Embed Size (px)
Citation preview
~.. .... ...-
TECHNICAL REPORT ARLCD-TR-80012
CHARCOAL REGENERATION -PART 11.
MODIFIED CARBON SURFACE ACTIVITY AND
REVERSIBILITY OF TNT ADSORPTION
J. HABERMANT. CASTORINA
S. SEMELH. KRAMER
JULY 1980
US ARMY ARMAMENT RESEARCH AND DEVELOPMENT COMMAND/LARGE CAUIBER
~I~I WEAPON SYSTEMS LABORATORYDOVER. NEW JERSEY
APPROVED FOR PUBLIC RELEASE: DISTRIBUTION UNLIMITED.
C=)1Cl
The views, opinions, and/or findings containedin this report are those of the author(s) andshould not be construed as an official Depart-ment of the Army position, policy or decision,unless so designated by other documentation.
Destroy this report when no longer needed. Donot return to the originator.
UNCLASSIFIlED TI AE(~mbt ne~! ________________
SECURITY CLASSIFICATION OFTHSPG foDaettr0
REPORT DOCMAENTATION PAGE BEFORE__COMPLETINGFORM
I. REPORT NUMBER ,12. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
Technical Report ARLCD-TR-80012 - ?72 (4. TITLE (And Sbt~to) S YEO EOT&PRO OEE
CHARCOAL REGENERATION - PART II.MODIFIED CARBON SURFACE ACTIVITY AND REVERSIBILIT _______________
(J1~TNTADSRPTONG. PERFORMING ONG. REPORT NUMBER
7.AUTHORgE) J. Haera 4. CONTRACT OR GRANT NUNSER(s)
T. Castorina 3627-01-001S. SemelH. KraMer
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT, TASK(ARRADOM, CWSLAREA A WORK UNIT NUMBERS
Energetic Materials Division (DRDAR-LCE) 1L,762720D048 S-10Dover. NJ 07801
It. CONTROLLING OFFICE NAME AND ADORESS III. REPORT OATE
ARRADCOM, TSD July 1980STINFO (DRDAR-TSS) IS. NUMBER OF PAGES
Dove~ NJ 780163I&PIMONSTO AINO AGENC NAt AODRIESS( d1iff~amt kern CtxuftoIins Oilft) 1S, SECURITY CLASS. (*I Lv. t*PWit
*Chemical Systems Laboratory UnclassifiedATTN: DRDAR-CLT-P _____________
Aberdeen Proving Ground, MD 21010 IS. LVIC'1NOONNOM
IS. DISTRIBUTION STATEMtNT (of thl* Rapwt)
Approved for public release; distribution unlimited.
17. OasyasouTtON &TATEWMNT (~44.th *Akk*O. 004"4 to 4844 A&A@ 4 1 UebAV RepeN)
III. URNIT.EEMANY NOTES
This study was funded under the Environmental Qktality R&D program adminis-tared by the DARL()4 Lead Laboratory for PAIILI, Aberdeen Proving Ground, MD0.
to. Rev 6OD C,.u ,e ~eUcw E4~I~ y us"tNWActivated carbon Water adsorption/desorpt ion Solvent rogeneradon ofPretroatuIent Pore size distribution activated carbonBatch cycling Adsorption/desorption cycling PorosityColumn cycling Adsorbed INT Rate of degradation ofChemical regeneration Isotherms carbon
ft ANIYRACT (00up 06 9~ ==O8 *ad 11~00 &gP 44 .Mbff
The TNT adsorption/solvent desorption cycling of activated carbon (PS300) hasbeen studied as a function of various pretreatments. Ithe preferential pluggingof pores by the pretreatments did not enhance the regenerative property of thePS300. In conjunction with water adsorption/desorption isotherms, the character-istics of a carbon with an improved cycling property has been elucidated. Theimproved property Is due to a shift to larger pore volume and a latger ratio ofpore opening to pore body diameters. Chemical regeneration of' activated carbon1containin irreversibly adsorbed 7WT has been sh oaNegbe
"Itt1J0Y CI.MMVATfON Or I~ pa" (obw ;ua 4w*00.
ACKNOWLEDGMENT
Professor F.J. Micale of Lehigh University, Bethlehem, PA iscited fork the"significant contributions he has made under contractDAAG29-76-u4 0100 on the water adsorption measurement, and the manyhelpful Ais ussions dealing with the interpretation- of thesemeasure76nts as related to the cycling of activated carbon.
NI'S GRiAaj
IDDC TA-B
JUat fif~ati±on
Avallabilit CodesDt IAvall anld/o)ZM a , P GC 41
TABLE OF CONTENTS
* - Page No.
Introduction 1
Experimental 1
Pretreatments 2
Water Adsorption Isotherms 3
Ultraviolet Analysis of Aqueous Solutions 3
Column Cycling (Adsorption-Desorption) 3
Batch Cycling (Adsorption-Desorption) 4
Chemical Regeneration 5
Results and Discussion 6
TNT Cycling 6
Water Adoorption/Desorption Isotherms 10
Desorption 12
Chemical Regeneration 12
Conclusions 13
References 14
Appendix 34
Distribution List 55
TABLES
Page No.
I Surface areas of virgin and pretreated carbons. 15
. 2 Rates of degradation of pretreated carbons. 16
3 Integrated areas under degradation curves. 17
4 Change in porosity of pretreated and other 18activated carbons compared to 40/80 FS300.
5 Column adsorption/desorption cycling. 19
6 Incremental column desorption for one cycle. 19
7 Effect of chemical regeneration on surface 20area of FS300.
8 Batch cycling of 40/80 mesh FS300/Argon/ 214 hr /1000Co
I - 9 Batch cycling of 40/80 mesh FS300/CH4 / 22600OC/4 hr
!+.
It
• '7
FIGURES
Page No.
1 Specific adsorption deactivation as a 23function of adsorption/desorption cycling.
2 Linear plots of specific adsorption deactivation 24as a function of adsorption/desorption cycling,according to a second order kinetic model.
3 Water adsorption/desorption on 40/80 FS300 25Sand hydrogen treated 40/80 FS300 at 10000C
for 4 hr.
4 Water adsorption/desorption on 40/80 FS300 26and 40/80 Witco 337.
5 Water adsorption/desorption on 40/80 FS300 27and 40/80 FS400.
6 Water adsorption/desorption on 40/80 FS300 28and 40/80 FS300 treated with NO2 for 72 hr.
7 Water adsorption/desorption on 40/80 FS300 29and TNT saturated-desorbed 40/80 FS300.
8 Change in body pore size distribution 30relative to 40/80 FS300.
9 Change in opening pore size distribution 31relative to 40/80 FS300.
10 Pore size distribution of the pore openings 32relative to the pore bodies for FS300,
OFS3O/H /1000C, Witco 337, and FS300/CH4/700OC/5 hr.
11 Pore size distribution of the pore openings 33relative to the pore bodies for FS300, FS400,FS300/4 hr/NO and PS300/72 hr/NO2.
I•
INTRODUCTION
The use of granular activated carbon (GAC)l for the treatmentof pink wastewaters at load and assembly Army Ammunition Plants hasbeen established as an effective method of pollution abatement.However, its major drawback is the need to incinerate the spentcarbon after the first breakthrough of RDX/TNT. The costeffectiveness of using GAC would be increased greatly by therepeated regeneration of the spent carbon in situ with solventelution. In addition, the solvent can be recovered economically byvacuum distillation and the eluted TNT/RDX reused.
Accordingly, the regeneration of GAC by solvent elution wasinvestigated and initially confined to TNT as the adsorbate(ref 1). It was concluded from this preliminary investigation thatthere is a loss of capacity of the carbon surface for TNT adsorptionwith each successive cycling by solvent (acetone) elution. Thisloss of capacity is due to the active nature of the carbon surfaceand/or the pore size distribution, leading to irreversibleadsorption of TNT.
The TNT adsorbate molecules are attracted first to reginns ofthe surface which have the highest energy. These regions areassociated with various functional groups and high energy sites withsurface carbon atoms having incomplete coordination numbers. Astudy of the relationship of energy site and/or pore sizedistributions representing the second phase of the investigation ofthe regeneration of carbon by solvent elution is the subject of thisreport.
