Desalination 279 (2011) 225234

Contents lists available at ScienceDirect

Desalination
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d e s a l

Inuence of impregnation ratio on coffee ground activated carbon as landll leachate

adsorbent for removal of total iron and orthophosphate
Sim Lan Ching a, Mohd Sufan Yusoff a,, Hamidi Abdul Aziz a, Muhammad Umar a, b
a
School of Civil Engineering, Engineering Campus, Universiti Sains Malaysia, 14300, Nibong Tebal, Penang, Malaysia
b
School of Civil, Environmental and Chemical Engineering, RMIT University, Melbourne 3001, Victoria, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Coffee grounds waste was used to prepare activated carbon with particle size between 1.18 and 2 mm by
Received 18 January 2011 chemical activation with sulphuric acid. The inuence of impregnation ratio on physical properties of
Received in revised form 19 May 2011 activated carbon and removal efciency of total iron and orthophosphate (PO4P) was studied. Impregnation
Accepted 5 June 2011
ratio of 0.5 was found optimum for preparing activated carbon with high pore surface area (224.7 m 2/g) and
Available online 24 June 2011
micropore volume (0.07986 cm 3/g). Batch experiments were conducted to study the effects of contact time,
Keywords:
pH, particle size and adsorbent dosage. Impregnation ratios of 2.5 and 0.5 were found to be optimum for the
Coffee ground adsorption of total iron and PO4P, respectively at doses of 10 g and pH 8.1. pH 13 was optimum for iron
Activated carbon removal while pH b 5 and N 11 was optimum for PO4P removal. The adsorption capacity increased with
Landll leachate increasing adsorbent dosage for total iron. However, this trend was exceptional for PO4P with uctuating
Semi-aerobic adsorption capacity. The smaller particle size yielded larger adsorption capacity.
Adsorption 2011 Elsevier B.V. All rights reserved.

1. Introduction coffee grounds as renewable fuel which reduces the needs for landll
dumping.
Coffee is produced all over the world. In 2008, according to the The adequacy and suitability of natural, renewable and low cost
International Coffee Organization the production of coffee materials (palm shell, pall ber, palm stone, bamboo dust, peat,
amounted to about 680,000,000 tons [1]. In Malaysia, 45% of all the chitosan, lignite, fungi, moss, bark husk, chitin, coir pith, maize cob,
waste dumped to landlls is composed of kitchen waste including pinewood sawdust, rice husk, sugar cane bagasse, tea leaves, and sago
coffee grounds [2]. The coffee grounds, which are the residues of the waste) as alternative precursors have currently been executed [6,7].
soluble coffee industry, are highly pollutant due to the presence of Various non-conventional low cost adsorbents have shown good
organic material that demands a great quantity of oxygen in order to adsorption capacities in removing COD from landll leachate such as
degrade [3]. The amount of carbon dioxide (CO2) produced by burning rice husk with 70% of COD removal [8] oil palm shell with 50% of COD
1000 g of coffee grounds was 538 g [4]. removal [9] and peats with 69% of COD removal [10].
In Malaysia, coffee production for 2009/10 was forecasted at Landll leachate is a complex wastewater produced as a result of
450,000 bags (60 kg) for both green coffee beans and Robusta coffee interaction of waste with water percolating through the body of a landll
beans. The import of green coffee beans for 2009/10 was forecasted at [1113]. To date, previous studies on non-conventional low cost
650,000 bags (60 kg) [5]. Usually, the locally produced coffee product adsorbents were limited to the application for basic dye or COD removal
is in soluble powder form. Initially, 1.86 kg of spent coffee grounds from landll leachate. Not much attempt has so far been made to obtain a
was produced for each kilogram of soluble coffee. With improvements comprehensive overview of leachate treatment in terms of the adsorption
to the industry, this gure dropped to 1.27 kg in 1967, and it was capacity of low cost adsorbents for other parameters such as ion
0.91 kg in 1997 [3]. phosphate and heavy metals. Thus, there is a need to explore the use of
A new Nestl plant in Shah Alam will feature Nestl standard of coffee grounds as low cost adsorbents for other parameters' removal and
environment-friendly manufacturing practices such as incineration of to optimize their removal conditions. Furthermore, limited investigations
spent coffee grounds and will eventually join the list of more than 20 about the effect of impregnation ratio on the adsorption capacities of
Nescaf factories around the world that are currently using spent activated carbons prepared from coffee grounds by sulphuric acid
activation have been reported.
The aims of this study were to characterize coffee grounds
activated carbon (CGACs) based on difference H2SO4 impregnated
ratios and to study the effects of shaking time, pH, adsorbent dosage
Corresponding author. Tel.: + 60 45996223; fax: + 60 45941009. and particle size on the adsorption of total iron and PO4 from leachate
E-mail address: [email protected] (M.S. Yusoff). produced at semi-aerobic Pulau Burung Landll Site (PBSL).

