Carbon 43 (2005) 1039–1048

www.elsevier.com/locate/carbon

Effect of adsorbent composition on H2S removal on sewagesludge-based materials enriched with carbonaceous phase

Adil Ansari, Andrey Bagreev, Teresa J. Bandosz *

Department of Chemistry, The City College of New York, 138th Street and Convent Ave, New York, NY 10031, USA

Received 28 April 2004; accepted 22 November 2004

Available online 25 January 2005

Abstract

In order to improve the carbonaceous phase content in sewage sludge derived adsorbents, dewatered sludge was physically mixed

with polystyrene sulfonic acid-co maleic acid sodium salt with the following ratios of polymer to sludge: 10:90, 30:70, 50:50 and

70:30. The samples, along with the pure precursors, were carbonized at 950 �C and then washed in water to remove the excess

of sodium salt/hydroxide. The performance of materials as H2S adsorbents was tested using a home-developed dynamic break-

through test. The samples, before and after adsorption process, were characterized by adsorption of nitrogen, potentiometric titra-

tion, thermal analysis and SEM. Differences in the performance were linked to the surface properties. It was found that mixing

polymer with sludge increases the amount of H2S adsorbed/oxidized in comparison with the adsorbents obtained from pure precur-

sors (sludge or polymer). Sewage sludge provides the catalytic centers for hydrogen sulfide oxidation whereas a carbonaceous phase

contributes to an increase in the dispersion of catalytic centers and provides more ‘‘storage space’’ in its micropores. There is an

optimal ratio in the composition of the precursors for which the best performance is achieved. When the content of the polymer

is too high, the ‘‘buffering capacity’’ of the sludge-derived phase is not enough to neutralize the suppressing effect of acidic surface

groups of a carbonaceous phase.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Activated carbon; Adsorption properties

1. Introduction

Acidic gases pollution creates environmental prob-

lems in the US, especially in the North East where acid

rains occur very frequently damaging the natural ecosys-

tems [1]. The main source of SO2 has anthropogenic ori-gin in fossil fuel burning. In spite of strict governmental

regulations (Acid Rain Program of Clean Air Act of

1990) still about 10 million tons of SO2 are generated

every year from power plants [2]. Another sulfur con-

taining gas, which contributes to acid rain is hydrogen

0008-6223/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.carbon.2004.11.042

* Corresponding author. Tel.: +1 212 650 6017; fax: +1 212 650

6107.

E-mail address: tbandosz@ccny.cuny.edu (T.J. Bandosz).

sulfide. Although there are many natural origins of

H2S (anaerobic decay of organic matter, volcanoes,

etc.) the municipal facilities are the significant sources

of its atmospheric emissions.

So far, to clean air from low concentration of sulfur

containing gases and volatile organic compounds(VOCs) (polishing) activated carbon-based adsorbents

are used [3–5]. Very often additional modifications are

applied to enhance carbon capacity for deodorization.

Impregnation with caustic is just one example [5–8].

According to EPA, 8.2 million tons of biosolids will

be generated in the US in 2010 [9]. So far the prognosis

predict that 70% will find beneficial use [land application

(48%), advance treatment (13.5%), other beneficial use(8.5%)] and 30% will be disposed [surface disposal/land-

fill (10%), incineration (19%), other (1%)].

1040 A. Ansari et al. / Carbon 43 (2005) 1039–1048

Adsorbents obtained by pyrolysis of sewage sludge

can be considered as complex pseudocomposite materi-

als. However, the process of carbonization of biosolids

has been studied in detail previously and it is described

in the literature [10–15] so far the most promising results

were obtained in our laboratory [16–20]. Recently, it hasbeen shown that by simple pyrolysis of sewage sludge

derived fertilizer, Terrene�, exceptionally good adsor-

bents for removal of sulfur containing gases can be ob-

tained. Their removal capacity is twice that of coconut

shell-based activated carbon [17]. The specific surface

feature of those materials resulting in the high H2S re-

moval capacity is still not identified. The predominant

influence of inorganic phase or combination of oxides,which are also quite commonly used as catalysts for

hydrogen sulfide oxidation or sulfur dioxide adsorption,

was ruled out since the capacity of pure inorganic phase

heated at 950 �C was found to be negligible [17]. The

data also showed that the oxidation of hydrogen sulfide

occurs until all micropores (mainly about 6 A in size),

likely within carbonaceous deposit or on the carbon/

oxide interface, are filled with the reaction products[17–20]. The form of that carbonaceous deposit is not

known but based on the previous results that deposit

is important from the point of view of the adsorption

capacity [20]. The products of oxidation immobilized

on the surface are stored there. Since pore volume seems

to be a limiting factor for the capacity of sewage sludge

derived materials, an increase in the content of carbona-

ceous deposit and pore volume with maintaining theimportant content of a catalytically active phase (sewage

sludge) seems to be the desired direction of research. In

such a case the products of oxidation would be stored in

a sufficient quantity in the pore system making the re-

moval capacity for sulfur containing gases exceptionally

high.

