The influence of surface modification of BPL carbons on aging

Carbon Vol. 34, No. 8, pp. 975-982,1996 Copyright 0 1996 Elsevier Science Ltd

Pergamon

Printed& Great Britain. All rights reserved ooO8-6223/96 $15.00 + 0.00

80008-6223(96)00059-0

THE INFLUENCE

OF SURFACE MODIFICATION CARBONS ON AGING

OF BPL

S. S. BARTON,~ M. J. B. EVANS,~ S. LIANG,~ and J. A. F. MACDONALD’** ‘Department of Chemistry and Chemical Engineering, Royal Military College of Canada, Kingston, Canada, K7K 5L0 bPhysical Protection Section, Defence Research Establishment Suffield, Medicine Hat, Alberta, Canada. TlA 8K6 (Received 11 October 1995; accepted in revisedform 27 February 1996) Abstract-The treatment of BPL carbon with various surface modifying agents, such as nitric acid, hydrogen, chlorine and ethylene, has been carried out. The treated and as-received carbon samples were aged under high humidity conditions and changes in the surface functional groups were examined using temperature programmed desorption, enthalpy of immersion determinations in water and water adsorption isotherms. Aging of many of the carbons at 84% RH for extended periods of time (up to 4 years) was seen to increase the amount of desorbed oxygen from the surface, increase the enthalpy of immersion in

water and alter the initial uptake of water vapour at low relative pressures. These changes appeared to be retarded by low temperature (< 180°C) chlorine treatment. Nitric acid oxidized BPL carbons containing high concentrations of surface oxides before storage or aging showed little indication of further oxidation when subjected to high humidity conditions. Copyright 0 1996 Elsevier Science Ltd Key Words-Surface

modification, chlorination, aging.

1. INTRODUCTION The surface of an activated carbon is in many ways a dynamic entity which may be influenced and altered by the method of storage of the carbon. It is well known that exposure to atmospheric conditions can diminish the practical uses of the carbon by adding oxygen containing functional groups to the surface, thus altering its adsorptive properties. A reduction in the chloropicrin adsorption capacity of aged carbons [l], a decrease in the efficiency of the removal of chloropicrin from humid air [ 2,3] and the deleterious effect of carbon aging on the retention efficiency of methyl iodide [4,5] are some examples. Treatment of the carbon surface with a stabilizing agent may be a means of diminishing the detrimental effects of storage [2,3,6]. The oxidation of carbon at ambient temperatures in the presence of air and water has long been recognized [ 71. King [S] demonstrated this by the extraction of oxalic acid from a sugar carbon that had been exposed to the air. The oxidation reaction postulated in the literature by Pierce et al. [9] was: C + H,O = CO, + Hz + C-O complex which was said to be dependent upon the reaction temperature. Two or more types of oxide complexes were later suggested [lo], the less stable decomposing to give CO, at low temperatures and a more stable complex that decomposes at high temperatures with the evolution of CO. The adsorption of water vapour on porous and non-porous carbon has been studied extensively [ll-151. Primary adsorption sites on the carbon

*To whom

all correspondence

should

be addressed.

surface capable of forming hydrogen bonds with water molecules influence the initial uptake of water vapour at low relative pressures. These primary adsorption sites, mainly carbon-oxygen complexes, vary in number and composition depending on the treatment of the sample. Dubinin and Serpinsky [ 161 developed a three parameter equation applicable to the water vapour isotherm which allowed the concentration of primary adsorption sites, a, to be determined from experimental isotherms. A modification of this equation [17] which introduced a fourth parameter allowed the fitting of water adsorption isotherms on super activated carbon (AX-21) and on microporous PVDC carbon over the entire relative pressure range. In this equation: h=

a c(a, + a)(l-exp(-k*(a-a,)‘))