EXPERIMENTAL
The activated carbon used in this study was FS300 Obtained fromCalgon Corporation, Pittsburg, PA. The carbon was ground to a 40/80mesh, washed thoroughly, and dried. All pretrested samples wereobtained from a master batch prepared in this way. Other GAC1s usedwere FS400 from Calgon Corporation and Wttco 337 (now designated asGrade 965) from Witco Corporation, New York, NY. These GAC's wereat-o ground to a 40/80 mesh for comparison purposes..
IThe nomenclature "activated charcoal" is synonymous with activatedcatbon.
I ,
S.1
* -- '
Pretreatments
Gas Treatments
Methane, hydrogen, and argon were used to pretreat FS300.The carbon was placed in a quartz tube (50 cm by 1.6 cm) and theproper gas was allowed to flow through the tube, at a flow rate of2-3 mL/min. After the air had been displaced, the tube was heatedin a tube furnace to the previously calibrated temperature, andallowed to remain at this temperature for the requisite number ofhours 2 . Cooling was carried out by allowing the gas to flow untilroom temperature was reached. The methane used was a 10% mixture ofmethane in 90% argon. Hydrogen and argon were high purity gases.
NO2 and NO Treatments
The NO and NO2 used were CP quality. The NO and NO2 treatedcarbons were prepared by allowing the vapors to displace the air instoppered flasks and allowed to remain in contact with the vapors atambient temperature for the requisite time. Precautions were takento absorb the excess vapors in caustic solution when preparing thesamples.
Some preparations were water-washed to remove the unreactedNO2 or NO. Others were washed with freon and/or acetone to removethe unreacted NO2 or NO. All preparations were thoroughly dried ina vacuum desiccator prior to use.
Silanization
Five srams of FS300 (40/80) tas heated overnight at 150*C.The carbon was placed in a one liter flask and covered with 75 mL ofmethyl cyclohexane. Ten oL of triwethylchlorosilane was added andthe mixture ref luxed under an air condenser for 5 hours. Thesolution was decanted, and 50 mL of n-propanol added to the carbonand refluxed for another hour. The alcohol was decanted and thecarbon washed with n-propanol followed by oven drying at 110%covernight.
Stearic Acid
Five grams of PS300 (40/80) was covered with 25 mL of 0.1%stearic acid in 952 ethanol in a beaker and allowed to stand over-night. The solution was decanted and the carbon allowed to air-dry.
f1ee Tabl~e 1
2
Water Adsorption Isotherms
Water adsorption isotherms were determined gravimetrically on aquartz spring balance (Worden ,'-;rtz Products, Inc) which had anabsolute sensitivity of 7 microgi'ps per gram of sample. The watervapor pressures were measured with a 100 mm Barocol (Datametrics,Inc) capacitive differential manometer sensor and associatedelectronics enabling accurate readings to better than 103 mm.
Ultraviolet Analysis of Aqueous Solutions
A Beckman DU spectrophotometer equipped with one centimetermatched quartz cells was used for the analysis of aqueous solutionsof TNT. A stock solution of TNT (100 ppm) was prepared bydissolving 100 mg of recrystallized TNT in one liter of distilledwater. Aliquots of the stock solution were taken to prepare I to 10ppm inclusively of standard TNT solutions. The absorbances of thesestandard solutions were measured on the spectrophotometer at 230 nmagainst a distilled water blank. The absorbances were plotted vsconcentration giving a linear plot. Unknown solutions were dilutedto a concentration in the range of I to 10 ppm. The absorbanceswere determined and their appropriate dilution factor applied to
give the concentration.
Column Cycling (Adsorption-Desorption)
Critical column variables were determined for the column used inthe column recycling part of this investigation. By taking intoaccount various particle sizes and weights of carbon a design was
developed which was suitable. This design allows a wide tolerancefor the pressure head over the column.
A 34 x 106 cm glass column with 19/22 joints at both ends wasplugged Just above a joint at one end with glass wool. A "U" shapedsiphon outlet was placed at this bottom end of the column and thetop end fitted with a one liter funnel to hold stock TNT solution.The funnel was equipped with an overflow tube to allow excesssolution to run back into the stock carboy (a 5-gallon polyethylenebottle). The solution was pumped to the funnel with a peristalticpump timed to be activated 50% of the time, so as to maintain aconstant head pressure (A v - 60 ml).
Eight grams of GAC was accurately weighed into a 250 mL beakerand slurried with ca 100 mL water. The slurry was poured into thefunnel and allowed to settle in the column with the aid of gentletapping and a stream of water where necessary.
3
The prepared column was clamped in the center of an automatic"20 position fraction collector (Buchler Instruments, Fort Lee,NJ). The fraction collector was equipped with an activator toperiodically advance an adjustable glass funnel so as to collect theeluent in one liter polyethylene bottles.
On column cycling was carried out by monitoring the effluentfor TNT concentration from the columns. The breakthrough point wasstandardized at an effluent concentration of 1 ppm. Loading thecolumn with TNT was terminated as soon as sufficient data wasgenerated to determine this point. The amount of TNT absorbed at 1ppm effluent concentration was considered as the capacity of thecarbon for comparison purposes. The specific adsorption of thecolumn was calculated by dividing the total amount of TNT adsorbedby the weight of the carbon in the column.
After breakthrough, the column was desorbed by adding 100 mLincrements of solvent, starting with 20% acetone/water in incrementsof 20% up to 100% acetone until one liter of eluent was collected.The column was then washed with distilled water until all acetonewas washed out (several pnrtions of 100 mL each). Afterwater-washing, the column was ready for adsorption. The acetoneextract was reserved for analysis.
The amount of TNT desorbed was determined by transferring theeluate to a one liter volumetric flask and diluting to the mark withacetone. A 10 mL aliquot was transferred to a 500 mL volumetricflask and diluted to the mark with water. A 10 mL aliquot wastransferred to a 250 mL beaker, diluted to 100 mL with water, heatedon a hot plate to evaporate ca 50 mL and insure the removal of theacetone. The remaining solution was transferred to a 100 mLvolumetric flAsk,diluted to mark and the concentration determined byUV spectroscopy. The amount of TNT desorbed was calculated as"followst
Amt. TNT " 1000 500 10 X - 0.SxC---W M x =
whe re:
C - concentration determined in ppm.
Batch Cycling (Adsorptlon-Desorptlon)
One hundred mg o0 GAC was accurately weighed into a Rlaonsstoppered I L Erlenmovei flask and 500 m. of a 140 ppm TNT solutionadded. This ratio of carbon to volume of stock TNT solution wasalways maintained to insure adsorption at the plateau region of the
4"..
I. .j
adsorption isotherm. The charged flask was placed on a wrist-actionshaker and shaken for 24 hours to assure equilibrium. The amountadsorbed by the carbon was determined by withdrawing an aliquot ofsupernatant solution from the flask and determining theconcentration of TNT by UV spectroscopy. The concentration, C1was subtracted from the initial concentration, Co , of stock TNTsolution and the specific adsorption calculated by the following:
C0 - CSpecific adsorption
2m
where:
m weight of carbon, mg2 - correction factor for the volume of TNT solution used.
The TNT solution was carefully decanted from the carbon andrinsed once with a small quantity of water. Fifty mL of C.P.acetone was added to the flask and the stoppered flask was shakenfor one hour. The acetone was decanted into a 250 mL volumetricflask. The process was repeated two more times with the washingbeing added to the volumetric flask. After dilution to the mark, a5 mL aliquot was pipetted into a 250 mL beaker containingapproximately 75 mL of water, and brought to a gentle boil on a hotplate to evaporate about 50 mL. This process expelled the acetonegiving a residual aqueous solution of TNT which was diluted tovolume (100 mL) in a volumetric. flask. The concentration wasdetermined by UV spectroscopy and the specific desorption calculatedas follows:
Specific desorption-- x --2x - 55 1000 m
where:
.m weight of carbon (100 mg)C w concentration of TNT, ppm
The entire operation was repeated for the next cycle ustin8 theeluted carbon remaining in the flask.
"Details of column cycling and batch cycling are presented in
the Appendix.
