Diffusion of hydrogen sulfide and methyl mercaptan onto microporous alkaline activated carbon

Di€usion of hydrogen sul®de and methyl mercaptan onto

microporous alkaline activated carbon

Hung-Lung Chiang

, Jiun-Horng Tsai

, Dai-Huang Chang

, Fu-Teng Jeng

a_{Department of Environmental Engineering, Fooyin Institute of Technology, Kaoshiung Hsien, Taiwan, ROC} b_{Graduate Institute of Environmental Engineering, National Cheng Kung University, Tainan, Taiwan, ROC}

c_{Graduate Institute of Environmental Engineering, National Taiwan University, Taipei, Taiwan, ROC}

Received 8 September 1999; accepted 17 November 1999

Abstract

Activated carbon kinetic studies show that both H2S and CH3SH yielded pore di€usion coecients from 10ÿ6 to

10ÿ8_cm2_{/s. Results indicated that pore structures could in¯uence e€ective di€usivity. Under the same adsorbate}

con-centration, CH3SH exhibited a greater e€ective pore di€usion coecient than H2S. This may be attributed to the fact

that CH3SH has both polar (±SH) and non-polar (±CH3) functional groups and dissolves into water easier, thus

providing more attraction for the activated carbon surface. In addition, the saturation vapor pressure of CH3SH is

lower than that of H2S. Therefore, CH3SH is easier to adsorb onto activated carbon than H2S. Ó 2000 Elsevier Science

Ltd. All rights reserved.

Keywords: Di€usivity; Hydrogen sul®de; Methyl mercaptan; Alkaline activated carbon

1. Introduction

Gaseous odors are emitted from various plants, such as sewage treatment plants and chemical industries (Vigneron et al., 1994). Odors from such plants are being treated by adsorption on activated carbon or impreg-nated activated carbon, scrubbing and neutralizing by chemical solution, condensation, mashing or catalytic combustion. Adsorption, however, is used most often to remove H2S and CH3SH from gas streams (Turk et al.,

1989; Koe and Tan, 1990). Therefore, several impreg-nated activated carbons have been developed for de-odorization (Tsutsui and Tanada, 1987; Ikeda et al., 1988; Tsai et al., 1992, 1999).

When removing adsorbate from air mixtures ¯owing into granular activated carbon, the following sequence

of reaction steps is possible (Jonas, 1978): mass transfer, surface di€usion, intragranular di€usion, physical ad-sorption, gas dead-sorption, chemical reaction and surface renewal. If the products of a chemical reaction are vol-atile and poorly adsorbed, they leave the macropore region and enter the ¯uid stream surrounding the carbon granule where they are swept away from the activated carbon by mass transport.

When Tien (1994) modi®ed the results of Jonas, he noted that di€usion was the predominant mechanism in activated carbon sorption. Recent research on sorption by dry soil grains revealed that di€usion processes govern the rate at which gas-phase species reach the intragranular soil surface. The structural similarity be-tween porous soil grains and activated carbon suggests similar mechanisms in both cases. Several related stud-ies support this conclusion (Garg and Ruthven, 1972; Ruthven and Derrah, 1972; Gray and Do, 1989a,b, 1990).

The purpose of this study was to investigate the physicochemical characteristics of ®ve activated

*_{Corresponding author. Tel.: 886-7-782-8663; fax:}

+1-886-7-782-9117.

E-mail address: hlchang@cc.fy.edu.tw (H.-L. Chiang).

0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 5 - 6 5 3 5 ( 9 9 ) 0 0 5 4 7 - 0

carbons. External di€usion, macropore di€usion, and micropore di€usion of H2S and CH3SH on the ®ve

ac-tivated carbons were calculated. 2. Theory

Three solid surface adsorption mechanisms were in-vestigated: external di€usion and internal di€usion (macropore di€usion and micropore di€usion). These mechanisms a€ected the pore structural distribution of the adsorbent (Yang, 1987).

2.1. External di€usion

External di€usion is the gaseous reactant transfer from the bulk gas stream to the external surface of the solid particle. The mass ¯ux being transferred from the gas stream to solid surface is given by Noll et al. (1992). NAˆkf_qape

p … ÿ CC s†; …1†

where NAis the mass ¯ux of the transferred species from

gas stream to adsorbent surface, kf the external ®lm

mass transfer coecient, apthe external surface area of

adsorbent, e the voids between adsorbents, qp the bulk

density of adsorbent, C the concentration of the trans-ferred species in the bulk of gas stream, and Cs the

concentration of the transferred species at the adsorbent surface.