(1)

h is the relative pressure, a is the adsorption, a, is the concentration of primary sites, c the ratio of adsorption-desorption rate constants, a, serves to trigger the start of the decline in adsorptive capacity and k governs the rate of this decline. Again, the primary adsorption sites are generally considered to consist of hydrophillic surface oxides. Treatment of carbon adsorbents with nitric acid has been shown to increase the concentration of these surface oxides [ 11,18,19] and reduction by hydrogen results in a significant decrease in a, [ 14). It was determined by Puri and Bansal [20] that treating carbon with chlorine rendered the carbon more water repellent. Also the pretreatment of graphite with chlorine at elevated temperatures resulted in a resistance to oxidation in air [Zl]. Treating the 91.5

S. S. BARTON et al.

916

degassed carbon with chlorine at relatively low temperatures, using a simple one step procedure, may stabilize the surface, thus diminishing the effects of high humidity aging on the adsorption properties. The chlorine treated carbons must not show a drastic reduction in adsorption capacity, indicative of pore closure, as observed by Verma and Walker [6] when elevated temperatures were used. Aside from the simple physisorption of chlorine suggested mechanisms for the interaction of chlorine with the carbon surface involve the saturation of olefinic bonds, an exchange with chemisorbed hydrogen or a dehydrohalogenation of the carbon to produce more olefinic bonds [ 221. Chemical treatment of carbon surfaces with various reagents, including ethylene, was previous studied by Schoderboeck et al. [23] as a method of influencing microporosity. Although high temperature treatment resulted in a reduction of the surface area, infrared spectroscopy indicated a reduction in the population of alcohol functionalities on the surface and the disappearance of the ketone absorbance band. In this study surface modified carbons were aged at 84% RH to determine the changes in the surface oxides due to humidity. The oxidised and as-received carbons were also stored dry as control samples. 2. EXPERIMENTAL

2.1 Treatment of the carbon Calgon BPL carbon was used throughout this study. The as-received (AR) sample was washed, dried, and sieved to 16X20 mesh (US Standard) but otherwise untreated. The nitric acid oxidized BPL carbons were prepared as in ref. 14. The AR and oxidised carbons were dried at llO”C, in air, after extensive washing. Both the AR carbon and the oxidised carbons were reduced in a flowing hydrogen/nitrogen mixture at 500°C for 2-3 hours

0.54

T

I

I

I

I

0 0.52

0

O

0

v

V

v.0 0.44

v I

1000

a 0 l

1040 Surface

I

I

1080

1120

I

1160

area/m’.g-’

Fig. 1. Variation of Dubinin micropore volume with surface area from nitrogen isotherms at 77 K: v AR; 0 reduced carbons; V nitric acid oxidised and 0 chlorinated carbons.

[ 141. Some samples were aged in a sealed desiccator over a saturated KBr solution delivering approximately 84% RH for an extended period of time. Laboratory temperatures varied from a low of 15 to 30°C over the time period (up to 45 months). Some samples were stored dry, immediately after removal from the oven, in a sealed desiccator. The BPL carbon samples to be treated with either chlorine or ethylene gas (both gases were obtained from Scott Specialty Gases, Inc at 99.5% purity) were first evacuated at 110°C overnight in a stainless steel reaction vessel contained in an oven. After allowing the system to equilibrate at a preset temperature (in the range of 50-170°C) the reacting gas was allowed to contact the degassed carbon for a specific period of time at a pressure of approximately 50 kPa. The reaction vessel was enclosed in the oven at constant temperature for the duration of contact time of the gas with the carbon. Any excess heat due to the reaction within the vessel was not monitored. Chlorine treated carbon samples were washed with large quantities of methanol, some for a week in a soxhlet extractor, after which they were dried in air at 110°C. This extensive methanol washing was carried out to remove any physisorbed species from the surface of the carbon and to ensure any changes occurring on aging in a humid environment were not already happening, in the washing procedure, had water been used. The chlorination procedure is similar to that used by Hall and Holmes [2] except restricted to low temperatures (170°C and below) and short reaction times (usually 30 minutes). Chlorinated samples prepared in this study were never boiled under reflux with the washing solution. After treating the carbon with ethylene the system was flushed with nitrogen and cooled to room temperature. These carbons were not washed but air dried at 110°C. Portions of the treated carbons were aged at 84% RH as above. These samples were redried at 110°C after aging and before characterisation.