0011-9164/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.desal.2011.06.011
226 S.L. Ching et al. / Desalination 279 (2011) 225234

2. Materials and methods after activation, carbonization and washing (w1) to that of the dry
coffee grounds (w0): yield% = (w1/w0) 100. The burn off is the
2.1. Analytical techniques weight loss percentage due to the activation and carbonization step.

Landll leachate samples were collected from PBLS using plastic 2.3. Characterization of adsorbents
containers and stored at 4 C until analysis was carried out. pH was
measured onsite using a portable pH meter (Hach, sens ion 1, USA). The N2 adsorptiondesorption isotherms of the activated carbons
COD was determined using Colorimetric Method (5220-D). NH4N were measured using an automatic adsorption instrument (AutoSorb
was determined using Nesslerization method (4500-NH3). The I, Quantachrome) at liquid nitrogen temperature (77 K). The specic
removal of total iron and PO4 P was determined using UV surface areas (SBET) of the activated carbons were calculated using the
spectrophotometer (HACH DR2500). PO4P and total iron concentra- BrunauerEmmettTeller (BET) equation by assuming the area of the
tion was determined using Molybdovanate and FerroZine Method at nitrogen molecule to be 0.162 nm 2. The micropore volume and
wavelength of 255 nm and 480 nm, respectively. The readings were microporous surface area were determined by the t-plot method [14].
taken after ltering the sorbents from the leachate samples. The The capacity of adsorbents to remove color can be evaluated
amount adsorbed was determined by taking the difference between through iodine adsorption from aqueous solutions using test
the initial concentration and the concentration of the solution at the conditions referred to as iodine number determination which
time of sampling. pH was adjusted by using 0.1 M H2SO4 or NaOH. indicates their relative activation level and the surface area available
for micropores [15]. Usually adsorbents with a high iodine number
2.2. Preparation of adsorbent have a high surface area and are suitable for adsorbing small
compounds [15]. The iodine number corresponding to the produced
Coffee grounds were collected from a local coffee making shop adsorbent was determined according to the procedure established by
located in Kampung Baru, Bukit Mertajam, Penang. The coffee grounds the Standard Test Method (ASTM D4607-86) [16]. According to this
were washed with boiled water to eliminate the impurities, dust and test, 0.1 g of activated carbon was placed in a 250 mL dry Erlenmeyer
water soluble substances followed by washing with distilled water. ask and fully wetted with 10 mL of diluted HCl (5% by weight).
The washed coffee grounds were oven dried at 105 C for 24 h prior to 100 mL of iodine solution (0.1 mol/L) was poured into the ask and
their activation. The dried coffee grounds were impregnated with a the mixture was vigorously shaken for 30 s. After ltration, 50 mL
weighted amount of H2SO4 aqueous solution. The impregnation ratio solution was titrated with 0.1 mol/L sodium thiosulfate. The iodine
is dened as the ratio of the weight (g) of H2SO4 to the weight of number expressed in mg/g was reported as the amount of iodine
precursor (g of dried coffee grounds). Coffee grounds activated adsorbed per gram of adsorbent at a residual iodine concentration of
carbons (CGAC) were prepared with different impregnation ratios 0.02 N.
namely IR 0.5, IR 1.5 and IR 2.5. The impregnation ration was Fourier transform infrared (FTIR) transmission spectra of carbon
determined according to the following formula. samples were obtained through potassium bromide technique [17] on
FTIR spectrometer (Model PerkinElmer Spectrum One) in 400
Impregnation ratio =
Weight of the H2 SO4 4000 cm 1 wave number range, with a 2 cm 1 resolution. FTIR was
Weight of the coffee ground activated carbon CGAC used to provide information on the chemical structure of the activated
carbon before and after adsorption. Pellets made of a mixture of
Dried coffee grounds were impregnated by putting 200 g of coffee 0.6 mg of activated carbon and 200 mg of KBr were pressed at
grounds each in 3 conical asks in contact with 600 mL of H2SO4 350 MPa, and oven-dried for 48 h at 110 C before analysis.
solutions for 24 h. The concentration of acid was 96% by weight. The
activated samples were subsequently oven dried at 105 C for 24 h. 2.4. Adsorbate
The resulting samples were then carbonized for 60 min at 600 C in a
vertical tube furnace (Carbolite Type MTF 12/38/250) under nitrogen The leachate samples used in this study were obtained from semi-
atmosphere. After cooling, the char was washed with boiled and aerobic PBLS, which is situated within Byram Forest Reserve in
distilled water in order to eliminate the residual organic and mineral Penang, Malaysia. The site has a natural marine clay liner with the
matter. Washing with water easily removed most of the residues from total area of 33 ha and it is equipped with a leachate collection pond.
activating agents. The completion of washing was carried out until the The PBLS was developed into a semi-aerobic system employing a
nal pH of washing water was neutral. It was then dried at 105 C for controlled tipping technique and leachate recirculation system in
24 h. The carbon product was sieved to a uniform size (1.182 mm) March 2001 and is one of the only three sites of its kind in Malaysia
and stored in closed bottles. The yield of the activated carbon [18]. The site receives approximately 1800 tons of solid waste daily
synthesis is dened as the ratio of the weight of the activated carbon from Penang Island [19]. No pretreatment was done for the collected