Chemical Regeneration
The following procedures were carried out on TNT saturated.S300 carbon from which the reversibly adsorbed TNT had been
•.S
$I(.4:
extracted with acetone, and therefore, contained only irreversiblyadsorbed TNT.
Alcoholic KOH: One gram of carbon was reacted at ambienttemperature with 50 mL of 10% alcoholic KOH to produce a red color.The mixture was heated to 50 0C, acidified with sulfuric acid anddiluted to ca 200 mL with water.
Nitric Acid: Two gram of carbon was covered with concentratednitric acid and heated gently fir one hour. The mixture was allowedto cool and then washed thoroughly with water.
Ozonation: One gram of carbon was covered with water containedin a tube into which a diffuser was placed. Generated ozone wasbubbled in for a period of two hours.
Oxalic Acid: One gram of carbon was convered with 100 mL of 5%oxalic acid and heated to boiling for 4 hours. The mixture wasallowed to cool and stand overnight.
Reduction Procedures: One gram of carbon was reacted in a1% sodium bicarbonate 1:1 ethanol-water solut.ion with excess sodiumdithionite.
One gram of carbon was mixed with ca 5 grams of zincamalgam and covered with 5% HIC. After 15 minutes, 100 mL of waterwas added and the carbon separated from the zinc amalgam byflotation.
3. One gram portions of carbon were reacted with I g ironin 100 cm3 of 30% lCI; 1 g tin in 100 mL 30% HCk, and I g N114 SH in100 mL of 5Z HCk.
A one gram samp e of carbon was refluxed for one hourwith 5 g KC12 # 6H20 and 2 cm concentrated lCI.
In all cases after the reacrion was terminated, the carbon was"separated from the reaction mixture, washed thoroughly and dried at1106 C.
RESULTS AND DISCUSSION
TNT Cycling
Activated cArbon shows chemical reactivity due to the presenceat the surfac%.. of various chemical functionalities (ref 2). Themicroporous structure contains constrictions narrow enough to hinder
6
_'7 W,
C7
the f ree passage of the TNT molecules and thereby renders some ofthe surface unavailable for its adsorption. Additionally, TNT mayadsorb at constricted pore openings where the potential force fieldis sufficiently high to prevent desorption and where chemicalreactivity may be promoted due to the long residence time of theadsorbate. Physically, adsorbed TNT is readily removed by acetoneextraction; however, a portion of the TNT is irreversibly adsorbedLeading to degradation of the adsorptive capacity. This capacitydiminishes with successive cycling as the irreversibly adsorbed TNTaccumulates. To obtain activated carbon which may be cycled withouta significant loss of capacity, we must consider the chemicalreactivity of the surface, the pore-volume distribution, and thesurface area-pore volume relationship. That these factors areinterrelated is evident in that reactivity leads to the closing of
a pores and loss of surface area.
The initial approach to this problem has been to pacify thesurface by reacting the high energy sites with molecules which wouldthen permit subsequently adsorbed TNT molecules to be easilydesorbed. To evaluate the cycling characteristics of various GACt s,it was necessary to establish an experimental method which wouldpermi~t adsorption/ des orpt ion cycling as rapidly as possible. Forthis purpose, we have designed a one point adsorption/desorptionexperiment; choosing the conditions so that the adsorption alwayoccurs at the plateau of the previously developed isotherm. Elutionis maximized by allowing an excess of solvent (acetone) to contactthe adsorbent-adsorbate complex for three periods of at least onehour each. This treatment was found to remove all the reversiblyadsorbed TNT. (Soxhlet extraction did not remove significantly moreTNT). Typically, carbons were batch cycled for six cycles to give areasonable estimate of the behavior of the carbon insofar asdegradation was cc-neerned, A few carbons were cycled for up to 18cycles (tables 8 and 9). The agreement between six cycles and manycycles is fairly close; hence, an extrapolation from 6 cycle resultscan be done with some confidence. Bly using this technique, a givenpretreated sample could be evaluated in 1 to 2 weeks depending onthe num\,er of cycles required. Promising candidates were thenBelecttod for pilot columns study.
Grano~lar activated carbons are manufactured in various meshsizes and comparing these various carbons would involve differencesdue to diffusion. A 40/80 mesh size obtained by grinding wastherefore utilized as the standard tesning size so as to allowcomparisons from one carbon to another * This size allowed an
3The periphery of a GAC particle may exhibit different propertiesfrom the central portion, therefore, any final testing must becatried out on the original particle size.
* 7
---- OFF O
increase in the loading and sharpened the breakthrough point of thecolumns. Testing in columns was carned out on a continuousadsorption basis to approach plant conditions as closely aspossible. The specific adsorption of the columns, for the purposeof comparison, was calculated by dividing the amount of adsorbed TNTby the weight of the carbon used in the column, even though the TNTis not adsorbed uniformly over the length of the column.
Table 1 shows the effects of the various pretreatments upon thesurface area of the FS300. It is evident that most of thepretreatments have lowered the surface area; a few treatments didnot significantly change the area; hydrogen treatment at 1000 0 Cincreased the surface area by 10%. It is also shown that treatmentswith such reagents as NO2 , NO, stearic acid, and methane attemperatures higher than 700%C have reacted with the carbon surfaceand closed off a portion of the micropore surface area.
Table 2 lists the values obtained from the linerization of thesecond order kinetic treatment of data from the cycling experiments
(batch and column) on various 40/80 mesh untreated and pretreatedactivated carbons. A second order kinetic model is used because thedata (the specific adsorption for consecutive cycles) fits thismodel, at least, to the extent of the cycling we have carried out.
The specific adsorption (zero intercept), Co , at zero cyclesand the slope of the equation were determined from a least squaresanalysis of the data. The correlation coefficient obtained is ameasure of the mathematical validity of the devised equation, andthe extent of deviation from the number one (1) is proportional tothe degree of scatter of the data points. The slope of the line isrelated to the rate of degradation with cycling. The nature of themodel is such that the slope is positive instead of an expectednegative one. For a standard of comparison the virgin carbon wascycled in triplicate for the batch experiments. The initialcapacities had a standard deviation of 0.01 and the slopes astandard deviation of 0.021.
As stated before, the experimental data follows the secondorder kinetic model (figs. I and 2). This relationship holds forboth the batch experiments and the column experiments. The specificadsorption for the batch is at the isotherm plateau and the specificadsorption for the column is calculated at thie I ppm breakthroughpoint.
%1-
To evaluate the carbons, the areas under the degradation curvesare integrated; this takes into account the rate of degradation andthe capacity of the carbon. The areas may then be compared to FS300as the standard. A significant increase in area indicates animprovement, and conversely, a decrease shows a poorer carbon.
The integrated linear plot gives the following equation:
A, (Area) imn2 /2 + n/CoWhere:
A = ydny 1/c
m slope, n cycle numberc = specific adsorption
Co = specific adsorption at n -o
Since the linear plot gives a value derived from a reciprocalrelationship, for ease of comparison we have taken the reciprocal ofA, so that a larger positive number shows an improvement over thestandard.
Table 3 shows the areas calculated from the degradation curvesof the various pretreated FS300's and two other commerciallyavailable carbons, namely FS400 and Witco 337. The significait
differetace at the 95% level for the batch experiments is 3 x £0-and for the column experiments, 2 x 10".
The GAC's which demonstrated some improvement in their cyclingcharacteristics as evidenced by a higher area for the batch cyclingexperiments were the 1000 *C hydrogen treated FS300, the FS400, andthe Witco 337. The FS300 pretreated with methane, NO2 and NO werechosen for column study for purposes of comparing their cyclingcharacteristics.
The results in table 3 show a significantly higher area in thecolumn cycling characteristics for the hydrogen treated carbon, theFS400, and the Witco 337. The other pretreated FS300 carbon, aspreviously indicated by the batch determinations show a lower areathan the virgin FS300. The improvement for the hydrogen treatedcarbon in the batch experiments was borderline, yet the columnexperiment improvement was significantly greater. Since thehydrogen treatments were carried out at different times, it ispossible that there may be slight differences due to variations inthe preparations. This appears to be a common occurance with GAC,i.e., relatively poor reproducibility of adsorption properties. The 4
fe FS400 and Witco 337 showed significantly higher areas under thecurves and, therefore, improved cycling characteristics over the
9
.7S.- :: . -4 ... . .. ;>
FS300. The water adsorption-desorption isotherms for the carbonshave given some insight into the reasons for these differences incycling characteristics.
