because of the variability in concentration, an average of 34,000 mg/kg(average of both sets of results) was used to represent the CS2concentration.

The samples were stored under refrigeration (at 0°C) prior to use. Excesssample (i.e., that not used in the experiments) will be returned to the sitefor treatment during the field implementation phase of the program. Aninventory of sample use was kept to account for all sample used in theexperiments. Treated samples aliquots not held for analysis and excesssoils were stored in sealed steel containers at room temperature.

3-1 WTTCO-6/17/96

4.0 EXPERIMENTAL CONDITIONS

The treatability experiments described in the following sections werecarried out on both non-Halby soils spiked with CS2 and on soilscontaining CS2 obtained from the site.

Using the average concentration of carbon disulfide values in the one-gallon subsamples of untreated soil, a theoretical peroxide requirementwas calculated. The stoichiometric requirement of oxidant as a percentageof the soil sample to be treated was derived from the above equation asthe concentration of CS2 in percent times 3.57S2. This value was thenadjusted for the presence of a carrier such as water (e.g., 50% hydrogenperoxide solution is 50% water) or other inert material.

Initial screening of the reaction between the oxidants and CS2 was doneon the non-Halby soils spiked with carbon disulfide. Fifty- to 100-gramsamples of non-Halby soil were containerized and spiked with CS2 at atarget concentration of 1,000 mg/kg. The soils were sealed, parafilmed,and allowed to equilibrate for 24 hours at approximately 21°C, with littleor no headspace.

The buffer used was a combination of sodium carbonate and sodiumbicarbonate. The composition of the buffer was one gram of sodiumcarbonate and one gram of sodium bicarbonate per 20 milliliters of water.In the case where sodium percarbonate or sodium perborate (inherentlybuffering compounds) was used as the oxidant source, buffer volume wasdecreased in order to conserve chemicals and not cause an excessivelyhigh pH condition in site soils.

The experiments performed on Halby soils used the same reagents andgenerally similar reaction conditions as those for the spiked soil samples.The sample to be treated must contain adequate water to allow mixing ofthe reagents and yet retain soil consistency. The maximum quantity ofwater as buffer or by dilution of 50% hydrogen peroxide was determinedthrough observations made during the course of the experiments. Small-scale studies on Halby soils were conducted on 100-gram soil samples;large-scale studies were conducted on 500-gram samples.

The calculation is expressed as weight of oxidant required equals the molecularweight of hydrogen peroxide times eight divided by the molecular weight of carbondisulfide.

4-1 WTICO-6/17/96

Reaction conditions were allowed to vary, depending on the objective ofthe experiments. For those experiments designed to determine conversionof CS2/ the reaction was carried out at atmospheric pressure and roomtemperature (25°C). For those experiments designed to evaluate the heatof reaction, the experiments were started at temperatures as low as 0°Cand allowed to reach an equilibrium temperature (no more than 80°C).

Small-scale studies were carried out using a reagent addition/oxidationtime of 30 minutes; large-scale studies were conducted using areagent/oxidation timeTSFShours to avoid excessive sample overheating.

4-2 WnCO-4/17/96

A R U O I 2 I I

5.0 EXPERIMENTAL PROCEDURE

5.2 SPIKED SOIL EXPERIMENTS

A set of initial screening experiments was carried out to determine theboundaries of the experimental parameters and the oxidant providing thebest performance in removal of carbon disulfide. These tests were used asa way to rapidly evaluate experimental success and sensitivity of reactionconditions.

Initial CS2 levels of CS2-spiked non-Halby soils and small-volume test.samples of Halby site soils were screened by ECD Environmental utilizingthe attached High Performance Liquid Chromatography (HPLC) method.A copy of the analysis procedure is attached.