For packed bed adsorption, the external ®lm mass transfer coecient (kf) was correlated by the Ranz±

Marshall equation following Eq. (2) (Noll et al., 1992): Sh ˆ2k_Dfrp

m ˆ 2:0 ‡ 0:6 Re

0:5 _Sc0:33_{; …2†}

where Dm is the molecular di€usivity, rp the radius of

adsorbent, Re the Reynolds number …Re ˆ …2rpVsqf=l††,

Sc the Schmidt number …Sc ˆ l=…Dmqf††, Sh the

Sher-wood number …Sh ˆ …2kfrp†=Dm†; l the viscosity of gas

stream, Vs the velocity of gas stream, and qf is the

density of gas stream. 2.2. Internal di€usion 2.2.1. Macropore di€usion

The importance of heterogeneous catalysis in mac-ropore di€usion has been a popular research subject. There are four kinds of macropore di€usion: (1) mo-lecular di€usion, (2) Knudsen di€usion, (3) Poiseuille ¯ow di€usion, and (4) surface di€usion. The di€usion mechanisms have been shown to a€ect adsorbent pore structure distribution (Ruthven, 1984).

The mathematical relationships for each macropore di€usion are shown below:

(1) Molecular di€usion:

DpˆD_sm; …3†

where Dp is the pore di€usivity, Dm the molecular

dif-fusivity and s is the tortousity factor.

Fuller et al. derived the molecular di€usivity (Dm) of

a two-component system as follows (Bird et al., 1960; Sherwood et al., 1975; Hines and Maddox, 1985; Szekely et al., 1976): Dmˆ 1:0  10 ÿ3_{ T}1:75 P …Pt†1=3 A ‡ P t … †1=3 B h i2 1 MA  ‡_M1 B 1=2 ; …4† where P is the pressure of gas stream, T the absolute temperature, …Pt†Aand …

P

t†B are the molecular

vol-umes of gases A and B, and MA and MB are the gas

molecular weights of A and B. Generally, the mean free path of molecules is smaller than the pore diameter of the adsorbent. The transport mechanism is therefore molecular di€usion.

(2) Knudsen di€usion: Knudsen showed that under these conditions the di€usivity per unit cross-sectional area of pore is given by

Dkˆ2₃rmÿˆ 9700rp  T M r …5† where Dk, rp, m , T and M are Knudsen di€usivity, pore

radius, mean molecular velocity, absolute temperature and molecular weight, respectively.

(3) Surface di€usion: The overall di€usion coecient is shown as:

D ˆ Dk‡ 1 ÿ e_e p p

 

KDs; …6†

where D is the overall di€usivity, Dk the Knudsen

dif-fusivity, ep the adsorbent porosity, K the equilibrium

constant, and Dsis the surface di€usivity.

2.2.2. Micropore di€usion

According to FickÕs law, one has

N ˆ ÿDoC_or; …7†

where N is the mass ¯ux of transferred species, D the e€ective di€usivity of transferred species, C the con-centration of transferred species, and r is the distance normal to the external surface.

The micropore di€usion is derived as follows (Noll et al., 1992): oq ot ˆ 1 r2 o or r2Dc oq or   : …8†

Assuming Dc is constant, Eq. (8) can be derived as Eq. (9). oq ot ˆ Dc o 2_q or2  ‡2_roq_or  : …9†

The initial and boundary condition are shown as fol-lows: q…r; 0† ˆ q0; q…rc; t† ˆ q0; oq_or   rˆ0 ˆ 0; then q ÿ q0 q0ÿ q0ˆ Mt M1ˆ 1 ÿ 6 p2 X1 nˆ1 1 n2exp  ÿn2p_r2₂Dct c  …10† and q ˆ 3 r3 c Z rc 0 q
r 2_{dr; …11†}

where q is the amount adsorbed (per unit volume of sorbent), q the average amount adsorbed in a pellet or particle, M1 the total mass uptake at a gas-phase

con-centration of C0and Mhis the cumulative mass uptake at

dimensionless time h. 3. Experimental

This research selected spent activated carbon (SAC), regenerative activated carbon (RAC), fresh activated carbon (FAC), impregnated-regenerative activated car-bon (RAC-N), and impregnated-fresh activated carcar-bon (FAC-N) as the experimental adsorbents. The activated carbons were made from coconut shell.