Water Adsorption/Desorption Isotherms
Water is unique as an adsorbate due to its unusual properties,whi-ch is explained by its intermolecular hydrogen bonding. Thenature of water adsorption in the low pressure region of theisotherm up to a relative pressure of 0.4 is due to the presence ofactive sites from oxygen complexes on carbon. The water moleculeshydrogen bond at these primary sites and act as secondary adsorptionsites, and via hydrogen bonds, adsorb other water molecules. Thisresults in complexes of water molecules forming in the pores (ref3). As the pressure rises, the probability of adsorption increasesdue to the increase in the number of secondary adsorption centers.The steep rise in the adsorption isotherm curve in this relativepressure region is associated with the condensation in the poreswhich occurs as a function of related pressure and pore diameteraccording to the Kelvin Equation.
On the adsorption side of the hysteresis loop, the vaporpressure is in equilibrium with menisci in the body of the pores.On the desorption side, the vapor pressure is in equilibrium withmenisci at the pore mouths. Thus, we can look at the hysteresisloop with the desorption leg desorbing in order of decreasing poreopenings and the adsorption leg adsorbing in order of increasingpore diameters (ref 4).
"Water adsorption/desorption isotherms shown in figures 3through 7, were measured on 40/80 mesh carbon samples of FS300,FS400, Witco 337, and a series of FS300 pretreated samples. Thewater isotherms measured on all samples exhibited the characteristicS-shaped isotherm where the amount adsorbed at relative pressuresbelow 0.4 was low due to surface hydrophobicity and where theadsorption increased rapidly in the relative pressure range of 0.4to 0.8. Pronounced hysteresis was observed in the desorption legfor all isotherms. Carbons which were treated with NO2 or NO showedan increased hydrophilicity, the isotherm (fig. 6) showing an earlyrise in the amount adsorbed. The general interpretation for theseisotherms is that the surface Is relatively hydrophobic, i.e.,contain a low concentration of polar sites, and a high concentrationof pores in the diameter range of 20 to 80A , where the poreopenings are smaller than the body of the pores. A detailedanalysis of the results, however, indicated that all carbon samplesshowed differences In the degree of surface hydrophilicity and poresize distribution in the range investigated.
11
'4 l ::
The pore size analysis from the adsorption/desorption isothermswere carried out according to two procedures. One procedure, usesthe adsorption/desorption isotherm for FS300 as the standard andcompares the results obtained from the other samples with thisstandard. The results depicted in figures 8 and 9 and tabulated intable 4, show the change in pore size distribution relative toFS300. The absolute pore size distribution cannot be obtained fromthe water adsorption isotherms because a standard, i.e., a carbonsample known to contain no pores, does not exist for these carbonsamples.
The results of the size distribution of the pore bodiesrelative to FS300 as shown in fig. 8 and in table 4 for FS400, Witco337, and different pretreated samples of FS300, show that FS400 hasa higher concentration of pores over the diameter range of 20 to80A, which is the reason for the higher surface area of thissample. The Witco 337 charcoal shows a lower concentration of poresbelow 50A, and a much higher concentration above 50A. Both themethane and hydrogen high temperature treatment of FS300 lead todecreases in the pore size distribution between 20 and 80A. In thecase of the hydrogen treatment, where the surface area has beenincreased, it must be assumed that this treatment shifts the poresize distribution to higher diameters (refs 5,6). Both the NO andNO2 treatment of FS300 leads to an increase in pore diameter below40 and 50A, respectively, and a decrease in higher diameter pores.Figure 9 and table 4, show the results of the analysis of sizedistribution for the pore openings relative to FS300. The resultsare similar to the changes shown for the pore bodies in that theWitco 337, and the methane and hydrogen treated FS300 all showinitial decreases in the pore size distribution. The results aredifferent, however, in that the decrease is over a much narrowerrange for Witco 337 and FS300/H 2 /1000C. In the second method, thepore size distribution is calculated from the hysteresis loop in thewater adsorption-desorption isotherms (figs. 3 through 7). Theresults of this analysis, demonstrated in figures 10 and 11, showthe pore size distribution of the pore opening relative to that ofthe body of the pore where the pore opening is smaller.
The results show that the hydrogen treated FS300 and Witco 337have the highest concentration of pore size openings in the range of30A to 80A with the Witco 337 being higher after extrapolating theexperimental results. The FS300 and FS400 all showed similar poreconcentration in this range. The NO and NO2 treated FS300 bothshowed a decrease in the concentration of pore size openings in thisrange.
The following conclusions can be drawn from this analysis. TheFS300 and FS400 have pores which are similar in shape. The FS400,
4however, has a higher concentration of mesopores The Witco has alower concentration of pores below 50A and a much higherconcentration of pores above 50A, leading to a significantly highersurface area. The pores in the Witco sample also exhibit a highratio of pore opening diameter to pore body diameter as compared toFS300 and FS400, in the range of 35A to 60A, but with a higher ratiobelow 35A. However, the absolute diameter of the pore openingdiameter increases above 35A as compared to all the carbonsinvestigated. The NO2 and NO pretreatments of FS300 have the commoneffect of closing the pores with small openings and decreasing thesize of larger pores, thus, decreasing the specific surface area.The methane pretreatment on FS300 has the effect of closing poresover a wider pore size range, but to a smaller extent than NO2 andNO. Hydrogen pretreatment of FS300 has the effect of increasing thedimensions of the pores in the range of 30 to 60A. The poreopenings are especially affected in that they are decreased below45A and increased above this value with a subsequent increase insurface area.
Desorption
As pointed out in the previous report (ref 1), there is alwayssome residual TNT remaining on the carbon due to chemisorptionand/or very strong physical adsorption. Diffusion problems are alsoimplicit in this connection as the time to desorb using acetone ismuch less than the time of adsorption. Table 5 shows the typicaldata derived from cycling a column of 40/80 mesh FS300. The firstcycle desorption removes the least amount of TNT, subsequentdesorptions remove significantly more. Although the analyticalprocedure has a relatively large error, it can be seen that eachcycle leaves behind some irreversibly adsorbed TNT.
The data in table 6 under incremental desorpLion (for onedesorption cycle) shows that most of the TNT is removed with thefirst few hundred mL of acetone, and that further exposure tosolvent removes only minuscule amounts.
"Chemical Regeneration
Most of the TNT adsorbed on activated carbon is easily desorbedwith acetone. The residual is chemisorbed or held by very strongphysical torces. Other solvents such as dimethyl formamide,
4 Mesopores have a width of from 20-500
12
dimethylsulf oxide, benzene, and triethanolamine at temperatures upto 80C were not effective in removing any more TNT, and weredifficult to remove from the carbon. An investigation was carriedout to determine if chemical reactions could solubilize thischemisorbed portion. As shown in table 7, chemical treatment of thecarbon resulted in varying degrees of surface area loss. Treatmentwith alcoholic KOH, high temperature reaction with magnesiumchloride hexahydrate, hydrochloric acid, zinc, nitric acid, oxalicacid, and ozonation all proved destructive to the surface area. Asthese methods would not be practical on a large scale, both forsafety and economic reasons, further investigation of this phase wasdropped. Previous attempts to chemically regenerate carbon have notbeen successful (ref 7).
CONCLUSIONS
The use of chemical regeneration is not feasible for thereactivation of GAC as it damages the pore structure, tends to becostly, and generates harmful by-products for disposal.
Due to the heterogeneous nature of the activated carbonsurface, there is an inevitable loss of TNT adsorptive capacityafter each successive adsorption/solvent desportion cycle. Thestandard carbon, FS300 as shown previously, has one third of itssurface area unavailable to TNT adsorption. This fact, plus theloss of surface area due to irreversible adsorption at the highenergy sites of the pore ipenings, accounts for its poor TNT cyclingcharacteristics.