CS2 Analytical Recovery

CS2 spiked non-Halby soil samples were used to determine the recoveryof CS2 from the soils. A calibration curve was prepared for the instrumentfrom 5,000 mg/kg to 625 mg/kg CS2. Several experiments were made tocheck the extraction and quantitation procedure (mini method validation).In addition, lower-concentration samples at levels as low as 60 mg/kgwere run to check the linear accuracy of the standard curve, which wasfound to be acceptable for screening purposes. After calibration of theHPLC, four samples (100 grams each) were spiked with CS2 from 1% to2% by weight (10,000 to 20,000 mg/kg) and allowed to equilibrate for 24hours in the soil at room temperature by the procedure described inSection 4.0. They were then extracted with methanol for at least 24 hoursand analyzed via the HPLC. Recoveries of the spikes from 70 to 85% wereobtained. These results indicated that the HPLC method was acceptablefor screening use. Continuing calibration samples run with the matrixspikes indicated that the HPLC maintained its standard response to CS2-

CS2 Oxidation in Spiked Samples

As a qualitative evaluation of non-Halby soils prior to spiking, an aliquotof non-spiked soils was mixed with 50% peroxide. No apparent reaction -was observed, indicating that the non-Halby soils did not have any readilyoxidizable material that would interfere with the experiments.

Several experiments were then run using caustic and 50% peroxide onnon-Halby soils spiked at 1,000 mg/kg CS2. For each experiment, CS2was added to 50 grams of soil and allowed to permeate the soil. The

5-1 • WTTCO-4/17/96

A f U O I 2 l 2

caustic (1 equivalent of sodium hydroxide) and the peroxide (8 molar •equivalents) were added to the soil simultaneously (these are thequantities necessary for theoretical 100% reaction of the CS2). In a secondexperiment, 24 molar equivalents of peroxide (three times theoretical) andthe same quantity of caustic (1 equivalent) were used. A controlexperiment using only caustic was run concurrently with those whereperoxide was added. For both soils treated with peroxide, the residualCS2 was less than 200 mg/kg (extrapolated below the lower standardvalue of 625 mg/kg).

The sample treated only with caustic also showed a correspondingly highdecrease in CSz probably due to the reaction of CS2 and OH'. Ihisreaction would not reach completion (conversion of the CS2 to sulfate)and would be very slow in the absence of more alkaline conditions (pH ofat least 10.5). As a result, caustic addition alone would leave the soil withlarge bisulfide and sulfide concentrations or a high pH, neither of which isappropriate for full-scale remediation. The sample treated with the 3xtheoretical quantity of peroxide reacted violently and with generation ofsignificant heat.

Initial experiments were also carried out with the solid oxidants(percarbonate and perborate)/and spiked native soils. The experimentsused buffered solutions of equal parts (by weight) of carbonate andbicarbonate to neutralize the acid generated by the oxidation reaction.The results of the initial screening of different oxidants on CS2 spiked soilsare summarized in Table 1.

As may be seen from the initial results, all of the tested oxidants were ableto achieve substantial CS2 removal. Caustic alone was able to achieveremoval of CS2 equivalent to those with peroxide oxidants. However, asnoted above, the reaction end products are bisulfide and sulfide, whichare undesirable.

5.2 HALBY SITE SOILS — SMALL-SCALE STUDIES

A series of batch experiments was carried out on 100-gram samples ofHalby site soils. The approximate average CS2 concentration of untreatedsoils, i.e., 34,000 ppm, was used to calculate the addition of oxidant to soilin experiments in this series. The purpose of the experiments was toexamine the extent of the oxidant reaction with the CS2 and to evaluateheat released: In addition, necessary oxidant doses, preferential oxidationof CS2 versus other organics and required reaction time were examined.The oxidants tested were hydrogen peroxide, sodium percarbonate,sodium perborate, and sodium persulfate. The observations to be made

5-2 WTTCO6/I7/96

OR5G8W.were focused on heat released during oxidaitt addition and the generationof vapors from the reaction mix. The reaction times were approximately30 minutes, which was the duration of addition of the oxidant. Reactionswere stopped at 30 minutes by adding dilute hydrochloric acid, andanalyses were performed within 15 minutes after reaction termination.The reaction between the oxidant and either CS2 or other oxidizablematerials in the soil was apparent upon observation of the reaction mass.The reaction produced effervescence within the soil as the oxidant wasadded. Twenty milliliters of the buffer solutions (equal volumes of 0.1 Nsodium carbonate and sodium bicarbonate) were added to the 100 gramsamples of the Halby soils. In the case of peroxide, 10 ml of 0.1 N sodiumhydroxide was added. The reactions were carried out in 200 ml samplejars with hand agitation. The temperature of the reaction mix wasmeasured with a glass thermometer inserted into the reaction mix.

The theoretical quantities of reactants (expressed as the complete oxidantformula) are summarized in Table 2 (assuming that the averageconcentration of C$2 in the Halby soils is 34,000 mg/kg and an 8:1 ratio ofoxidant to CS2 is theoretically correct).