The objective of this paper is to evaluate the impor-

tance of the synergetic effect of surface chemistry andporosity of sewage sludge derived materials enriched

with carbonaceous phase on their performance as

hydrogen sulfide removal media. Based on the hypothe-

sis formulated above, new materials are synthesized

from the physical mixture of organic polymer and sew-

age sludge and then tested as hydrogen sulfide adsor-

bents. Their performance in the process of hydrogen

sulfide removal from moist air is linked to the surface

Table 1

Names of samples, yields on various stages, bulk density and ash content

Sample Content of dry sludge [%] Yield [%] Yield af

CS 100 40 40

CPS 10/90 69 38 31

CPS 30/70 37 43 25

CPS50/50 20 39 17

CPS 70/30 10 33 11

CP 0 35 8

features such as porosity, content of inorganic matter,

dispersion of that matter, and overall surface chemistry

expressed as the pH value. The detailed study presented

here helps to identify the synergetic effect occurring in

the case of the mixture of the particular components

used in this research.

2. Experimental

2.1. Materials

Adsorbents were prepared from dewatered sludge

(from Wards Island Water Pollution Control Plant,NYC DEP), polystyrene sulfonic acid-co maleic acid so-

dium salt and mixtures of both precursors with follow-

ing weight percentage of polymer to dewatered sludge

10:90, 30:30, 50:50, 70:30. In all cases the pyrolysis

was done in a horizontal furnace at nitrogen atmosphere

with heating rate 10 deg/min. The final pyrolysis temper-

ature was 950 �C with holding time 1 h. Since polymer

(polystyrene sulfonic acid-co maleic acid sodium salt)contains a lot of sodium, after preparation the materials

were washed in a Soxhlet apparatus to constant pH of

water leachate. The composite adsorbents are referred

to as CPS, which is followed by numbers representing

the polymer to dewatered sludge ratio. The adsorbent

obtained from dewatered sludge is referred to as CS

and carbon obtained from polymer—CP. The names

of adsorbents, and their yields are collected in Table 1.

2.2. Methods

2.2.1. Evaluation of H2S sorption capacity

A custom-designed dynamic test was used to evaluate

the performance of adsorbents for H2S adsorption from

gas streams as described elsewhere [17]. Adsorbent sam-

ples were ground (1–2 mm particle size) and packed intoa glass column (length 370 mm, internal diameter 9 mm,

bed volume 6 cm3), and pre-humidified with moist air

(relative humidity 80% at 25 �C) for 1 h. The amount

of water adsorbed was estimated from the increase in

the sample weight. Moist air (relative humidity 80% at

25 �C) containing 0.3% (3000 ppm) of H2S was passed

through the column of adsorbent at 0.5 L/min. The flow

rate was controlled using Cole Palmer flow meters. The

ter washing [%] Bulk density [g/cm3] Ash content [%]

0.55 75

0.64 65

0.28 46

0.12 32

0.09 15

0.10 1

A. Ansari et al. / Carbon 43 (2005) 1039–1048 1041

breakthrough of H2S was monitored using electrochem-

ical sensors. The test was stopped at the breakthrough

concentration of 100 ppm. The adsorption capacities

of each adsorbent in terms of mg of hydrogen sulfide

per gram of adsorbent were calculated by integration

of the area above the breakthrough curves, and fromthe H2S concentration in the inlet gas, flow rate, break-

through time, and mass of sorbent. For each sample the

test was repeated at least twice.

The adsorbents exhausted after H2S adsorption are

designated by adding an addition letter, E to their

names.

2.2.2. Characterization of pore structure of adsorbents

On the materials obtained sorption of nitrogen at its

boiling point was carried out using ASAP 2010

(Micromeritics). Before the experiments, the samples

were outgassed at 120 �C to constant vacuum

(10�4 Torr). From the isotherms, the surface areas

(BET method), total pore volumes, Vt, (from the last

point of isotherm at relative pressure equal to 0.99), vol-

umes of micropores, Vmic, mesopore volume Vmes alongwith pore size distributions were calculated. The last

three quantities were calculated using density functional

theory, DFT [21,22].

2.2.3. Study of surface chemistry

The surface properties were evaluated first using

potentiometric titration experiments [23–25]. Here, it is

assumed that the population of sites can be describedby a continuous pKa distribution, f (pKa). The experi-

mental data can be transformed into a proton binding

isotherm, Q, representing the total amount of proton-

ated sites, which is related to the pKa distribution by

the following integral equation:

QðpHÞ ¼Z 1

�1qðpH; pKaÞf ðpKaÞdpKa: ð1Þ

The solution of this equation is obtained using the

numerical procedure [26], which applies regularization

combined with non-negativity constraints. Based on

the spectrum of acidity constants and the history of

the samples, the detailed surface chemistry shall be

evaluated.