The TNT adsorption/desorption cycling behavior of the variouspretreated and other carbons were compared to the standard carbon,FS300, by integrating the areas under the linear degradation curves,which were derived from a second order kinetic model. Theintegrated areas depend upon the rate of degradation (slope) and thespecific capacity of the carbon for the TNT (y intercept). Thehighest areas will have high intercepts and a small slope. By meansof water adsorption/desorption isotherms it was found thatpretreated carbons which had lower integrated areas and hence,poorer TNT cycling capacity than the standard FS300, showed adecrease in the concentration of higher diameter pores. Conversely,the other carbons .,th higher integrated areas and hence, better TNTcycling capacity than the standard FS3CI showed pore sizedistributions that had an increased concentration of pore diametersin the higher range, i.e., from •OA to IOOA, and a higher ratio ofIpore opening diameter to pore body diameter.
13
The observations summarized above, have lead to anunderstanding of the requirements for a GAC which would be moresuitable for the adsorption/solvent regeneration cycling of TNT.This improved GAC should have an increased pore concentration in thepore diameter range of from 50A to 100A, with a high ratio of poreopening diameter to pore body diameter. The larger pore openingswould tend to minimize the blockage of pores by the irreversibleadsorption of TNT at the high energy sites of the pore openings.The present carbon, high in micropore concentration, shouldtherefore, be replaced by a commercially available carbon high inthe lower diameter size range of mesopores. To this end, carbonsshould be screened by TNT adsorption/desorption cycling and poresize distribution determinations for optimum regenerative behavior.
REFERENCES
1. T.C. Castorina, J. Haberman, and J. Sharma, "CharcoalRegeneration, Part I; Mechanism of TNT Adsorption," TechnicalReport ARLCD-TR-77065, ARRADCOM, Dover, NJ, Nov 1977.
2. J.B. Donnet, Carbon, 6, 161 (1963).
3. M.M. Dubinin, Chemistry and Physics of Carbon,Vol 2, MarcelDekker, NY, 1966.
4. A.J. Juhola and E.O. Wiig, Amer. Chem. Soc. 71, 2069 (1949).
5. J. Holmes and P.H. Emmet, J. Phys. Colloid Chem. 51, 1276(1947).
6. H.L. McDermot and J.C. Arnell, J. Phys. Chem. 58, 494, (1954).
7. Summary Report, "The Advanced Waste Treatment ResearchProgram," US Dept of Health, Education and Welfare, April 1965.
14
Aa
Table 1. Surface areas of virgin and pretreated carbons.
Treatment Surface area, m, /g
*FS300 (virgin) 1026TNT-sat/soxh-ext 578CH4/4 hrI600*C 1044C114/4 h-r/650-C 976CI{4/3 hr/700*C 1044C114/5 hr/7000C 1065CH /4 hr/750*C 649ArI4 hr/10000C 1049H2/4 hr/1000*C 1157N02/72 hr/H20 694N02/72 hr/freon 696N02/72 hr 777N02/72 hr/freon/acetone 537N402/18 hr 650N02/4 hr 6931402/4 hr/freon 642NO0/72 hr/H120 842NO/4 hr/freon 8111;0/4 hr/ 1120 805NO/4 hr/freon/acetone 732Silanized 959Steart. acid 741Witco 1311FS400 1119
i)s
Table 2. Rates of degradation of pretreated carbons
Batch
CorrelationTreatment CAcoefficient
FS300 (virgin) 0.128 0.49 0.92TNT-sat/soxh ext 0.294 0.42 0.88CH414 hr/600*C 0.169 0.41 0.90CH4/4 hr/650*C 0.184 0.45 0.96C114/3 hr/7000C 0.080 0.44 0.91CE4/5 hr/700*C 0.120 0.44 0.86CH4/4 hr/7500C 0.096 0.28 0.69Ar /4 hr/l000*C 0.141 0.44 0.89H2/4 hr/1000*C 0.135 0.52 0.95N02/72 hr/H20 0.150 0.35 0.86
*N02/72 hr/freon 0.124 0.33 0.94*N0 2/72 hr 0.304 0.38 0.99
N02/72 hr/freon/acetone 0.199 0.29 0.79N02/18 hr 0.116 0'&32 0.68N02/4 hr 0.205 0.35 0.97
*N02/4 hr/freon 0.139 0.32 0.82NO/72 hr/H20 0.161 0.38 0.87NO/4 hr/freon 0.207 0.40 0.96NO/4 hr/H20 0.060 0.44 0.87
*NO/4 hr/freon/acetone 0.083 0.33 0.67Silanized 0.231 0.52 0.91Stearic acid 0.234 0.47 0.96Witco 337 0.063 0.63 0.97PS400 0.041 0.55 0.75
Column
FS300 (virgin) 0.250 0.50 0.94CN/4 hr/700*C 0.353 0.48 0.93H 2 /4hr/ 1000*C 0.335 0.76 0.981402/2 hr 0.468 0.33 0.93NO/4 hr/H20 0.289 0.40 0.78Witco 0.183 0.63 0.98PS400 0.232 0.57 0.89
All carbons were PS300, 40/80 meeh, unless otherwise listed.
16
Table 3. Integrated areas under degradation curves
I-3
Area x 10
Treatment Batch Column
FS300, virgin 69 61
CH4/5 hr/700*C 63 53H•2/4 hr/1000*C 72 72
N02/72 hr 54 55NO/4 hr./H20 53 50FS400 86 68
Witco 337 94 78TNT sat'd/extracted 51 -
CH4/4 hr/6000C 56 -
dCH4/4 hr/6500C 60 -
CHI/3 hr/7000C 66 -
CH4/4 hr/7500C 43
Ar/4 hr/10000 C 62N02 /72 hr/H 0 50
N02/72 hr/freon 49S.N01/72 hr/acetone 41
N02/18 hr. 48
N02 /4 hr 48
N02/4 hr/freon 47
NO/72 hr/H20 34 -
NO/4 hr/freon 53 -
NO/4 hr/freon/acetone 51 -
Silanized 64 -
Stearic acid 59 -
Significant diff. at95% level 3 2
All carbons were FS300, 40/80 mesh, unless
otherwise listed.
17
A
s~d. ~ Ioo 00000 a 0 I
P4~ P4+
toV LA 0000L o
w Cow
914 H W Wjm
0 cca
0.0541.4
1.44 040 ok
U)m I
zc U
oo~18
Table 5. Column adsorption/desorption cycling.
to Breakthrough,Cycle Specific ads. Total specific Specific Percent
number at 1 ppm adsorption desorption desorption
1 0.48 0.53 0.39 73.6
2 0.39 0.42 0.36 85.7
3 0.36 0.39 0.32 82.0
4 0.34 0.38 0.34 89.5
5 0,32 0.38 0.30 78.9
* 6 0.29 0.34 0.33 97.17 0.30 0.36 0.33 91.7
Table 6. Incremental column desorption for one cycle.
Increment Amount
number Volume mL desorbed/g
1 300 1.7502 100 0.165
3 100 0.6064 200 0.0105 100 0.014
6 100 0.0097 100 0.006
Total 1000 1.960
Amount adsorbed was 2.52g - 77.8% desorbed.
19
' '......................... ' .. "
Table 7. Ef fect of chemical regenerationon surface area of PS300.
Treatment Surface area, mn2/
None 945Sulfuric acid 848Reduction 318Reduction 655Oxalic acid 714Ozone 723Nitric acid 469
40/80 mesh, FS300 was saturated with TNT andthen soxhiet extracted with acetone.
20
-'4
Table 8. Batch cycling of 40/80 mesh FS300/Argon/4 hrs/1000*C.
Cycle number Specific adsorption
1 0.412 0,423 0.344 09365 0.346 0.327 0.278 0.279 0,,27
10 0.2711 0,2612 0.2313 0.2414 0.2315 0.2216 0.1517 0.1518 0.15
*0.235u+1.806C
corr. coeff * .917
21
WEE..K7-7'
7 To,,A
Table 9. Batch cycling of 40/80 mesh FS300/CH4/600'C/4 hr.