2.2 Characterisation

of the carbon

The apparatus for water adsorption and the temperature programmed desorption (TPD) method are described elsewhere [ 141. The nitrogen adsorption isotherms were determined at 77 K using a Micromeretics ASAP 2000 apparatus after evacuation of the samples at 383 K to <7 x 10m3 mbar. The nitrogen isotherms were fitted to both the BET and the equation [24] for a surface area Dubinin-Radushkevich (D-R) equation for a micropore volume [25] using 0.162 nm’ for the area of a nitrogen molecule and 0.813 g.mL-’ for the density of liquid nitrogen. The enthalpy of immersion in water for the treated and aged samples was determined using a calorimeter also described elsewhere [ 261. The total chlorine content of the carbons was determined by neutron activation analysis. Standards for the procedure were prepared by a wet impregnation of BPL carbon with a known concentration of

The influence

of surface modification

of BPL carbons

911

on aging

0.5

0.4 ;

ul

Q z

0.3

g

0.2

z m a

0.1

0 F

0.0

0.4 ;

WI

4

0.3

$ F 2

0.2

0 E -z

0.1

0.0

C )

0.2

0.4

RELATIVE

Fig. 2. Water

adsorption

isotherms

0.6

0.8

0.0

0.2

at 298 K: 0,

0.4

RELATIVE

PRESSURE

0.6

0.8

1 .o

PRESSURE

0, AR, AR-stored; A, A, 1.3 hour nitric acid oxidised, same-stored; and v, V, 4 hour nitric acid oxidised, same-stored.

n , 2 hour nitric acid oxidised, same-stored;

Table 1. Parameters of eqn (1) fitted to water adsorption and a series of nitric acid oxidised BPL carbons Carbon BPL-AR BPL-AR-S BPL-1.3 BPL-1.3-S BPL-2 BPL-2-S BPL-4 BPL-4-S

a, mmo1.g 0.711 0.683 3.18 2.50 1.98 1.95 3.13 3.05

1

c g.mmol-’ 1.91 1.86 1.92 2.22 2.20 2.24 2.08 2.23

KC1 followed by drying at 110°C in air. The Cl-BPL carbons and the standard were subjected to bombardment by thermal neutrons in a SLOWPOKE-2 research reactor. Some of the Cl nuclei became radioactive ($Zl*). As $Cf* decayed, “fingerprint” y-rays were recorded, and then the spectrum analyzed for the element of interest. The spectrum from the chlorinated carbon sample was then compared to the y-spectrum of the irradiated standard. Since the standards used in the determination were impregnated BPL carbons the matrices of the standards and the chlorinated sample were very similar. Thus, the analytical results can be looked upon with some confidence.

q,

isotherms for AR BPL carbon before and after storage

1

k mmolg 0.0388 0.0403 0.0644 0.0478 0.0575 0.0536 0.0562 0.0475

a, 47.9 47.4 36.8 36.8 36.1 37.1 37.8 39.1

3. RESULTS

a, mm0l.g

’

26.1 26.1 24.6 24.9 23.3 23.5 24.5 24.7

AND DISCUSSION

Figure 1 shows the micropore volumes vs the surface areas of the AR, nitric acid oxidised, chlorinated and reduced BPL carbon samples before aging. The oxidised and chlorinated samples tended to have slightly lower micropore volumes and surface areas than the AR BPL carbon while the samples reduced in hydrogen, with cleaner surfaces, had higher values. These slight changes in surface area were deemed not to be significant and the pore structure of the carbon was only slightly changed by the various treatments. These values are quoted per gram of sample of adsorbent with the understanding that the BPL carbon used initially contained not only oxygen

978

S. S. BARTON et al. 0.50

0.5

0.45 0.40 i

0.35 m G 0.30 > Sz 0.25 k g 0.20 :: Q

0.15

,::u

, ,

0.10

0.2

0.0

0.4 RELATIVE

0.6

1

0.8

1 .o

0.15

1

1.0 i 2

,

:

8 ?I 0

0.8

0.6

0.4

2 s

0.2

300

500

750

900

TEMPERATURE/“C Fig. 4. Desorption of carbon dioxide and carbon monoxide from surface (open bars), AR and (striped bars), AR-aged 45 months.

groups but 5% ash. It was not determined in this study whether the surface modifying procedures and extensive washing altered the ash content, function

Table 2. Characteristics

of AR BPL carbon

Carbon

Treatment

BPL-AR BPL-AR2 BPL-AR3 BPL-red BPL-red2

0.2

0.4 RELATIVE

Fig. 3. Water adsorption isotherms at 298 K: 0, AR and 0. AR-aged 45 months, curves are fitted eqn (1).

0.00

0.0

PRESSURE

and reduced

As-received Aged 33 weeks/ 84% RH Aged 45 months/ 84% RH Reduced/h, Reduced/aged 21 months 84% RH

0.711 1.46 1.24 0.323 0.878

0.8

1 .o

PRESSURE

Fig. 5. Water adsorption isotherms at 298 K: V, AR-red and V, AR-red aged 21 months.

but the actual amount of carbon in these samples does vary. Figure 2 shows the water adsorption isotherms at 298 K for the AR carbon and three nitric acid oxidised carbons. Also shown are the isotherms on the same carbons after dry storage at ambient temperatures for over two years. The water isotherms on the stored carbons appear to reproduce the previously determined isotherms, indicating little surface oxide change on dry storage over this time period. Adams et al. report similar findings for samples stored sealed and dried for a period of 18 weeks. These BPL carbons must have a high degree of homogeneity within the batch, as is reflected in the water isotherm reproducibility on two different samples of about 100 mg selected from approximately 20 grams. The parameters of eqn (1) for the water adsorption isotherms are given in Table 1 for the carbons before and after dry storage (stored carbons indicated by S). The a, values refer to the saturation amount adsorbed, when h= 1. Clearly both sets of data are well described by a single curve. Figure 3 shows the water adsorption isotherms at 298 K on the AR carbon and the same carbon aged for 45 months at 84% RH. (The carbon was dried after aging at 110°C) The aged sample exhibits a greater affinity for water at low pressures indicating an increased concentration of primary sites on the surface. These results correspond to those of Adams et al. [7] although their measurement of water

BPL carbon

a, mmo1.g~’

0.6

before and after aging under high humidity

a, mmo1.g’ 26.1 23.9 23.5 26.4 23.8

TPD (0) mmol.g-’ 1.29 2.14 2.47 0.60 1.89

hi J.g-’ 47.2 56.5 58.3 35.3 59.1

dn

conditions kJ.mol-’ 1.8 2.4 2.5 1.3 2.5

The influence of surface modification of BPL carbons on aging i

PI 2.0 c

z

E

-F

BPL-2

BPL-2-A

979

BPL-2-s

1.8

1.6 1.4 1.2 1.0 0.8

0.6 0.4 0.2 0.0 300

500

750

800

300

500

TEMPERATURE

750

900

300

500

750

900

/“C

Fig. 6. Desorbed carbon dioxide and carbon monoxide from carbon surfaces: CO*, (clear bar) and CO, (striped bar); first series to 900°C 2 hour nitric acid oxidised carbon, second series, same carbon aged 45 months and third series, same carbon stored dry 27 months.