2.2.4. pH

The pH of a carbon sample suspension provides

information about the acidity and basicity of the carbon

surface. A sample of 0.4 g of dry carbon powder was

added to 20 mL of distilled water and the suspension

was stirred overnight to reach equilibrium. Then the

pH of suspension was measured.

2.2.5. Thermal analysis

Thermal analysis was carried out using TA Instru-

ment Thermal Analyzer. The instrument settings were:

heating rate 10 �C/min and a nitrogen atmosphere with

100 mL/min flow rate. For each measurement about

25 mg of a ground adsorbent sample were used.

2.2.6. SEM/EDX

Scanning electron microscopy images were obtainedat Zeiss-LEO using LEO 1550 FESEM, and on LEO

1455 VP SEM with Tungsten source in Lamont Doherty

Earth Observatory of Columbia University. The as re-

ceived samples were mounted using silver support.

3. Results and discussion

The yields of materials collected in Table 1 show that

with an increase in the content of the polymer yield de-

creases. This behavior is expected based on the very

small yield of adsorbent when only polymer is used as

a precursor [27]. Decomposition of organic acidic

groups, condensation of aromatic rings and dissolving

of sodium species during washing results in only few

percent of the initial mass left after all steps of prepara-tion. On the other hand, from dried sludge 40% weight is

preserved, which is owing to the high content of an inor-

ganic matter in the initial sludge (over 30% in dry sludge

[16,17]). After washing yields decreases for all samples

but the sludge-based one. That decrease is most pro-

nounced for the samples with a high content of polymer.

In those samples the large content of soluble sodium

species (likely chloride and hydroxide) is present [27].In the case of the CS sample the yield is almost constant

in spite of the detection of some soluble salts in the

leachate. The reason for this in a compensating effect

of washing which, besides removal of water-soluble spe-

cies, causes hydration of other compounds (mainly cal-

cium- and magnesium-based) and incorporation of

water into their chemical structure.

Table 1 also lists the density of materials and theirash content. As expected, higher ash content leads to

the higher bulk density of the samples. The polymer-de-

rived sample is extremely light, which must be related to

the developed pore structure [27]. Those differences in

the density will have an effect on the capacity when mass

units of the adsorbent are used.

The differences in the performance of our materials as

hydrogen sulfide adsorbents are presented in Fig. 1. Thecapacity is arbitrary measured only to 100 ppm. Based

on the shape of the curves and the breakthrough time,

the best materials are those with low content of polymer

and high content of sludge. They outperform the sludge-

derived sample. It is interesting that when the content of

polymer is higher than 30%, that carbonaceous phase is

no longer beneficial for the process of H2S removal.

The breakthrough capacity results calculated fromthe breakthrough curves are collected in Table 2. The

best performing adsorbent is the sample obtained from

0

20

40

60

80

100

120

0 20 40 60 80 100

Time [min]

H2S

Con

cent

ratio

n [p

pm]

CSCPS 10/90CPS 30/70CPS 50/50CPS 70/30

Fig. 1. H2S breakthrough capacity curves.

Table 2

H2S breakthrough capacity result (at 100 ppm)

Sample H2S

breakthrough

capacity

[mg/cm3]

H2S

breakthrough

capacity

[mg/g]

Water

adsorption

[mg/g]

pHin pHE

CS 10 19 93 7.38 7.15

CPS 10/90 29 45 74 8.66 8.22

CPS 30/70 31 111 134 9.00 8.98

CPS 50/50 21 160 245 8.70 6.60

CPS 70/30 16 180 173 7.21 4.93

CP 0 0 150 6.02 –

pHin—pH before H2S adsorption; pHE—pH after H2S adsorption.

1042 A. Ansari et al. / Carbon 43 (2005) 1039–1048

30% of polymer and 70% sewage sludge-polymer mix-

ture. Its capacity is three times higher than that on the

sample derived from pure sludge, CS. When the capaci-

ties per unit mass are considered, an increase in the

capacity is observed when the content of polymer in-

creases. This is due to the low density of materials. Usu-

R2 = 0.9629

0

20

40

60

80

100

120

140

160

180

200

0 20 40 60 80 100 120

content of dry sludge [%]

H2S

bre

akth

roug

h ca

paci

ty [m

g/g]

Fig. 2. Dependence of H2S breakthrough cap

ally adsorbers have a limited size so the capacity per unit

volume is important. On the other hand, the adsorbent

should have certain mechanical stability to ensure the

proper dynamic conditions of the bed.