3.1. Regeneration of spent activated carbon

50 g of SAC (Chinese Carbon, Taiwan) was placed into a vacuum oven (10ÿ1_±10ÿ2 _{mmHg) at room}

tem-perature, under nitrogen (99.95%), for 2 h. Next, 10 g of the pretreated SAC was put into an oven at 400°C for 1.5 h. The high purity nitrogen gas was used to cool the oven. The regenerative-spent activated carbon was stored in gas-sealed vials until experimentation. 3.2. Preparation of impregnated activated carbons

50 g of activated carbon (RAC or FAC) was placed into an oven and dried with ¯owing nitrogen gas (140°C) for 6 h. The pretreated activated carbons were immersed in 500 mlÿ1 _{N NaOH solution (Merck, Germany) and}

stirred for 30 min. The immersed activated carbons were kept in a vacuum oven for 30 min and in a dryer (glass container with silica gel) for 200 min (stationary times)

at room temperature. The immersed activated carbons were then ®ltered from the impregnation solutions and dried in an oven at 130°C for 60 h. The prepared alka-line activated carbons were stored in gas tight, nitrogen gas-®lled containers before use.

3.3. Physical characteristics

The activated carbons were stored in an oven at 105°C and dried for 48 h. The physical characteristics of the activated carbon, including speci®c surface area, micropore area, total pore volume, micropore volume and pore diameter were measured with N2…g† adsorption

using an ASAP 2000 Micropore Analyzer (Micromeri-trics, USA) at 77 K using liquid N2.

3.4. Amount of alkaline on activated carbon

The quantity of alkaline or NaOH on each activated carbon was determined by extraction and titration processes. The impregnated activated carbons were placed in a vacuum oven (1±10ÿ1_{mm Hg, 105°C) for 24}

h. HCl solution (1 N) was then added, and the mixture stored at 25°C for 24 h. To separate the supernatant, the sample was centrifuged at 3000 rpm for ten min. The supernatant was then titrated with 1 N NaOH solution until pH 7.

3.5. Sorption experiment

A simulated blend of cylinder gases was passed through a glass column 20 cm in length and 28 mm in diameter. The bottom of the adsorption column was packed with a layer of 10-cm glass beads. 5±10 g of activated carbon (radius 0.5 mm) was packed in the column for each run. Cylinder gases of H2S (8000 ppm)

and CH3SH (99.9%) were certi®ed by suppliers (Scott

Gas Company, USA). The in¯uent concentrations of H2S and CH3SH ranged from 30±200 ppmv in the

ad-sorption system. The ¯ow rate, which ranged from 2.0± 10 l/min, was controlled by a mass ¯ow meter (Sierra Series 9000, USA).

H2S and CH3SH stream constituents were analyzed

by a Gas Chromatograph (HP-6890) equipped with a Pulsed Flame Photometric Detector (PFPD) and chro-matographic column (G.S.Q.: 30 m, B: 0.53 mm). Temperature of injector, column, and detector were 180, 150, and 220°C, respectively. Retention times of H2S

and CH3SH were 4.1 and 5.93 min, respectively. The

adsorption capacities of H2S and CH3SH on each

acti-vated carbon in the gas stream were analyzed by the GC/ PFPD and calculated by column adsorption kinetic curves. Quality control was also conducted to ensure experimental data performance.

4. Results and discussion

4.1. Physico±chemical characteristics of activated carbons Physico±chemical characteristics of the ®ve activated carbons are shown as Table 1. The sequence of BET surface area, micropore area, pore volume and micro-pore volume were as follows: FAC > FAC-N > RAC > RAC-N > SAC. The pore diameter distribution was SAC > RAC > RAC-N > FAC  FAC-N. The amount of alkaline equivalent on the activated carbons was FAC-N > RAC-N > FAC > RAC > SAC. Results indi-cated that SAC was the least e€ective adsorbent of the ®ve activated carbons. NaOH impregnation changed their physicochemical characteristics.

4.2. External di€usion

At 298 K, the viscosity of air, H2S, CH3SH, air-H2S

gas mixture and air-CH3SH gas mixture are 184.6,

126.5, 95.8, 170.5 and 143.5 lP, respectively. The Rey-nolds number, Schmidt number and Sherwood number were 16.1, 1.1, and 8.0 for the H2S-air mixture system,

and 12.0, 2.0 and 8.1 for the CH3SH-air mixture system.