adsorption was from a flow of humid air through a bed of carbon, not on a degassed sample in the absence of air as in this study. In Fig. 3 the aged carbon also showed a slightly lower adsorption capacity at high relative pressures. Fitted curves are obviously quite different for the two water isotherms. TPD results on the AR and 45 month aged (BPLAR3) carbon samples are shown in Fig. 4 and indicate an increase in the measured amount of both CO and CO, evolved from the aged sample at each temperature up to 900°C when compared to the AR sample. As seen in Table 2 the measured differences in the characteristics, including the enthalpy of immersion, of the AR and aged samples are substantial. The majority of the oxidation of the AR carbon under these high humidity conditions appears have occurred in the first 8 months with little additional change over a further 37 months. No significant change in the nitrogen surface area was observed. Similar results were observed for the hydrogen reduced sample of the AR carbon, BPL-red, and its aged equivalent, BPL-red2. Figure 5 shows that upon aging this sample, the amount of water adsorbed at lower relative pressures changes significantly. For example, at a relative pressure of 0.5, the aged sample adsorbed 17 times the amount of water vapour adsorbed by the reduced sample at the same relative pressure. Again at saturation the aged sample adsorbed slightly less water vapour than did the unaged sample. After aging the reduced sample a significant increase in the amount of CO and CO2 desorbed by TPD was observed, demonstrating that re-oxidation of a clean surface can take place readily at high relative humidities and ambient temperatures. A higher enthalpy of immersion was also observed for the aged sample as noted in Table 2. The molar enthalpy of adsorption, &I, obtained by dividing the enthalpy of immersion by the saturation amount

adsorbed, a,, is a measure of the intensity of the surface-adsorbate interaction and has previously been shown to correlate with calculated a, values [ 141. The aged BPL samples have similar molar enthalpies regardless of whether or not an initial reduction was carried out. The carbons which initially had low concentrations of surface oxides, before aging, showed substantial increases in the total amount of oxygen complexes desorbed as CO and CO,, paralleled by a change in shape of the water adsorption isotherm. The previously highly oxidized carbons were less susceptible to the effects of storage in a humid environment. The TPD profiles for CO and CO, desorption from a 2 hour nitric acid oxidised carbon, BPL-2, and the complimentary aged, BPL-2-A, and stored dry samples, BPL-2-S, are shown in Fig. 6. These profiles indicate that the oxidised carbons change little on storage or aging. The total oxygen desorbed from the samples was 4.0 mmol.g-i for the sample soon after oxidation, 4.2 mmol.g- ’ after 2 years stored dry and 4.5 mmol.g-’ after almost 4 years at 84% relative humidity. The water adsorption isotherms at 298 K on three chlorine treated BPL carbon samples are shown in Fig. 7 along with the isotherm on the as-received carbon. These modified carbons were treated with chlorine for 30 minutes at temperatures from 50 to 170°C and all show a slight decrease in the saturation, a,, values. Little change may be observed in the low relative pressure portion of the isotherms, hence these chlorinated samples are not more hydrophobic than the as-received carbon. Previously it had been shown that physisorbed chlorine on carbon cloth drastically affected the shape of the water adsorption isotherm while chemisorbed chlorine had very little influence [22]. Extensive methanol washing of the chlorinated BPL carbons and drying at 110°C appeared to elimi-

980

S. S. BARTON

0.0

0.2

0.4

0.6

1 .o

0.8

1 P f

RELATIVE

PRESSURE

Fig. 7. Water adsorption isotherms at 298 K: 0, AR, n , Cl-BPL carbon 5o”C, A, Cl-BPL carbon 170°C and V, Cl-BPL carbon 100°C.

I

7

I

I

3

,

,

,

,

60

90

120

150

Reaction

C

Temperoture/“C

Fig. 8. Relationship between HCI desorbed from TPD and carbon chlorination reaction temperature.

et

al.