The results of the experiments using solid oxidants (percarbonate andperborate) are provided in Table 3. The experiment employing 50%hydrogen peroxide was characterized by an excessively vigorousexothermic reaction. Because of this, no data were taken on this sample.Sodium persulfate was also tested under these conditions at an 8:1 molarratio of oxidant to CS2; however no reaction or reduction of CS2 wasobserved.

In addition to these small-scale studies, qualitative testing was performedto evaluate the effect of no buffer addition on the oxidation reactions using -H2C>2 and percarbonate. It was found that both reactions proceeded* ;̂? falthough the H2O2 reaction was less efficient under these conditions.^)

Experiments H-4 and H-5 were carried out to determine whether therewas a relationship between the oxidant added and the residual CS2. Only10% of the theoretical, dose of oxidant was added. CS2 reductionsobserved corresponded roughly to the reduced quantity of oxidant added,suggesting a proportional'relationship.

Initial experiments to measure reaction temperature rise were conductedusing the solid oxidants. These experiments were conducted at roomtemperature. Attached are two graphs (Figures 1 and 2) indicating thetemperature increase with solid oxidant (perborate and percarbonate)addition over time. .The graphs indicate that the equilibrium temperature

5-3 WTTCO-6/17/9S

A R U O I 2 I U

of the mixtures was in the 60-80°C range3, This temperature is the resultof heat generation in the reaction mix and cooling of the reaction vessel bythe air in the laboratory (in the fume hood). The reaction container, a 200-ml sample jar, was not insulated. It appears that in this temperature rangethe heat generation is equal to the heat loss from the uninsulated sides ofthe jar.

Calculations of the heat released in the reaction as measured by thetemperature increase in the first few minutes of the reaction indicated thatthe heat release is about 20 kcal/g-mole of solid oxidant. As the reactionproceeded, the increment of temperature increase per quantity of oxidantadded decreased substantially. In these reactions, some volatiles wereobserved to be emitted from the reaction mixture, particularly in the firstfive minutes of the reaction. Experiments to quantify and identify thenature of volatile emissions are presently being conducted.

The solid reagents reacted readily with contaminants in the soil. Mixingof the solids was not perfect, and the reaction mass was not homogenous'in temperature. Since the powders were added at the top of the soil mass/it was hotter (as much as 10°C higher) there than at the bottom of themass. The reaction mass resembled "mousse" in consistency after addition

Nof the solid reagents.

following these experiments two additional experiments were conductedat constant temperature to see whether there was a significant impact ofvolatilization of the CS2 when the reaction temperature was observed toreach the 60-80°C level. The addition rates were the same as those used in,the previous experiments. For the percarbonate addition, the initialtemperature was 0°C, and after percarbonate addition it increased to 34°C,where it was maintained until all of the reagent was added. There werefew or no CS2 vapors emitted from the reaction. The perborate reactionwas much slower, and the reaction temperature did not increase to above5°C over the 30-minute oxidant addition period. It was not until thesample was allowed to warm to room temperature (approximately 17°C)after the 30-minute oxidant addition period, that the sample with sodiumperborate started to react. For all oxidants, a 30-minute oxidant additiontime was used to allow for comparison of these results with those of theprevious experiments.

Although this temperature is above the boiling point for C$2 (46°C), and well abovethe flash point, there were no signs that the reaction mix had combustedspontaneously.

5-4 WTTCO-6/17/96

A R U O I 2 I 5

OfflSffWLFor the reactions carried out at lower temperature/ the percarbonateappeared to react readily. The reaction was much less vigorous, withlesser heat generation than for peroxide and a lower equilibriumtemperature in the reaction mass. No volatilization of any consequencewas observed for these reactions.

5.3 HALBY SITE SOILS — LARGE-SCALE STUDIES

Several large-scale studies were carried out to provide adequate samplefor analysis by a certified laboratory using EPA approved procedures.The purpose of these analyses was to formally document the resultsobtained in the experimental program. A secondary objective of theexperiments was to evaluate reaction effectiveness (conversion of CS2 bythe various oxidants) so that field treatment could be fine-tuned to the

. existing"concentrations.. One special experiment was also made to clarifythe potential for byproduct formation during the oxidation process.