The external ®lm mass transfer H2S and CH3SH

coef-®cients were 14.8 and 11.2 cm/s, respectively. 4.3. Macropore di€usion

The di€usion of adsorbates in macropore-molecular di€usion, surface di€usion, and Knudsen di€usion, were considered the di€usion mechanisms. Results of the H2S

and CH3SH macropore di€usivity into the ®ve activated

carbons (SAC, FAC, RAC, RAC-N, and FAC-N) are shown in Table 2.

4.3.1. Molecular di€usion

According to Eq. (4), molecular di€usivity a€ects the temperature, pressure, molecular weight and adsorbate characteristics. In this research, the molecular di€usivity of the Air-H2S and Air-CH3SH systems were 0.185 and

0.138 cm2_{/s, respectively. The molecular weight and}

volume of H2S is smaller than CH3SH, so the molecular

di€usivity of H2S was greater than CH3SH.

4.3.2. Knudsen di€usion

When the mean free path of the adsorbate is greater than the pore diameter of the activated carbon, the collision of molecules and the pore wall can inhibit molecule transport. The Knudsen di€usivity is a func-tion of the pore radius, temperature and molecular weight (Eq. (5)). Analysis indicated that the Knudsen di€usivity of H2S was between 4:33  10ÿ3 and

5:04  10ÿ3 _cm2_{/s and that of CH}

3SH was between

3:51  10ÿ3 _{and 4:24  10}ÿ3 _cm2_{/s. The Knudsen}

di€u-sivity was directly proportional to the pore radius under identical adsorbate and adsorption temperature condi-tions. Since CH3SH molecular weight is greater than

H2S, the Knudsen di€usivity of H2S was greater than

CH3SH in an adsorption system.

4.3.3. Di€usion of transient region

The di€usivity of the transient region can be calcu-lated as

Table 1

Physico±chemical characteristics of activated carbons

Adsorbents BET surface

area (m2_/g) Micropore_{area (m}2_/g) Pore volume_(cm3_/g) Micropore_{volume (cm}3_/g) Pore_{diameter (A)} Alkaline_{equivalent (meq/g)}

SAC 285 186 0.170 0.088 17.6 0.75 RAC 794 610 0.429 0.282 16.1 1.75 FAC 1282 864 0.623 0.394 14.6 1.88 RAC-N 664 505 0.338 0.238 15.1 2.63 FAC-N 1188 789 0.581 0.335 14.5 3.13 Table 2

Macropore di€usion of H2S and CH3SH on activated carbons

Adsorbents H2S (cm2/s) CH3SH (cm2/s) Dk 103 D  103 Dk 103 D  103 SAC 4.33 4.23 3.64 3.55 RAC 4.63 4.52 3.90 3.79 FAC 4.19 4.09 3.52 3.44 RAC-N 5.04 4.91 4.24 4.11 FAC-N 4.18 4.08 3.51 3.43

1 Dˆ 1 Dk‡ 1 Dm 1  ÿ 1  ‡_NNB A  XA  ; …13†

where D is di€usivity of the transient region, NAand NB

are the molecular transport ¯ux of A and B, and XAis the

molar fraction of A. Assuming the molecular transport ¯ux of A is equal to that of B…NAˆ NB†, the equation

simpli®es to 1 Dˆ 1 Dk‡ 1 Dm: …14†

The di€usivity of H2S was between 4:08  10ÿ3 and

4:91  10ÿ3 _cm2_{/s and that of CH}

3SH was between

3:43  10ÿ3_{and 4:11  10}ÿ3_cm2_{/s in the transient region.}

4.4. Micropore di€usion

Governing model equations with similar boundary and initial conditions have been solved by several in-vestigators. In this research, a numeric method and Fortran least squares program were used to solve the governing equation.

The adsorption di€usivity of H2S and CH3SH on the

®ve activated carbons are shown as Table 3.

4.4.1. E€ective di€usivity of H2S adsorbed on activated

carbon

The e€ective di€usivity of H2S that was adsorbed on

activated carbon was between 2:1  10ÿ8 _and

6:49  10ÿ6 _cm2_{/s. When the in¯uent concentration of}

H2S was increased from 50 to 100 ppmv, the e€ective

di€usivity increased: SAC from 5:35  10ÿ6 _to

6:49  10ÿ6_cm2_{/s, RAC from 1:02  10}ÿ7_{to 1:19  10}ÿ7

cm2_{/s, FAC from 7:3  10}ÿ8_{to 8:8  10}ÿ8_cm2_/s,

RAC-N from 2:1  10ÿ8_{to 2:7  10}ÿ8_cm2_{/s, and FAC-N from}

2:1  10ÿ8 _{to 4:0  10}ÿ8_cm2_/s.