Cycle number Specific adsorption
1 0.382 0.313 0,344 035 0,316 0.327 0.*278 0.269 0.23
10 0.24
1*0.174n+2.453 corr coeff. -. 909C
22
PORO
0
4 4
0
03'
0
-4-
CC'
4,N'
C" -
0Witco 337
CFS400PS300
0.4
0.
0
cycle w~b41'
Figure 2. Linear plots of speific adsorption~ deactivtwion a.s at
funct ion of adsorption/de-wrption cyc1ling ac~cording to aV econid orde~r kineti metodel.
24
il7I7
7,' _7
I I-777
bo
0
.0
0..
.0 r44
00
0)00
254
C4 0
00
-4T
.4z.
03
0NW
4,4 4.P
Vita Pe~jopy O
260
I U
00
0N
4 0Co
0
'-4
0
00
lq '0
-I r�4
0Co0
04 N �' Jo0 � 0
. 0%%%4���** ,� - -� N
4 �
� C1.4 �0.
N 0� 0.� 0..4 d-4 �J
w
-� 00
'.0Iai
�I4:44
�fO'zI N
AIn p-4
U/9 p.qJoopv o'��
* 27
I
04
$40
.44
4-
Oo
00
0
-4'
Nj
4114
.... ..... ivelpsqovp O0
914
280
00
0.0
co
00
0
w
N44
00
0I C2 P4
Al.'
0004 C
000
a- /1
41 C (4) M 41
44I
t4 U
~%to
300
u U
0a 0 OD
X0
o 0 0 0oo 0 0
o 4) v4*-4 -. w
- *.4-P4
a 0
*~C *. . *
lo01004 *so oloioo ;*
31 1.
04
0'
0U 0
to000 e
r -.
0 u0,-r4
C3
1A1
40 a0e44
'"MICA I.O
32~
00
>04
o d 0
o '-.T
w01
N C"O
4-4
1 0C-
0.101.
33J
APPENDIX
34
I•-
I Figure A-1. Autoimat ic fract ion collector with carbont column.
35
"a
Table A-i. Small columna adsorption/desorption cyclingas a function of particle size of FS300.
o... . . ~Cumulative TNTb
Volume Retained, m Effluent, mgThru liters Coarse ' 4 oarse 40/80
1st Adsorption1 116 118 1.70 0.02 227 236 5.90 0.03 342 354 1.30 0.024 444 472 13.90 0.205 542 590 18.40 0.06 632 704 24.90 0.07 719 818 28.30 0.01 796 932 38.40 0.0
2sn Desorption0.1 174 674 57.9 72.30.2 591 805 72.1 86050.3 617 831 75.4 89.00.4 626 837 76.4 89.0065 636 847 78.3 91.01.0 644 853 79.0 91.7
2nd Adsorption1 120 120 1.1 8.12 240 241 0.0 0.0
3 360 362 2.4 0.184 478 484 3.0 0.085 596 604 3.9 0.11006 668 729 52.7 0.0017 757 854 35.8 0.008 841 979 41.6 0.50!iii 2nd Desorption
0.1 646 852 76.9 87.1
0.2 699 896 83.1. 91.8S0.3 716 912 84.7 93.4
b 0.4 722 920 86.0 94, 0• 0,5 728925 87.0 ?,S0.6 732 928 87.2 95.2
I,.0 743 936 88.3 95.7
i Coarse 3.001 8.1 1.0 6.37S40/60 3.031 7.2 1.0 5.67
bi bAverage rate 1.4 em 3/min,; adsorption from aqueous solutiou, dosorptiotn
Swith acetone.
36
• ••,•°-.
Table A-1. Small column adsorption/desorption cycling as afunction of particle size of FS300 (Cont'd)
Cumulative TNT
SVolumeb Retained, mg Effluent, mgThru liters Coarse 40/80 Coarse 40/80
3rd Adsorption1 131 134 5.5 2.32 257 269 10.1 2.13 375 414 19.0 8.5
% 4 459 524 27.7 2.25 540 632 30.8 4.016 623 780 67.4 2.17 701 910 77.9 2.28 752 1039 80.0 2.1
3rd Desorption0.1 576 847 76.6 8140.2 657 891 87.4 85.70.3 685 916 91.0 88.20.4 704 934 93.6 89.80.5 725 954 96.3 91.60.6 734 968 98.5 94.11.0 746 988 99.3 96.2
4th Adsorption1 132 137 8.0 3,32 267 Z74 5.1 2.93 323 383 56.4 2.94 409 480 25.9 15.05 501 618 63.0 17.56 600 743 56.0 31.0
4th Desorpt ion0 1 444 516 73.7 69.50 2 584 659 97.4 89.00.3 606 681 100.5 91.80.4 616 691 102.5 93.00.5 626 701 104.5 95.01.0 651 721 107.0 97.1
8Particle size FS300 Wit, Length, c Diameter. cm Volume, c
Coarse 3.001 8.1 1.0 6.37t.40/80 3.031 7.2 1.0 5.67
bAverage rate 1.4 cm h/in; adsorption from aqueous solution, desorptionwith acet-,ne.
37
A,
Table A-2. Small columna adsorption/desorption cycling
as a function of particle size FS300.
Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80
1st Adsorption
1 153 153 0.10 0.1002 306 306 0.00 0.003 459 459 0.13 0.004 612 612 0.60 0.225 764 765 0.70 0.246 916 918 4.00 0.287 1065 1070 4.60 0.318 1206 1222 12.3 0.809 1339 1365 30.5 9.50
10 1445 1506 32.0 12.011 1540 1633 59.0 18.012 1622 1764 72.0 24.0
1st Desorption0.1 1049 1186 64.6 67.20.2 1232 1333 75.8 75.40.3 1272 1348 78.4 76.90.4 1385 1361 83.3 77.10.5 1496 1390 92.2 78.80.6 1507 1419 92.9 80.41.0 1517 1530 93.5 86.7
2nd Adsorption1 145 145 0.60 0.702 290 290 0.70 0.503 435 435 0.40 0.504 580 599 0.70 0.505 746 749 3o9 0.806 887 847 8.6 02.207 1160 1041 13.8 6.o8 1225 1058 18.4 8.6
* 3
aparticle Size FS300 t Length•em Diameter,cm Volume, cm
Coarse 4.000 8.1 1.0 6.37b 40/80 4.00 7.9 1.0 6.35
Average rate 1.7 cm3 /min; adsorption from aqueous solution, desorptionwith acetone.
38r ............................................
Table A-2. Small columna adsorption/desorption cyclingas a function of particle size FS300. (cont'd)
Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80
2nd Desorption0.1 959 814 78.2 69.00.2 994 863 81.1 73.20.3 1019 981 83.2 83.20.4 1045 997 85.3 84.60.5 1059 1010 85.4 85.71.0 1191 1038 97.2 88.0
3rd Adsorption1 110 120 0.10 0.202 225 240 0.33 0.603 330 360 0.30 0.304 480 480 0.35 0.305 649 600 1.06 0.406 806 767 12.6 22.87 1055 911 20.4 34.68 1189 1020 35.0 56.3
3rd Desorption0.1 772 620 64.9 60.80.2 955 932 80.4 81.60.3 963 962 81.0 84.60.4 979 979 82.4 86.20.5 994 991 83.6 87.41.0 1013 914 85.2 89.6
aparticle Size FS300 Weight, Lengthcm Diametercm Volume. cm3
Coarse 4.000 8.1 1.0 6.37b 40/80 4.00 7.9 1.0 6.35
Average rate 1.7 cm /min; adsorption from aqueous solution, desorption
with acetone.
39
A,
Table A-3. Small columna adsorption/desorption cycling
as a function of particle size FS300.
Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80
Ist Adsorption1 154 154 0.20 0,212 308 308 0.23 0.193 462 462 0.24 0.364 616 616 0.37 0.375 770 770 0.18 0.476 924 924 0.37 0.177 1078 1078 0.71 0.608 1232 1232 0.60 0.509 1386 1386 0.64 0.60
10 1540 1540 0.59 0.5511 1694 1694 0.58 0.5412 1848 1854 0.56 0.5213 2002 2002 0.59 0.6014 2156 2156 0.83 0.5215 2300 2300 0.97 0.5916 2454 2454 0.82 0.5817 2608 260M 0.63 0.4918 2761 2762 1.04 0.4619 2913 2916 2.30 0.5320 3064 3070 3.10 0.4621 3205 3215 4.00 0.6022 3344 3360 6.20 0.8423 3481 3505 8.20 0.5724 3625 3660 11.0 0.5925 3761 3821 19.4 8.30
26 3891 3969 25.0 15.4027 4017 4104 29.8 20,4028 4136 4230 36.4 24.0
29 4245 4384 46.0 32.0
5' 3aParticle Site FS300 2t!ht.,g Lejhcm Dinmeter,em Volume, cm
Coarse 10.000 21.2 1.0 16.640/80 10.00 21.9 1.0 17.2
b 3Average rate 4 cm /min; adsorption from aqueous solution, desorption
with acetone.
40
.....
Table A-3. Small columna adsorption/desorption cycling
as a function of particle size FS300. (Cont'd)
Cumulative TNT
Volumeb Retained, mg Effluent, mugthru liters Coarse 40/80 Coarse 40/80
1st Desorption0.1 2640 3250 62.6 75.80.2 3431 3760 80.8 87.60.3 3585 3885 84.6 90.80.4 3602 3906 85.4 91.20.5 3610 3921 85.0 91.91.0 3670 3751 86.4 92.3
2nd Adsorption1 123 123 0.24 0.302 246 246 0.40 0.103 369 369 0.63 0.184 484 484 0.60 0.235 599 599 0.68 0.426 714 714 0.35 0.407 829 829 0.40 0.408 944 944 0.62 0.389 1060 1060 0.62 0.65
10 1176 1176 0.54 0.5211 1292 1292 0.54 0.5212 1408 1408 0.52 0.4213 1524 1524 0.40 0.20.14 1643 1643 0.60 0.4015 1762 1762 0.50 0.4616 1871 1871 0.40 0.5017 1990 1990 0.50 0.6018 2109 2109 0.50 0.4019 2228 2228 0.070 0.4020 2347 2347 0.60 0.3821 2466 2466 0.68 0.4422 2585 2585 0.73 0.4623 2700 2700 1.30 0.4224 2808 2815 6.9 0.70
25 2912 2930 11.3 0.83
aParticle Size PS300 Weight,'g Length~cm Diameter,cm Volume, cm3
Coarse 10.000 21.2 1.0 16.6b 40/80 10.00 21.9 1.0 17.2
Average rate 4 cm i/mixj; adsorption from aqueous solution, desorptionwith acetone.
41
Table A-3. Small columna adsorption/desorption cycling
as a function of particle size FS300. (Cont'd)
_ _ _ Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80
2nd Adsorption26 3013 3043 14.0 1.9227 3108 3154 20.0 4.3028 3199 3261 24.0 10.029 3286 3357 29.3 19.330 3365 3447 36.0 24.6
2nd Desorption0.1 2460 2630 73.1 75.70.2 2715 2968 80.8 85.40.3 2785 3015 82.8 87.80.4 2792 3029 83.1 88.00.5 2812 3043 83.6 88.21.0 2874 3106 85.4 89.4
3rd Adsorption1 125 125 0.10 0.002 250 250 0.10 0.003 375 375 0.09 0.084 500 500 0.50 0.335 636 636 0.10 0.306 772 772 0.70 0.497 907 908 0.91 0.558 1043 1044 0.60 0.489 1155 1155 0.54 0.65
10 1267 1268 0.70 0.4011 1379 1380 0.70 0.6012 1527 1529 1.23 0.3113 1657 1660 1.70 0.1814 1787 1792 2.30 0.9515 1927 1936 3.10 0.3016 2064 2074 3.00 1.5017 2201 2212 3.48 2.3818 2311 2324 7.50 8.1019 2421 2428 12.00 16.0020 2525 2503 15.00 34.0021 2623 2580 21.0 52.0
3aParticle Size PS300 Weight~g Length~cm Diameter~cm Volume, cm
SCoarse 10.000 21.2 1.0 16.640/80 10.00 21.9 1.0 17.2
bAverage rate 4 cm3 /min; adsorption from aqueous solution, desorptionwith acetone.
42
• ,. 1 ,• , ....... .. ..- ,. . ,."•'
Table A-3, Small columna adsorptioni/desorption cyclingas 8 function of particle size FS300. (cont'd)
Cumulative TNT
Volumeb Retained, mng Effluent, mgthru liters Coarse 40/80 Coarse 40/80
3rd Desorption0.1 1065 1520 40.0 57.60.2 1773 2038 66.6 78.20.3 2194 2479 82.4 95.20.4 2243 2532 84.9 97.00.5 2265 2550 84.6 98.00.6 2277 2562 85.8 98.31.0 2295 2581 86.2 99.2
a 3Patil Size FS300 Weightg Length~cm Diameter,cm Volume, cm
Coarse 10.000 21.2 1.0 16.6
b 40/80 10.00 21.9 1.0 1.bAverage rate 4 cm 3 min; adsorption from aqueous solution, desorption
with acetone.
I. 43
"q,
Table A-4. Small columna adsorption/desorption cyclingas a function of particle size FS300.
Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80Adsorption
1 154 154 0.31 0.242 308 308 0.51 0.603 462 462 0.46 0.314 616 616 0.34 0.345 770 770 0.75 0.806 924 924 0.61 0.717 1078 1078 0.63 0.718 1232 1232 0.84 0.719 1386 1386 0.49 0.37
10 1540 1540 0.55 0.3911 1694 1694 0.73 0.6812 1848 1848 0.81 0.7013 2002 2002 0.64 0.3414 2156 2156 0.70 0.4215 2310 2310 0.62 0.7416 2464 2464 0.79 0.8217 2618 2618 0.70 0.7218 2872 2872 0.79 0.7219 3040 3040 0.57 0.6020 3208 3208 0.30 0.3521 3376 3376 0.29 0.6322 3544 3544 0.11 0.2723 3542 3542 0.11 0.1724 3722 3722 0.24 0.2925 3872 3872 0.61 0.2126 4022 4022 0.23 0.6027 4172 4172 0.32 0.6528 4327 4327 0.36 0.1029 4482 4482 0.10 0.4030 4627 4627 0.36 0.2131 4772 4772 0.21 0.3432 4917 4917 0.12 0.32
38Partlcle Size FS300 Weght.Lg Length•cm Diametercm Volume, emr
Coarse 20.00 12.3 2.0 39.4b 40/80 20.00 12.3 2.0 39.4Average rate 4 cm3 /min; adsorption from aqueous solution, desorptionwith acetone.
44
'At " '
* Table A-4. Small columna adsorption/desorption cyclingas a function of particle size FS300. (Cont'd)
Cumulative TNT
Volumeb Retained, mg Effluent, mgthru liters Coarse 40/80 Coarse 40/80
Adsorption (cont 'd)33 5062 5062 0.18 0.4334 5205 5205 0.24 0.6335 5348 5348 0.59 0.3636 5491 5491 0.56 0.0337 5634 5634 0.56 0.5038 5776 5777 1.3 0.6039 5917 5921 2.8 0.6840 -6055 6065 5.0 0.5441 6190 6208 7.7 0.5142 6325 6351 8.1 0.4143 6461 6494 9.6 0.7044 6604 6647 12.0 2.445 6745 6798 14.6 4.946 6880 6847 20.0 6.047 7011 6995 24.5 10.048 7125 7107 30.0 12.5
*49 7242 7211 38.0 18.050 7354 7308 43.0 29.0
Desorption0.1 3380 3870 46.0 52.40.2 5460 5775 74.0 78.30.3 5900 6330 80.1 85.70.4 6079 6511 82.5 88.40.5 6169 6584 85.3 89.21.0 6349 6786 86.4 92.2
3aParticle Size FS300 Weiltht, Leghc Diameter.em Voue
Coarse 20.00- 12.3 2.0 39.4b40/80 3 2Q.00- 12.3 2.0 39.4
bAVexage rate 4 cm /midn; adsorption from aqueous solution. desorptionwith acetone.