The format of the experiments was as follows: 500 grams of Halby site soilwere placed in a 2-liter resin kettle. An aliquot of the same soil wassampled at the time the resin kettle was charged in order to establish abackground (untreated) concentration of CS2- Experiments were carriedout with each of the oxidants successfully screened so fan sodiumpercarbonate, sodium perborate, and hydrogen peroxide. Thepercarbonate and perborate were added directly as solids to the Halbysoils. The hydrogen peroxide was added as a 25% solution to minimizelocal overheating of the reaction mix. One hundred ml of buffer solutioncomposed of 50 ml each of 0.1 N sodium carbonate and sodiumbicarbonate were added to the reaction mix prior to the addition ofpercarbonate or perborate. One hundred ml of carbonate/bicarbonatebuffer were used for the hydrogen peroxide trial: fifty ml of carbonate (0.1N) were added to the Halby soils prior to the addition of hydrogenperoxide, and fifty ml of bicarbonate buffer solution (0.1 N) weresubsequently added to make a 25% peroxide solution (from 50% stock).The individual experiments with the oxidants were carried out at themolar ratios provided in Table 4 of oxidant to CS2 (assuming an initialCS2 concentration of 34,000 ppm).

Note that the stoichiometric ratio of oxidant to CS2 which is needed tooxidize all CS2 is 8 moles of oxidant (as equivalent hydrogen peroxide) toeach mole of CS2- The experiment using an 8:1 molar ratio of hydrogenperoxide was not considered practical as a soil treatment method for soilquantities with this high of an average CS2 content. This is because of thesignificant heat generation of the reaction. A lower molar ratio (4:1) wasused rather than additional dilution, which causes the reaction mix tobecome a thin slurry,

5-5 WnCO-6/17/96

A R t » O I 2 f 6

The tabulated molar ratios indicated above were based on initial CS2concentrations of 34,000 mg/kg. In practice, laboratory analysis of eachsample tested indicated significantly lower initial levels of CS2- Theresults presented on Figure 3 reflect the actual molar ratios of oxidantsadded to the levels of CS2 based on CLP method analysis.

The reaction mass was continuously cooled and agitated throughout theaddition of oxidant. Cooling was achieved through the use of an externalice bath. The temperature of the kettle could be lowered by submergingthe bottom of the kettle to remove the heat generated by the reaction. Thereaction mass was kept below 25°C throughout the experiments to .minimize the loss of CS2 through volatilization. Temperature wasmonitored continuously by hand. Agitation of the reaction mix wascarried out by manually stirring the reaction mix.

All of these samples were analyzed by Envirotech Research using EPA-approved procedures. The samples were analyzed for carbon disulfide byEPA procedure No. SOWOLM01.8. Analyses were also made forchemical oxygen demand (COD) and pH. All samples were shippedunder chain of custody on ice in a sealed container to maintain sampleintegrity.

The data from experiments H-6 through H-21 are provided in Table 5.Again, the actual untreated CS2 concentrations are used in calculating thepercent reductions in CS2 rather than the previous assumption of 34,000ppm CS2. These calculations are provided in Table 6.

Reaction conditions were similar to those for the small-scale trials, exceptthat the oxidant was added over an 8-hour period to avoid a high heatgeneration rate. Although previous small-scale studies indicated that noexternal buffer solution was required for the percarbonate solid oxidant,aqueous buffer solutions were employed in these large-scale trials to aidin heat dissipation and to maintain consistent reaction conditions amongoxidants.

The data indicate that 90+ % conversion (residual concentrations of CS2 inthe range of 400 mg/kg, based on raw soil levels) can be reached bypercarbonate and perborate oxidation at a molar ratio of oxidant to CS2 of8-ll:l. This is at or near the projected stoichiometric ratio of 8 moles ofoxidant to one mole of CS2. Levels of 200 mg/kg or lower can beobtained at molar ratios of 15-17:1. Using the relationship identified inFigure 3, the quantity of oxidant can be adjusted to achieve a desiredresidual concentration of CS2 ~

5-6 - WTTCO-6/I7/96

Because of potential safety and cost concerns with caustic use for pHadjustment in the H2O2 trials, the 0.1 N carbonate/bicarbonate bufferused with solid oxidants was employed. These trials were not assuccessful as those for other oxidants tested. Further trials are in progressusing caustic at a 1:1 molar equivalent of CS2 to caustic.