Dependences of the capacity per unit mass and vol-

ume on the content of dry sludge (in a dry precursor)are collected in Fig. 2. When the capacity per unit mass

is considered, and the polymer derived sample is ex-

cluded from the consideration, as being an ‘‘outsider’’

and owing to the reason discussed below, the linear

dependence is observed indicating the positive effect of

addition of the carbon from polymer. On the other

hand, the dependence of the capacity expressed per unit

volume of the carbon bed on the composition of the pre-cursors shows a maximum in the performance at about

50% of the dry sludge content. This trend, besides being

linked to the trend in the bulk density of materials, is

likely caused by the synergetic effect of the final compo-

sition where both, pores from the polymer-derived car-

bon and catalytic metal content of sludge-derived

adsorbent are combined leading to the improvement in

the capacity for H2S removal. When only carbon is pres-ent, the lack of catalytic centers for oxidation can be a

limiting factor [28–30]. As shown previously, even the

carbon with the optimal pore size and volume for H2S

molecules are not able to work as efficient adsorbents

when the surface chemistry in unfavorable for dissocia-

tion of hydrogen sulfide [29,30]. Non-linear behavior

can be also linked to the differences in the yields of

adsorbents derived separately from each precursor.It is interesting that the capacities per unit volume of

the adsorbents obtained for our best performing samples

are comparable to those of coconut-shell based carbons,

which are considered as best ‘‘virgin’’ carbon for air

desulfurization [5,16]. On the other hand, the capacities

per unit mass, especially for CPS 30/70 is twice higher

than that for coconut shell based activated carbon [5].

Comparison of these values clearly indicates superiorityof CPS 30/70 where the same performance can be

0

5

10

15

20

25

30

35

0 20 4 0 6 0 8 0 100 120content of dry sludge [%]

H2S

bre

akth

roug

h ca

paci

ty [m

g/cm

3]

acity on the composition of materials.

0

50

100

150

200

250

0 0.2 0.4 0.6 0.8 1

Relative pressure, p/po

Amou

nt a

dsor

bed

[ STP

cc/

g]

CPS 10/90

CPS 10/90ECPS 30/70

CPS 30/70E

0

100

200

300

400

500

600

0 0.2 0.4 0.6 0.8 1

Relative pressure, p/po

Amou

nt a

dsor

bed

[ STP

cc/

g]

CPS 50/50

CPS 50/50ECPS 70/30

CPS 70/30E

Fig. 3. Nitrogen adsorption isotherms.

Table 3

Structural parameters of adsorbents studied

Sample SN2[m2/g] Vmic [cm

3/g] Vmes [cm3/g] Vt [cm

3/g]

CS 94 0.017 0.096 0.113

CSE 32 0.004 0.129 0.133

CPS 10/90 128 0.039 0.067 0.106

CPS 10/90E 80 0.015 0.102 0.117

CPS 30/70 316 0.111 0.118 0.229

CPS 30/70E 57 0.011 0.098 0.109

CPS 50/50 581 0.212 0.154 0.366

CPS 50/50E 166 0.053 0.111 0.164

CPS 70/30 948 0.377 0.162 0.539

CPS 70/30E 334 0.134 0.128 0.262

CP 1240 0.518 0.342 0.860

CPE 1240 0.518 0.342 0.860

A. Ansari et al. / Carbon 43 (2005) 1039–1048 1043

obtained with twice less money (if the prices of the final

products were comparable).

Differences in the performance of our materials as

H2S adsorbents have to be linked to differences in their

pore structure and surface chemistry. The nitrogen

adsorption isotherms and the pore size distributions cal-

culated from those isotherms are collected in Figs. 3 and4. As shown in Fig. 3, the samples differ significantly in

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

1 10 100 1000 10000

Pore width [ ]

Incr

emen

tal p

ore

volu

me

[ STP

cc/g

]

COS 50/50CPS 50/50ECPS 70/30CPS 70/30E

Incr

emen

tal p

ore

volu

me

[ STP

cc/g

]

Fig. 4. Pore size d

the nitrogen uptake, which is related to the content of

carbonaceous phase from polymer. Higher is content

of that phase, more nitrogen is adsorbed as a result of

the more developed pore structure [27].

Analysis of the structural parameters collected in

Table 3 clearly shows the dependence of the surface area

and porosity on the composition of the precursor. The

most microporous is CPS 70/30 and the least—CPS10/90. For all samples, the volumes of mesopores increased

with an increase in the content of polymer. It is owing to

the contribution of either sludge [16,17] or polymer [27]

to this parameter. In the case of a sludge-derived phase,

those pores are within the inorganic matter whereas in

the phase derived from polymer, which is mainly car-

bon, the mesopores are formed as a result of the release

of off-gases (SO2, H2S, CO2, hydrocarbons) and migra-tion of metals in the hot carbonaceous matrix [27].