4.4.2. E€ective di€usivity of CH3SH adsorbed on

activat-ed carbon

When the in¯uent concentration of CH3SH was

creased from 50 to 100 ppm, the e€ective di€usivity in-creased: SAC from 5:02  10ÿ6 _{to 5:78  10}ÿ6 _cm2_/s,

RAC from 1:36  10ÿ7_{to 1:42  10}ÿ7_cm2_{/s, FAC from}

3:21  10ÿ7 _{to 3:95  10}ÿ7 _cm2_{/s, RAC-N from}

7:2  10ÿ8 _{to 8:7  10}ÿ8 _cm2_{/s, and FAC-N from}

1:03  10ÿ7_{to 1:27  10}ÿ7_cm2_/s.

Results indicated that the larger the pore diameter, the larger the e€ective di€usivity and the higher the in-¯uent concentration, the higher the e€ective di€usivity.

The e€ective di€usivity of H2S and CH3SH on

RAC-N from 30 to 200 ppmv is shown as Fig. 1.

When the in¯uent concentration of H2S was

in-creased from 30 to 200 ppmv, the e€ective di€usivity of H2S adsorbed on RAC-N increased from 1:63  10ÿ8to

3:89  10ÿ8 _cm2_{/s. When the in¯uent concentration of}

CH3SH was increased from 30 to 200 ppmv, the e€ective

di€usivity of CH3SH adsorbed on RAC-N ranged from

5:31  10ÿ8_{to 1:03  10}ÿ7_cm2_/s.

Given the same in¯uent concentrations of H2S and

CH3SH, the e€ective di€usivity of CH3SH was greater

than that of H2S. This can be attributed to methyl

mercaptan having both a polar functional group (S±H) and a non-polar functional group (C±H). Methyl mer-captan tends to be adsorbed more easily on activated carbon than H2S. Additionally, a high concentration of

adsorbate in the micropores of the activated carbon may increase the van der WaalÕs force (due to condensation). The saturation vapor pressure is 19.807 atm for H2S and

Table 3

E€ective di€usivity of H2S and CH3SH on activated carbons

Adsorbents H2S (cm2/s) CH3SH (cm2/s) 50 ppmv 100 ppmv 50 ppmv 100 ppmv SAC 5:35  10ÿ6_0:96†a _{6:49  10}ÿ6_{0:97† 5:02  10}ÿ6_{0:90† 5:78  10}ÿ6_0:91† RAC 1:02  10ÿ7_{0:91† 1:19  10}ÿ7_{0:99† 1:36  10}ÿ7_{0:92† 1:42  10}ÿ7_0:90† FAC 2:1  10ÿ8_{0:92† 8:8  10}ÿ8_{0:97† 3:21  10}ÿ7_{0:92† 3:95  10}ÿ7_0:92† RAC-N 2:7  10ÿ8_{0:89† 2:7  10}ÿ8_{0:94† 7:20  10}ÿ8_{0:91† 8:7  10}ÿ8_0:92† FAC-N 2:1  10ÿ8_{0:92† 4:0  10}ÿ8_{0:94† 1:03  10}ÿ7_{0:92† 1:27  10}ÿ8_0:91†

a_{Regression coecient of adsorption curve (experiments) and ®tting curve (di€usion model).}

Fig. 1. E€ective di€usivity of H2S and CH3SH onto RAC-N.

is 1.943 atm for CH3SH at 25°C (Reid et al., 1988).

These values indicate that the CH3SH is more easily

condensed in the micropores of activated carbon. Therefore, the e€ective di€usivity of CH3SH is larger

than that of H2S.