5-7 WITCO-6/17/96

f l R l * O I 2 ! 8

6.0 DISCUSSION OF RESULTS TO DATE

6.1

The results obtained indicate that all of the oxidants (except persulfate)were effective in oxidizing tine C$2 in the spiked soils. It was apparentfrom the data that the oxidation reaction occurred readily with all of theoxidants used except persulfate. The reaction between hydrogen peroxideand CS2 at concentrations of CS2 greater than about 1,000 ppm generatessignificant heat (through oxidation of both sulfur and carbon).

6.2 HALBY SITE SOIL

The small- and large-scale experiments with Halby Site soils provided,results similar to those obtained with spiked non-Halby soils: generationof an estimated 20 kcal (of excess heat)/mole of €82 oxidized, and ademonstration of a general correlation between oxidant quantity added

• \ and CS2 quantity removed. The reaction variable having the most impactL o • ' • on the oxidation efficiency is pH, which should be kept at 8.5 or higher for

best results. Sodium persulfate was ineffective as an oxidant.

For large-scale experiments, sixteen-times-longer (8 hour-) oxidation timeswere required to avoid excessive heat generation. This factor will need, tobe considered in full-scale process economics. In addition, hydrogenperoxide oxidation using alkaline buffers instead of caustic was not aseffective as other oxidants tested, likely because of the low resultantoxidation pH. Caustic-hydrogen peroxide oxidations will be evaluatedfurther in the fourth phase of testing presently in progress.

ERM,INC . . . . . . - 6-1 . WTTCO31024.04.01-6/17/96

M U O I 2 I 9

7.0 CONCLUSIONS TO DATE

The results of the treatability study experiments show that CS2 can beremoved from Halby soils by treatment with any of the oxidants (exceptpersulfate) evaluated in this program.

The conversion of CS2 is dependent on the quantity of oxidant used,expressed as the molar ratio of oxidant to CS2- The relationship is shownin Figure 3, attached. It is noted that, in certain samples (samples H-6, H-12, H-16, and H-18), the COD after treatment increased by approximately16,000 mg/kg. This would indicate that COD-refractory compounds inthe soil are being oxidized to COD-amenable compounds. For the H2O2trials, the low pH may have caused preferential oxidation of organicsother than CS2. The two remaining elevated COD samples were runusing much less than stoichiometric quantities of oxidant; however, thisshould not have resulted in an increased degree of organics oxidation toCOD-amenable materials. Thus, the rise in COD observed in the H2O2trials cannot be explained.

ERM.INC 7-1 WTTCCXJ1024.04.01-6/17/96

A R U O I 2 2 0

Figure 1Reaction Mix Temperature Increase with

Sodium Percarbonate Addition

M »5 z8 *o •

MS 0

ororo

Reaction Temperature

Cumulative Grams of Percarbonate Added

15Time, mlns.

> i• >« Zs*9 •==p o

O

rororo

Figure 2Reaction Mix Temperature Increase with

Sodium Perborate Addition

Temperature of Reaction Mix

Cumulative Grams of Perborate Added

Time, mlns.

o.55i -.XO

Figure 3Conversion of Carbon Disulfide Versus Mole Ratio of Oxidant

to Carbon Disulfide

O

roroCJ

6 8 10 12

Molar Ratio, Oxidant to Carbon Dfsulfkte14 16 18

Qffi&KAL

Table 1 Non-Hctlby Spiked Soils Reacted at Varying Oxidant: CS2 MolarRatios

Sample Size<S>

50

50

SO

100

100

100

100

100

InitialCarbon

DisulfideCone,

(mg/kg)

1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,000

Oxidant

None

H202(8:l)

H2P2 (24:1)

None

H202(8:l)

Perborate(8:1)

Percarbonate(8:1)

None

Causticand/orBuffer

NaOH

NaOH

NaOH

50/50 Carb.-Bicarb.

50/50 Carb.-Bicarb.

50/50 Carb.-Bicarb,

50/50 Carb.-Bicarb.

None

Un reactedCarbon

Disulfide,mg/kg

150

*

*

503

157

144

213

918

PercentReduction

85

NA

NA

49.7

84.3

85.6

78.7

8.2

All experiments carried out at room temperatureSoil samples allowed to react for 24 hours prior to analysis* assumed negligible, based on substantial observed volatilization of CS2 from the sample and vigorousreaction upon addition of peroxide. No CS2 peaks were observed in the analysis of the treated samples.