The differences in the pore structure of our materials

before and after H2S adsorption are collected in Fig. 4.

With an increase in the content of polymer an increase

in the volume of very small pores, smaller than 10 A,

is observed indicating their origin from polymer-carbon

phase. After H2S adsorption, the volume of those pores

and pores smaller than 30 A significantly decreases as a

0

0.002

0.004

0.006

0.008

0.01

0.012

1 10 100 1000 10000

Pore width [ ]

CPS 10/90CPS 10/90ECPS 30/70CPS 30/70E

istributions.

1044 A. Ansari et al. / Carbon 43 (2005) 1039–1048

result of deposition of H2S oxidation product. That de-

crease supports our hypothesis about the importance of

micropores in the H2S immobilization process on carbo-

naceous adsorbents [28,29]. The biggest decrease in the

volume of micropores (Table 3) is noticed for CPS 50/

50E, which had high H2S removal capacities, both inmass and volume units.

To check if the apparent porosity is the result of

physical mixture of the precursors or the synergetic ef-

fect, the theoretical volumes of micropores were calcu-

lated assuming the physical mixture of materials and

taking into account the yields of the washed adsorbents.

The results are presented in Fig. 5 where solid line shows

the theoretical calculation results. The good agreementof the data obtained and theoretical calculations proves

that from the points of view of the porous structure our

materials are the physical mixtures of the adsorbent

phases coming both from the sludge and the polymer

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 4 0 6 0 8 0 100

Content of dry sludge [%]

Mic

ropo

re v

olum

e [ c

m3 /g

]

Fig. 5. Dependence of the volume of micropores on the composition

of the precursor. The solid line represents theoretical calculations of

volume assuming the physical mixture of the compounds.

0

5

10

15

20

25

30

35

0 0.2 0.4 0.6

volume of micropores [cm3/g]

H2S

bre

akth

roug

h ca

pcity

[m

g/cm

3]

Fig. 6. Dependence of H2S breakthrough c

precursors. The similar relationship exists for the spe-

cific surface area.

Since it is proposed that one of the reason of lower

H2S removal capacity of the sewage sludge derived

materials in comparison with catalytic carbons is their

low pore volume related to the low content of organicmatter [17,20], the dependence of the capacity of the

adsorbents addressed in this research on the volume of

micropores was analyzed. The results are presented in

Fig. 6. If the performance of materials were a linear

function of the pore volume the breakthrough capacity

per unit mass would show a linear trend. Analyzing

our data, an indication of the limiting capacity at Vmic

about 0.2 cm3/g is found which suggests that there is acritical amount of carbon derived from polymer, which

is beneficial for the performance of materials as H2S

adsorbents. To high amount of that carbonaceous phase

would likely decrease the capacity as suggested by a

sharp drop for the pure CP adsorbent. This implies that

the carbon derived from polymer has some surface fea-

tures, which do not favor retention of hydrogen sulfide

on the surface. When in mixture with proper amountof the sludge derived phase, the carbon phase is able

to enhance the properties of the final product, likely

due to the chemical effects originating from an inorganic

phase of the sludge. When the H2S removal capacity is

expressed per unit volume, at Vmic about 0.2 cm3/g a

well-defined maximum is found on the dependence of

the capacity on the volume of micropores. This clearly

shows that adding carbon derived from that specificpolymer increases the capacity until the suppressing ef-

fect of the carbon surface on H2S adsorption becomes

predominant.

To check the products of surface oxidation both pH

values and TA results should be taken into consider-

ation. The changes in the pH after H2S adsorption sug-

gest differences in the selectivity of oxidation (Table 2).

While for samples having 30% of polymer in the compo-sition or less the pH is only slightly affected, for CPS50/

0

20

40

60

80

100

120

140

160

180

200

0 0.2 0.4 0.6

volume of micropores [cm3/g]

Bre

akth

roug

h ca

pcity

[ m

g/g]

CP

apacity on the volume of micropores.

Table 4

Weight losses [%] in nitrogen at various temperature ranges, increase in the weight loss [%], DW, and expected sulfur content from H2S breakthrough

capacity, Sbth. [%]

Sample <200 �C 200–380 �C 380–500 �C 500–1000 �C Total weight loss DW Sbth

CS 2.60 0.22 0.27 10.25 13.34

CSE 5.02 4.47 0.91 3.75 14.15 0.81 1.78

CPS 10/90 6.78 1.32 1.00 8.74 17.84

CPS 10/90E 12.35 3.19 1.41 4.57 21.52 3.68 4.23

CPS 30/70 8.00 3.12 1.42 11.43 23.97

CPS 30/70E 10.81 7.97 4.16 17.91 40.85 16.88 10.43

CPS50/50 5.07 3.94 1.16 9.57 19.74

CPS50/50E 11.01 5.26 10.75 7.31 34.33 14.56 15.04

CPS 70/30 5.35 2.48 1.05 7.86 16.74

CPS 70/30E 11.19 10.0 6.14 8.28 35.61 18.87 16.9

CP 2.68 2.0 1.0 11.38 17.06

CPE – – – – –

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 200 400 600 800 1000

Temperature [°C]