nate any reversibly held, physisorbed, chlorine from the samples. The enthalpy of interaction, &I, for these chlorinated BPL carbons was the same as the AR BPL carbon. The desorption of the surface complexes at temperatures reaching 900°C indicated chlorine was predominately released from the carbon surface as HCl. The total amount of chlorine desorbed from the surface of all of these BPL carbon samples however did not account for the total chlorine content as determined by neutron activation analysis. This was not unexpected as the stability of carbon-chlorine complexes is well known [27]. The amount of chlorine released from the samples on temperature programmed desorption to 900°C did appear to change with the temperature of the initial carbon-chlorine reaction, suggesting the presence of various forms of surface chlorine compounds with differing stabilities. Hall and Holmes [3] have previously suggested that the chemisorption of chlorine takes place in at least two distinct ways. The relationship between temperature of reaction and hydrogen chloride desorbed for a number of chlorine treated BPL carbon samples is shown in Fig. 8. Neutron activation analysis of these carbon samples indicate little difference in their chlorine content, 1.38 kO.074 mmol.gg’. Figure 9 shows the TPD profile of desorbed CO and CO, from a sample of BPL carbon chlorinated at 170°C and for the same sample aged 26 weeks at 84% R.H. Admittedly the aging of this sample was of a much shorter duration than some of the previous examples, but only small changes in the carbon were evident from TPD of the surface oxides, water adsorption and molar enthalpy of immersion as shown in Table 3. Comparing the changes in the AR sample that was aged 33 weeks, BPL-AR, to the chlorinated sample aged for a similar period of time, 26 weeks, Cl-BPL-2, differences are evident. Upon aging of the nonchlorinated sample the concentration of desorbed surface oxides increased 66% while the increase in the oxides from the chlorinated sample was limited to 6.5%. Calculated primary site concentrations, a,, from the water isotherms are more than double for the non-chlorinated carbon but show only a slight increase for this chlorinated carbon. The enthalpy of immersion in water for the chlorinated aged sample was about 30% less than for the non-chlorinated aged sample while the saturation amounts of water adsorbed are very similar. The water adsorption isotherms for the aged AR and aged chlorinated carbons are shown in Fig. 10. It is worth noting that the concentration of desorbed oxygen from TPD experiments indicates that all the chlorinated samples in this study contain as much oxygen (eg. CL-BPL-2, 1.23 mmol.g-‘) as the as-received sample (BPL-AR, 1.29 mmol.g-‘) contrary to the assumption by Hall and Holmes [2] that the oxygen-containing surface groups were replaced by chlorine. Included in Table 3 are some preliminary studies

The influence

of surface modification

of BPL carbons

500

750

981

CL-EPL-2-A

CL-EPL-2

300

on aging

900

500

300

TEMPERATURE

750

900

/“C

Fig. 9. Desorbed carbon dioxide and carbon monoxide from carbon surfaces: CO,, (clear bar) and CO, (striped series to 9OO’C, 170°C chlorinated BPL carbon and second series, same carbon aged 26 weeks. Table 3. Characteristics Carbon

of chemically

Treatment

Cl-BPL-2 Cl-BPL-2-A BPL-C,H, BPL-C,H,-A

treated

carbons

a, mmol.g-’

before and after aging under high humidity TPD (0) mmol.g-’

As-produced Cl-170°C Aged 26 weeks/84% RH As-produced C,H, 165°C

0.95 1.1 0.86

1.23 1.31 0.777

Aged 11 weeks/84% RH

1.1

1.22

TPD (Cl) mmol.g-’

bar); first

conditions dI? kJ.mol-’

0.76 0.67

1.7 1.8 1.5

1.8

0.45 0.40 0.35

7 01

0.30

-

0.25

-

0.20

-

0.15

-

0.10

-

7

0, > 2 I7 rY : 2

ol &

0.30

2

0.25

i

0.20

0

kj

0.15

-

0.10

-

0.05 0.0

0.2

0.4

0.6

0.8

1 .o

-.

0.00 0.0

RELATIVE

PRESSURE

Fig. 10. Water adsorption isotherms 33 weeks and V, Cl-BPL-2-A

298 K: A, AR2 aged aged 26 weeks.

on the ethylene treated BPL carbon showing similar aging effects. After treatment of the carbon with ethylene at 165°C for 5 hours, the concentration of oxides desorbed from the surface and the enthalpy of immersion were both reduced, but after only 11 weeks of aging at 84% R.H. these changes had been reversed. Contrary to the aged reduced BPL carbons and the

0.2

0.4 RELATIVE

0.6

0.8

1 .o

PRESSURE

Fig. 11. Water adsorption isotherms 298 K: V, BPL-C,H, and 0, BPL-C,H,-A aged 11 weeks.

aged AR carbons, which both showed a decrease in the saturation amount of water adsorbed, a,, upon aging, the entire water adsorption isotherm of the ethylene-treated aged carbon shifted upwards to greater adsorption at lower relative pressures, shown in Fig. 11. This effect is, as yet, unexplained.