Table 2 Small-Scale Halby Site Soils Experimental Matrix

Grams HalbySoil

100

100

100

VolumeBuffer, ml

20

20

10mlO.lNNaOH

GramsPercarbonate

38.7

GramsPeroxide

25.4

GramsPerborate

56.9

,,.^A ^.500 '^/jkCSv ;•-' ^ *J 'i

A R I » O I 2 2 5

ORlfflWUL

TableS Small-Scale Halby Soil Samples Reacted Under ControlledTemperature (£ 25°C) Conditions

Sample ID

Untreated

H-2

H-3

H-4

H-5

•H-6

H-7.

Buffer

Carb/Blcarb

Carb/ Bicarb

Carb/ Bicarb

Carb/Bicarb

Carb /Bicarb

Carb/Bicarb

Carb/Bicarb

Oxidant

none

Percarbonate

Perborate

Percarbonate

Perborate

Percarbonate

Perborate

RatioOxidanfcCSa

NA

8:1

8:1

0.8:1

0.8:1

8:1

8:1

%CS2

Reduction

5%

not measured*

not measured*

approx. 10%

approx. 11%

99.9%

80%

These experiments were made to check temperature rise and heat generation. They are not consideredreliable for calculating C$2 reductions.

A R U O I 2 2 6

Large-Scale Halby Soil Study Molar Ratios

Hydrogen Peroxide

2:1

4:1

__

Sodium Percarbonate

2:1

4:1

8:1— ,.. ...

Sodium Perborate

2:1

4:1

8:1.

A R U O I 2 2 7

Table 5 Performance Data from Treatment of Halby Site Soils

Sample ID

H-6

H-7

H-8

H-9

H-10

H-ll

H-12

H-13

H-14

H-15

H-16

i» H-17

4T H-18O— H-19ro£9 H-20CO

H-21

Sample Make-up

Untreated Halby Sail/Buffer (carb/bicarb) Solution

Halby Soil/Buffer Solution /Percarbonate

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solurlon/Percarbonate

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solution /Percarbonate

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solution /Perborate

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solution/ Perborate

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solution/HzOa (25%)

Untreated Haiby Soil/Buffer Solution

Halby Soil/Buffer Solution/H2O2 (25%)

Untreated Halby Soil/Buffer Solution

Halby Soil/Buffer Solution /Perborate

OxIdaul/CSjMolar Ratio

Nome

4.0

None

8.0

None

17.0

None

2.88

None

11.3 ,

None

6.8

None

15.28

None

15.11

CS2 (p'p'tti)

17,000

5,000

17,000

440

16,000

4.1

23,000

6,600

12,000

400

10,000

3,800

8,900

1,200

18,000

170

pH

9.02

10.03

8.90

10,06

8.86

10.26

9.10

9.96

9.32

10.32

9.58

7.33

9.67

6.71

9.40

10.47

COD (ppm)

33,400

46,500

34,500

30,500

24,600

25,000

38,000

54,100

33,200

28,200

27,600

43,300

30,800

46,400

34,900

29,900

Reduction

——

70

—

97

—

99.9

—

71

—

96

—

62

' 86.52

99.06

Table 6 _ Calculations for Percent Reduction in .C$2 Concentrations

Oxidant

Percarbonate

Perborate

H2O2

ExperimentNumber

#6

#7

#8

#9

#10

#11

#12

#13

#14

#15

#20

#21

#16

#17

#18 :

#19

Actual Stoichiometric Ratio'3'

34,0000.25 x ———— = 0.50

17,000

34,000.0.5 x ———— =1.0

17,000

34,000l .UA — L.it

- 16,000

34,0000.25 A ""' '" -i 0.36

23,000

34,0000.5 x ———— , * 1.41

12.000

34,000l.U A — I. GO

18,000 - . ,

34,000\J.LJ x — u.aj

10,000

34,000 "05 x " ——— = 1.91

8.900

ActualMolar Ratio

4.0

8.0

17.00

2.88

11.3

15.11

6.8

15.28 .

% CS2

Reduction

70

97

99.9+

71

96

99.06

62

86.52

(a) assumed Stoichiometric ratio xexpected CS2 cone, in soil

actual CS2 cone, in soil—---actual Stoichiometric ratio

Attachment ACarbon Disulfide SampleAnalysis

OWfflNW,•r?^<*

CARBON BISULFIDE SAMPLE ANALYSIS

A.1 INTRODUCTION

An open loop isocratic HPLC analytical system was used to screen soilsamples and determine the amount of CS2 in spiked, treated, anduntreated small-scale soil samples.