Der

ivat

ive

of w

eigh

t los

s [%

/°C]

CPS 10/90CPS 10/90ECPS 30/70CPS 30/70 E

0

0.05

0.1

0.15

0.2

0.25

0.3

0 200 400 600 800 1000

Temperature [°C]

Der

ivat

ive

of w

eigh

t los

s [%

/°C]

CPS 50/50CPS 50/50ECPS 70/30CPS 70/30E

Fig. 7. DTG curves in nitrogen.

A. Ansari et al. / Carbon 43 (2005) 1039–1048 1045

50 and CPS70/30 a decrease of few pH units is noticed

clearly indicating the presence of sulfuric acid [28–30].

That sulfur species can be detected from TA analysis.

As seen form Table 4 a significant increase in the

amount of desorbed species occurred for the CPS50/

50E and CPS 70/30 E samples in the temperature range

200–380 �C and 380–500 �C. The former range repre-

sents the removal of strongly adsorbed SO2 (fromdecomposition of sulfuric acid) [16,20,28–31] and the

latter—the removal of elemental sulfur [32]. For the

samples with the predominant content of sludge, based

on the previous studies and on changes in pH, the lower

temperature weight loss is related to the removal of sul-

fur from the meso- and macropores of this material

[17,20].

In Table 4 we also compare the total increase in theweight loss after H2S adsorption (DW) to the expected

increase in weight assuming that all H2S adsorbed was

oxidized to sulfur (Sbth). Even though the comparison

is very rough, and, as indicated above in some cases

SO2 and sulfuric acid are also the oxidation products,

relatively good agreements between those qualities are

found suggesting that sulfur is the predominant surface

reaction product.Detailed patterns of the weight loss for our samples

before and after H2S adsorption are seen in Fig. 7. Add-

ing carbon from polymer definitely increases the weight

loss at temperatures higher than 380 �C, which is related

to deposition of sulfur in the small pores of carbon.

When the content of polymer carbon is high (70% poly-

mer in the initial ‘‘wet’’ mixture) its addition to sludge is

no longer beneficial from the point of view of the dis-posal of spent adsorbents since the high intensity of

the peak at about 300 �C is related to the removal of sul-

furic acid (SO2) [29,31]. That acid decreases significantly

the surface pH of the spent adsorbent and thus sup-

presses the dissociation of hydrogen sulfide [28,33].

The speciation of oxidation products is related to the

surface chemistry of materials, which changes after addi-

tion of polymer (Fig. 8). Adding the CP carbon causes

an increase in the acidity manifested by higher amount

of carboxylic groups present on the surface (pKa < 7).

Those groups were formed when reactive edges of

graphene layers were exposed to air. The presence of

certain amount of sludge likely inhibits this reaction

and also contributes to the catalytic effect of oxidation

to elemental sulfur. That reaction is enhanced by suchsludge components as iron oxides or calcium oxides as

demonstrated in the literature [17,34,35]. Besides acidic

0

1

2

3

4

5

6

7

8

2 3 4 5 6 7 8 9 10 11

pKa

f(pKa

) [m

mol

/g]

CSCPCPS 10/90CPS 30/70CPS 50/50CPS 70/30

Fig. 8. pKa distributions.

1046 A. Ansari et al. / Carbon 43 (2005) 1039–1048

groups having their origin in the carbonaceous phase,

the materials studied are also rich in basic components.

These components, have likely their origin in the inor-

Fig. 9. SEM images of CS (A), CP (B), CPS70/30 (C,D), and CPS50/50 (E,F).

(A), 100 nm (B), 2 lm (C), 1 lm (D), 3 lm (E), 1 lm (F).

ganic sludge-derived phase. Indeed, the pKa distribu-

tions presented in Fig. 8 clearly show an increase in

the amount and variety of species having the pKa values

greater than 8 with an increase in the content of this

phase.

The SEM micrographs of the adsorbents derivedfrom pure sludge (A) and pure polymer (B) along with

the polymer-dewatered sludge mixtures with ratios equal

to 70:30 (C and D) and 50:50 (E and F) are presented in

Fig. 9. Analyzing the texture, differences in porosity are

clearly seen. The surface of CS looks compact with a low

volume of pores. Various levels of brightness indicate

differences in the surface chemistry, a consequence of

the sludge composition [19]. In the case of CP we havea uniformly porous foam-like structure with pore of

sizes about 200 nm [27]. Moreover, the indication of

the smaller pores can be noticed in the carbonaceous

matrix. Those pores are responsible for high surface

area of the adsorbent. The mixture of sludge and poly-

mer with 10% of dry sludge (CPS 70/30) (Fig. 9C and

Size bars seen in the micrographs represent the following lengths: 2 lm

A. Ansari et al. / Carbon 43 (2005) 1039–1048 1047

D) leads to the porous material where ‘‘flaky’’ structure

can be seen along with similar elements as in Fig. 9B.