S. S. BARTONet al.

982 4. CONCLUSIONS

AND SUMMARY

Chlorination of the carbon surface at relatively low temperatures (< 180°C) followed by extensive washing in methanol to remove any physisorbed chlorine species appears to limit the rate of aging of BPL carbon in a high humidity environment. Some fixed chlorine species on the surface that cannot be removed by heating in a vacuum to 900°C may limit the available sites for atmospheric oxidation at high RH. Further investigations will include a low temperature pretreatment of the surface with ethylene, which has been shown to reduce the concentration of oxides desorbed during TPD experiments and reduce the enthalpy of immersion, followed by a chlorination step in an attempt to produce a hydrophobic stable surface. REFERENCES 1. M. Smisek and S. Cerny, In Active Carbon p, 145. Elsevier, New York (1970). 2. C. R. Hall and R. J. Holmes, Carbon 30, 173 (1992). 3. C. R. Hall and R. J. Holmes, Carbon 31, 881 (1993). 4. B. H. M. Billinge, J. B. Docherty and B. J. Bevan, Carbon 22, 83 (1984). 5. V. R. Dietz, Carbon 25, 31 (1987). 6. S. K. Verma, P. L. Walker, Jr, Carbon 30, 837 (1992). 7. L. B. Adams, C. R. Hall, R. J. Holmes and R. A. Newton, Carbon 26, 451 (1988). 8. A. King, J. Chem. Sot. 842, (1933).

9. C. Pierce, R. N. Smith, J. W. Wiley and H. Cordes, J. Amer. Chem. Sot. 73, 4551 (1951). 10. R. N. Smith, C. Pierce and C. D. Joel, J. Phys. Chem. 58, 298 (1954). 11. S. S. Barton, M. J. B. Evans, J. Holland and J. E. Koresh, Carbon 22, 265 (1984). 12. H. F. Stoeckli, F. Kraehenbuehl and D. Morel, Carbon 21, 589 (1983). 13. M. J. B. Evans, Carbon 25, 81 (1987). 14. S. S. Barton, M. J. B. Evans and J. A. F. MacDonald, Carbon 29, 1099 (1991). 15. P. J. M. Carrott, Carbon 30, 201 (1992). 16. M. M. Dubinin and V. V. Serpinsky, Carbon 19, 402 (1981). 17. S. S. Barton, M. J. B. Evans and J. A. F. MacDonald, Carbon 30, 123 (1992). 18. S. S. Barton, M. J. B. Evans and J. A. F. MacDonald, Adsorption Science and Technology 10, 75 (1993). 19. R. C. Bansal, J.-B. Donnet and F. Stoecki, In Actiue Carbon, p. 180. Marcel Dekker, New York (1988). In Symposium on Carbon, 20. B. R. Puri and R. CBansal, Vlll-3-1. Carbon Society of Japan, Tokyo (1964). 21. D. W. McKee and C. L. Spiro, Carbon 23, 437 (1985). 22. S. S. Barton, M. J. B. Evans, J. E. Koresh and H. Tobias, Carbon 25, 663 (1987). M. Islam, R. W. Coughlin, K. Mans23. P. Schoderboeck, field-Matera and E. Davis, Carbon 31, 1351 (1993). 24. S. Brunauer, P. H. Emmett and E. Teller, J. Amer. Chem. Sot. 60, 309 (1938). 25. M. M. Dubinin, E. D. Zaverina and L. V. Radushkevich, Zh. Fiz. Khim. 21, (1947). 26. S. S. Barton, M. J. B. Evans. B. H. Harrison and J. R. Sellors, J. C&l. & Polymer Sci. 260, 726 (1982). 27. B. R. Puri, In Chemistry and Physics of Carbon, Vo1.6, p. 258. Marcel Dekker, New York (1970).