45
- r -n ON If C1 tc)
e %L)' 4
0
l-ri
.,...'00 0 CD0 000C)0 40000 00 6 * % 9 * S 4 9 * i
oH .uOOOOOO 00 0 0 4
n00
j0040 - 4 J V)N-40eqD"M0 mfCJ-4CJ M nN e - - --4 0'ON N'N 14 IN
9. th 9i 14 . *;900 00 04 0
to 0J
4) tw
00aG'4 OND-. k.nt-4 N 0
"v40
oilto
"411
464
4-1 ) e 000000 C 00 %1-%QC1 (In0 000 e
41!0
~ .,4-44000 ~400 -4 -40.-400
0 CO
0.0.
V) w000010 \ 0 - 0r
-H.q4 4(nmc nM e , 4c C-4 M M(4
' C14 N 4
91W. () 1 * s * # * * * * * . . . . . .w
0)0 C000
CA 4
w- 0jc
0 tl 4 C C o r N N 0 f,%
00
tj C4 cn %cr kA %D .-4 N m 4 t D 4N 44
%Ttm %
'.4)
in)
$4JL UN*
1'47
V .r # a 0 0 - 9 . . . 0 * * * 9 . * 9 9 0 & 9 *
v-4tto0000 000000 000000 0(00000
0
94
0)'d0W 4N000 ý ) 0 -4 0-,1 C'000 00000 o N f)10 LM
.,.4. 4JOOOOOO 000000; 000000~c 000000 ;
000
u00 03
00 *0-r
4.1CV .-4 cMN"NNc'N-l fnmmr4'NN~ N~mni lrNN(JN~qNC
;It1 43o000 000000 000000 000000 oo.;6
9v3
(1)
4)
14 ~t~f li 00 00O'
-4-
.14
~p 14 14 .4 0 \k . 2
41~ 0 - C'jC'J ý4.N C14 0-n-4 -4 C'JN0
0-4 0 0 0 0 0 00 0 00000; C C
4.4
0 v
~J0000000000 0000 ) r4.
0.0 a)4
0 a)
Q 0
411
U Vol
Vq4 "4N ~ 5 C 'N N ~ I
o44 o oo o 0000004 - - 000900n Lt) 14
.11 f
caa
i0
0.tG 1.4 0 )
W)N~~4 NO
49
-4J
00
0 ID r-.u Ln ) o0 a% Ch cS'0m t) u M M -4 0 LI)iun S
U 4 )0 0 0 00 00 0C)000 0 00 000 0000 0
0 SH
u0
44* -H
QO0O'~ 14.ru C;-'o 01) 0000000 000000 0000 0000 .
lu41
U4 .) % 0ON N 0% 0 otn N 0in r, n O 0 0
mm 0 N N "M N N m M %) -lk N 0
5" (0n-.
14 a4
0 0
17 U4
r-4C c~ 00 00000 000000
4))o0 0~ 4 r ~lr.or. 4 t'o
- 00 00 00 00 0000
00 " )r
4)0 0 C
.V) 0~
0
01)0000000 0 0 0 C 0C%0 0 0,0 0
-4
.4k o "Nm.y4 \ n%
ý4
4M 4) tD. ) I0 114 0 1
.5.4 .....
0
0 .
0 0)
enI '400000C Z\ mML ý-00 0 0 P,4a\ \IC
UI 000 00 OC O O OO O O0) 0 0C 0C
01
.... .
4-r
UIV
-It-z -C~ 04 0 w% ý -% C4r, D L
44 0-'C4C\-Dt.4
W0~0000000 00000 000
ej0
0 000000-10 00000000000
tKEl0
(A 0. 0
00
-' .-~4-I 4,Z O 0Or. ~ ~ ~ D O 0 f '
.4,.j~ * * p * p 5 5 5 35 5 5 S S
-r-4 0 1 0000000N L
u0
0. cw mr)$4 *'00000
0 000 000
J-4
m 0 0
14
-54
DISTRIBUTION LIST
Commiander/DirectorV Che-. cal Systems Laboratory
U.S. Army Armament R&D CommandATTN: A. Hillsmeier
D. Renard (9)Aberdeen Proving Ground, MD 21010
CommanderU.S. Army Armament R&D CommandATTN: DRDAR-CG, W A.H. Light, Jr.
DRDAR-TD, Dr. R. WeigleDRDAR-TDS, Mr. V. LindnerDRDAR-LC, GOL D.P. WhalenDRDAR-LCM, Mr. L. SaffianDRDAk-LCE, Dr. R.F. WalkerDRDAR-LCE, Mr. T. Castorina (25)
DRDAR-TSS (5)Dover, NJ 07801
Directorballistic Research LaboratoryU.S. Army Armament R&D CommandATTN: Technical LibraryAberdeen Proving Ground, HD 21010
CommanderRadford Army Ammunition PlantATTN. Dr. W.T. BolleterHercules, Inc.Radford, VA 24241
CommanderBadger Army Ammunition PlantBaraboo, WI 53913
Comma nderIndiana Army Ammunition Plant
* lCharlestown, IN 47111.
Commander* lHolston Army Ammunition Plant*: Kingsport, TN 37660
I
:-As. -
[I
CommanderLone Star Army Ammunition PlantATTN: Technical LibraryTexarkana, TX 75501
CommanderMilan Army Ammunition PlantMilan, TN 38358
CommanderTooele Army Ammunition PlantTooele, UT 84074
CommanderIowa Army Ammunition PlantSilas Mason, Mason & Hanger, Inc.ATTN: Technical LibraryMiddletown, IA 52638
CommanderJoliet Army Ammunition PlantJoliet, IL 60436
CommanderLonghorn Army Ammunition PlantMarshall, TX 75670
Defense Technical Information CenterCameron Station (12)Alexandria, VA 22314
CommanderLouisiana Army Ammunition PlantShreveport, LA 71130
CommanderNewport Army Ammunition PlantNewport, IN 47966
CommanderVolunteer Army Ammunition PlantChattanooga, TN 37401
Cotma nde rKansas Army Ammunition PlantParsons, KS 67357
56
. . . .7
"? "• '::< • ••"• ,", .. • <' •.. ; ,,.., . ., ,,., .• - •¢ • ... . % .;• -,,•:, '• • <.'•, .: .' • -. :: - ";"<..• •::','; ',,.'., <: •• .' ., , ,''•
* CommanderU.S. Army Research OfficeATTN: Dr. G. WymanBox CM, Duke StationDurham, NC 27706
CommanderNaval Surface Weapons CenterWhite Oak LaboratoryATTN: Technical Library
WieDr. J. lioffsommerWieOak, Silver Spring, MD 20910
DirectorDARCOM Field Safety ActivityATTN: DRXOS-ESCharlestown, IN 47111
CommanderUS Army Training &Doctrine CommandATrTN;. ATEN-MEFort Monroe, VA 23651
Commanderw US Naval Sea Systems Command
ATTN: SEA-0332Dr. A. B. Amstar
Washington, DC 20362
Lawrence Livermore LaboratoryATTN; Technical LibraryP.O. Box F;08Livermore, CA 94550
CommanderFort DetrickU.S. Army Medical BioengineeriagR&D LaboratoryATTN: Dr. D.H. RosenblattFrederick, MD 21701
CommanderU.S. Army Armament MaterielReadiness CommandATTN: DRSAR-LEP-LRock Island, IL 61299
A S7
DirectorU.S. Army TRADOC Systems AnalysisActivityATTN: ATAA-SL (Technical Library)White Sands Missile Range, NM 88002
Weapon System Concept Team/CSLATTN., DRDAR-ACWAberdeen Proving Ground, MD 21010
DirectorU.S. Army Ballistic Rsch LaboratoryARRADCOMATTN. DRDAR-TS B-SAberdeen Proving Ground, MD 21005
Technical LibraryATTN: DRDAR-CLJ-L
Aberdeen Proving Ground, HD 21010
Benet Weapons LaboratoryTechnical LibraryATTN: DRDAR-LCB-TLWatervliet, NY 12189
U.S. Army Materiel SystemsAnalysis ActivityATTN: DRXSY-MPAberdeen~ Proving Ground, MD 21005
Chief, Demolition BranchSpecial Forces SchoolATTN. Staff Sgto MonahanFor~t Bragg, NC 28307
58