When the content of sludge increases, the texture be-

comes more compact and two phases are distinguished:

highly porous carbon and a denser sludge-derived phase

(Fig. 9F). In spite of this inhomogeneity, which was pos-tulated above based on the changes in the porosity and

surface area (physical mixture), the distribution of inor-

ganic phase looks very homogeneous. The bright dots

seen in Fig. 9E represent inorganic phase, which likely

catalyzes hydrogen sulfide oxidation [17,20,34,35]. Its

high dispersion in the pore system provides sites for

H2S oxidation to sulfur, which is stored in the pore vol-

ume enhanced by the polymer based carbonaceousphase.

4. Conclusions

Based on the results described above, addition of

polymer to the sludge improves the H2S removal capac-

ity of adsorbents compared to the materials derivedfrom the pure precursors. It is important to mention

that there is an optimal content of carbon from polymer

(about 50/50 mixture of the initial precursors), which

can lead to an improved capacity. For that enhancement

a highly dispersed inorganic phase and the expanded

pore volume coming from a polymer derived carbon

are a necessity. When the content of that carbon extends

its critical value, the capacity decreases as the surfacearea of carbon from polymer becomes predominant

and acidic groups suppress dissociation of hydrogen sul-

fide in the local pore system. Moreover, those groups

likely contribute to exhaustion of the ‘‘buffering’’ capac-

ity of the basic nature of adsorbent phase derived from

sewage sludge.

Acknowledgments

This research was supported by NYSERDA Grant #

7653. The authors are grateful to Ms. Dee Breger of La-

mont Doherty Earth Observatory of Columbia Univer-

sity, Mr. David Frey of Ziess-LEO, and Dr. Boris

Reznik of Laboratorium fur Elektronenmikroskopie,

Universitat Karlsruhe for help with SEM images. Exper-imental contribution of Ms. Anna Kleyman is appreci-

ated. TJB thanks Dr. Jacek Jagiello for providing

SAIEUS program.

References

[1] Manahan SE. Environmental chemistry. 6th ed. Boca Raton,

FL: CRC Press; 1994.

[2] <http://www.epa.gov/air/caaac/dieselreview.pdf>.

[3] Turk A, Sakalis S, Lessuck J, Karamitsos H, Rago O. Ammonia

injection enhances capacity of activated carbon for hydrogen

sulfide and methyl mercaptan. Environ Sci Technol 1989;23(10):

1242–5.

[4] Turk A, Mahmood K, Mozaffari J. Activated carbon for air

purification in New York City�s sewage treatment plants. Wat Sci

Tech 1993;27:121–6.

[5] Bandosz TJ, Bagreev A, Adib F, Turk A. Unmodified versus

caustics-impregnated carbons for control of hydrogen sulfide

emissions from sewage treatment plants. Environ Sci Technol

2000;34(6):1069–74.

[6] Bagreev A, Bandosz TJ. A role of sodium hydroxide in the process

of hydrogen sulfide adsorption/oxidation on caustic-impregnated

activated carbons. Ind Eng Chem Res 2002;41(4):672–9.

[7] Przepiorski J, Oya A. K2CO3-loaded deodorizing activated

carbon fiber against H2S gas: factors influencing the deodorizing

efficiency and the regeneration method. J Mat Sci Lett 1998;17:

679–82.

[8] Boudou JP, Chehimi M, Broniek E, Siemieniewska T, Bimer J.

Adsorption of H2S or SO2 on an activated carbon cloth modified

by ammonia treatment. Carbon 2003;41(10):1999–2007.

[9] <www.epa.gov>, Biosolids regeneration, use, and disposal in the

United States: EPA530-R-99–009. US EPA: Washington, DC;

September 1999.

[10] Piskorz J, Scott DS, Westerberg IB. Flash pyrolysis of sewage

sludge. IEC Process Des Dev 1996;25:265–70.

[11] Chiang PC, You JH. Use of sewage sludge for manufacturing

adsorbents. Can J Chem Eng 1987;65:922–8.

[12] Lewis FM. Method of pyrolyzing sewage sludge to produce

activated carbon US Patent 4,122,036; 1977.

[13] Sutherland J. Preparation of activated carbonaceous mate-

rial from sewage sludge and sulfuric acid US patent 3,998,757;

1976.