5. Conclusions

In¯uent concentrations of H2S and CH3SH were 30

to 200 ppmv, the e€ective di€usion coecients were 10-8

to 10-6 _cm2_{/s. CH}

3SH exhibited an e€ective pore

di€u-sion coecient greater than that of H2S at the same

adsorbate concentration. This may be attributed to the fact that CH3SH has both polar (±SH) and non-polar

(±CH3) functional groups with a strong anity toward

the activated carbon surface. Furthermore, the satura-tion vapor pressure of CH3SH (1.943 atm) is lower than

that of H2S (19.807 atm) at 298 K. This indicates that

CH3SH is easier to condense in micropore carbon than

H2S. Results indicate that in¯uent concentration and

pore structure can a€ect the e€ective di€usivity. Acknowledgements

The authors express their sincere thanks to the National Science Council, Taiwan, ROC for its support (NSC 87-2211-E-006-004 and NSC-88-2211-E-006-009) of this study.

References

Bird, R.B., Stewart, W.E., Lightfoot, E.N., 1960. Transport Phenomena. Wiley, New York.

Garg, D.R., Ruthven, D.M., 1972. The e€ect of the concentra-tion dependence of di€usivity on zeolite sorpconcentra-tion curves. Chem. Eng. Sci. 27, 95±99.

Gray P.G, Do, D.D., 1989a. Adsorption and desorption of gaseous sorbates on a bidispered particle with freundilch isotherm: I. theory analysis. Gas Sep. Purif. 3, 193±200. Gray, P.G., Do, D.D., 1989b. Adsorption and desorption of

gaseous sorbates on a bidispered particle with freundilch isotherm: II experimental study of sulphur dioxide sorption on activated carbon particles. Gas Sep. Purif. 3, 201±208.

Gray, P.G., Do, D.D., 1990. Adsorption and desorption dynamics of sulphur dioxide on a single large activated carbon particles. Chem. Eng. Comm. 96, 141±154. Hines, A.L., Maddox, R.N., 1985. Mass Transfer:

Fundamen-tals and Application. Prentice-Hall, New Jersey.

Ikeda, H., Asaba, H., Takeuchi, Y., 1988. Removal of H2S,

CH3SH and (CH3)3N from air by use of chemically treated

activated carbon. J. Chem. Eng. Jpn. 21, 91±97.

Jonas, L.A., 1978. Reaction steps in gas sorption by impreg-nated carbon. Carbon 16, 115±119.

Koe, L.C.C., Tan, N.C., 1990. Comparison of ®eld and laboratory H2S adsorption capacity of activated carbon.

Water Air Soil Pollut. 50, 969±976.

Noll, K.E., Gounaris, V., Hou, W.S., 1992. Adsorption Technology for Air and Water Pollution Control. Lewis Publishers, Michigan.

Reid, R.C., Prausnite, J.M., Poling, B.E., 1988. The Properties of Gases and Liquids, fourth ed., McGraw-Hill, New York. Ruthven, D.M., 1984. Principles of Adsorption & Adsorption

Process. Wiley, New York.

Ruthven, D.M., Derrah, R.I., 1972. Sorption in davison 5a molecular sieves. Can. J. Chem. Eng. 50, 743±747. Sherwood, T.K., Pigford, R.L., Wikle, C.R., 1975. Mass

Transfer. McGRAW-Hill, New York.

Szekely, J., Evans, J.W., Sohn, H.Y., 1976. Gas-Solid Reac-tions. Academic Press, New York.

Tien, C., 1994. Adsorption Calculations and Modeling. But-terworth-Heineman, Maryland.

Tsai, J.H., Chiang, H.L., Tsai, C.L., Hsu, Y.C., 1999. Adsorp-tion of hydrogen sul®de and methyl mercaptan mixture gas on alkaline activated carbon. J. Chin. Inst. Environ. Eng., in press.

Tsai, J.H., Jeng, F.T., Yen, S.H., 1992. Treatment of NH3and

H2S mixture gas by sorption. J. Chin. Inst. Environ. Eng.

22, 211±218.

Tsutsui, S., Tanada, S., 1987. Adsorption of hydrogen sul®de, dimethyl sul®de and their binary mixture into pores of n-containing activated carbon. Chem. Pharm. Bull. 35, 1238± 1242.

Turk, A., Sakalis, E., Lessuck, J., Karamitsos, H., Rago, O., 1989. Ammonia injection enhances capacity of activated carbon for hydrogen sul®de and methyl mercaptan. Envi-ron. Sci. Technol. 23, 1242±1245.

Vigneron, S., Hermia, J., Chaouki, J., 1994. Characterization and Control of Odours and VOC in the Processes Indus-tries. Elsevier, Amsterdam.

Yang, R.T., 1987. Gas Seperation by Adsorption Processes. Butterworth, Boston.