A.2 HPLC SYSTEM

A.2.1 CONFIGURATION

The HPLC system is composed of the following components:Spectraphysics 8815 Isocratic High Pressure Pump

Rhedyne 7125 Injector Valve with 10 |il loop

Timberline Column Oven at 30.0°C

Jones 25 cm x 4.6 cm C18 (5 |im) separation columnLinear Model 200 UV/VIS Detector

Spectraphysics SP4400 IntegratorThermo Separation Products Winner on Windows Data SoftwarePackage

Computer (386SX 25 Mhz CPU) System using Windows 3.1,WordPerfect for Window 5.1, Excel 4.0 for Windows and Quick Basic

Acrodisc CR PTFE (0.2 |im) Syringe FilterCole-Partner Model 8891 Ultrasonic Mixer

Ohaus Analytical Plus Balance

Laboratory Glassware, Reagents and Supplies

100.0 ml volumetric flasks (Class A)

Volumetric pipets (Class A)

Pipet bulbs100 |il gas tight Hamilton Syringe

HPLC grade acetone (Baker Analyzed Solvent)

ERM,INC . A-l WTTCp-31024.04.Ql-6/17/96

CWGflWLHPLC grade methanol (Baker Analyzed Solvent)

HPLC grade water (Baker Analyzed Solvent)

A.2.2 STANDARDS AND SAMPLE PREPARATION

The CS2 standards were prepared by diluting a known amount of reagent-grade CS2 with HPLC-grade methanol in a 100-ml volumetric flask. A 50-ml volumetric pipet was then used to transfer an aliquot from the firstflask to another 100-ml volumetric flask, which was filled to volume withHPLC-grade methanol. This method was repeated twice more and usedto generate a four-point curve from 1,000 ppm to 625 ppm.

Soil samples were processed for analysis using the following procedure.1. Tare a 40 ml VQA vial on analytical scale.

2. Weigh out sample placed directly in the VOA vial.

3. Add 1 N HC1 to yield a pH <2 (approx. 1 g).4. Weigh sample on analytical scale (to determine the quantity of acid

added) and recrod weight in logbook.

5. Dilute sample to volume with HPLC-grade methanol (with no headspace). Record weight of result.

6. Let stand for 24 hours.7. Draw approximately 3 mis of sample into syringe.

8. Remove needle and attach filter (Acrodic CR PTFE (0.2 |im)).9. Filter sample into small vial and mark vial with sample number.

MOBILE PHASE PREPARATION

The HPLC mobile phase was prepared to produce a solution of 82%volume methanol and 18% of water. The solution was prepared bymixing 2,000 ml of HPLC-grade methanol with 439 ml of HPLC-gradewater.1. Depending on the amount of mobile phase being mixed, measure out

220 ml of HPLC H2O for every 1,000 ml of HPLC methanol.2. Degas new mobile phase with helium for 5 minutes and keep a slight

overpressure of helium during HPLC operation.

wrrco-6/i7/96

/ W O I 2 3 2

A.3.0

A.3.1 CALIBRATION CURVE

Four calibration points were used. Each point was injected two times.Before each analysis, a sample of methanol spiked with a known amountof CS2 was injected on the HPLC to check the accuracy of the instrument.Spike recovery was consistently greater than 70%, which was deemedadequate for screening purposes.

A.3.2 HPLC ANALYSIS PROCEDURE

Liquid Chromatographic ParametersColumn Jones 25 cm x 4.6 CIS (5 um particle size)

\ ' "Mobile Phase 82% Methanol -18% Water

Flow Rate 1.0 mls/min

Detector Sensitivity 0.1 AUFS '

Volume Injected 60|ll

Wavelength 315 ran

Analytical Procedure

The standard solutions were injected a sufficient number of times toensure reproducibility. Then each sample was injected. The concentrationof the analytes in the sample was calculated by comparing the peak area ofthe sample to that of the standards. Thermo Separation Products' Winneron Windows software was used to electronically reduce the rawintegration data. .

ERM,INC. . . A-3 ' WTTCO-3102404.01-6/17/96

. A R l + 0 1 2 3 3