[14] Lu GQ, Low JCF, Liu CY, Lau AC. Surface area development of

sewage sludge during pyrolysis. Fuel 1995;74:3444–51.

[15] Lu GQ, Lau DD. Characterization of sewage sludge-derived

adsorbents for H2S removal. Part 2: Surface and pore struc-

tural evolution in chemical activation. Gas Sep Purif 1996;10:

103–9.

[16] Bagreev A, Bandosz TJ, Locke DC. Pore structure and surface

chemistry of adsorbents obtained by pyrolysis of sewage sludge-

derived fertilizer. Carbon 2001;39:1971–8.

[17] Bagreev A, Bashkova S, Locke DC, Bandosz TJ. Sewage sludge-

derived materials as efficient adsorbents for removal of hydrogen

sulfide. Environ Sci Technol 2001;35(7):1537–43.

[18] Bashkova S, Bagreev A, Locke DC, Bandosz TJ. Sewage sludge

derived materials as adsorbents for H2S and SO2. In: Kaneko K,

Kanoh H, Hanzawa Y, editors. Fundamentals of adsorption, vol.

7. Chiba Japan: IK International; 2002. p. 239–46.

[19] Bagreev A, Bashkova S, Reznik B, Zibat V, Bandosz TJ.

Heterogeneity of sewage sludge derived materials as a factor

governing their performance as adsorbents of acidic gases. In:

Rodriguez-Reinoso F, editor. Proceeding of conference on char-

acterization of porous solids, vol. 6. Amsterdam: Elsevier; 2003.

p. 217–24.

[20] Bagreev A, Bandosz TJ. Efficient desulfurization adsorbents

obtained by pyrolysis of sewage sludfge derived fertilizer modified

with spent mineral oil. Environ Sci Technol 2004;38:345–51.

[21] Olivier JP. Modeling physical adsorption on porous and nonpo-

rous solids using density functional theory. J Porous Mater

1995;2:9–17.

[22] Lastoskie CM, Gubbins KE, Quirke N. Pore size distribution

analysis of microporous carbons: a density functional theory

approach. J Phys Chem 1993;97:4786–96.

[23] Bandosz TJ, Jagiello J, Contescu C, Schwarz JA. Characterization

of surfaces of activated carbons in terms of the acidity constant

distribution. Carbon 1993;31:1193–200.

1048 A. Ansari et al. / Carbon 43 (2005) 1039–1048

[24] Jagiello J, Bandosz TJ, Schwarz JA. Carbon surface character-

ization in terms of its acidity constant distribution. Carbon

1994;32:1026–8.

[25] Jagiello J, Bandosz TJ, Putyera K, Schwarz JA. Determination of

proton affinity distributions for chemical systems in aqueous

environments using stable numerical solution of the adsorption

integral. J Colloid Interface Sci 1995;172:341–6.

[26] Jagiello J. Stable numerical solution of the adsorption integral

equation using splines. Langmuir 1994;10:2778–85.

[27] Hines D, Bagreev A, Bandosz TJ. Surface properties of porous

carbon obtained from polystyrene sulfonic acid-based organic

salts. Langmuir 2004;20:3388–97.

[28] Adib F, Bagreev A, Bandosz TJ. Analysis of the relationship

between H2S removal capacity and surface properties of unmod-

ified activated carbons. Environ Sci Technol 2000;34(4):686–92.

[29] Adib F, Bagreev A, Bandosz TJ. Effect of surface characteristics

of wood based activated carbons on removal of hydrogen sulfide.

J Colloid Interface Sci 1999;214(2):407–15.

[30] Adib F, Bagreev A, Bandosz TJ. Effect of pH and surface

chemistry on the mechanism of H2S removal by activated carbons.

J Colloid Interface Sci 1999;216(2):360–9.

[31] Rodrigues-Mirasol J, Cordero T, Rodriguez JJ. Effect of oxygen

on the adsorption of SO2 on activated carbon. In: Extended

abstracts of 23rd Biennal conference on carbon, College Park,

July 1997. p. 376–7.

[32] Chang CH. Preparation and characterization of carbon–sulfur

surface compounds. Carbon 1981;19:175–86.

[33] Bandosz TJ. On the adsorption/oxidation of hydrogen sulfide on

unmodified activated carbon at temperatures near ambient. J

Colloid Interface Sci 2002;246(1):1–20.

[34] Bagreev A, Bandosz TJ. H2S adsorption/oxidation on adsorbents

obtained from pyrolysis of sewage sludge-derived fertilizer using

zinc chloride activation. Ind Eng Chem Res 2001;40:3502–10.

[35] Bagreev A, Bandosz TJ. H2S adsorption/oxidation on materials

obtained using sulfuric acid activation of sewage sludge derived

fertilizer. J Colloid Interface Sci 2002;252:188–94.