Table 1
Characteristics of leachate from Pulau Burung landll site.

Raw 1 2 3 4 5 6 Average

pH 7.89 7.71 7.12 7.67 7.43 7.57 7.5

COD (mg/L) 1200 1010 1450 1620 1773 1815 1478
BOD5 (mg/L) 271 256 184 215 338 218 247
BOD5/COD 0.23 0.25 0.13 0.13 0.19 0.12 0.17
TKN (mg/L) 1771.84 1699.4 1711.36 1644.16 1563.52 1520.5 1651.8
NH4N (mg/L) 2609 2670 2208 2388 2405 2780 2510
Nitrite (mg/L) 43.3 36.7 42.5 46.7 36.7 43.3 41.5
Nitrate (mg/L) 1.33 1.67 3.2 1.33 2.67 2.33 2.1
PO4P (mg/L) 284 276 290 220 250 240 260
Color point (Pt.Co) 4153 4030 3527.5 3670 3679.5 3720.5 3796.75
SS (mg/L) 434 366 450 490 458 463 443.5
Total iron (mg/L) 4.25 4.61 4.42 4.82 4.73 4.59 4.57
 S.L. Ching et al. / Desalination 279 (2011) 225234 227

leachate samples. The collected samples were stored in darkness at containing leachate (100 mL in each case), at different impregnation
4 C with minimum exposure to surrounding air to keep the leachate ratios of coffee ground activated carbons at pH 8.1 and temperature
characteristics unchanged. The characteristics of leachate samples 30 C. The asks were placed in an orbital shaker (PROTECH Model
were based on 6 samplings collected for 3 months from January to 720) and agitation was provided at 300 rpm for 150 min. The particle
March, 2010. All experiments were undertaken in triplicates. The sizes were varied at 1.182, 0.61.18, 0.4250.6, 0.30.425 and 0.15
detailed composition is shown in Table 1. 0.3 mm. The concentrations of total iron and PO4P were similarly
The COD strength of the leachate varied between 1010 and measured.
1200 mg/L, while the BOD5 concentration was less than 500 mg/L.
This gives a BOD5/COD ratio of about 0.17, which indicates that most 2.6. Calculation analysis
of the organic compounds in the landll leachate were poorly
biodegradable. Ammonium strength is also a major pollutant in the In this work, calculation analysis was carried out to determine the
landll leachate and its strength was around 2510 mg/L. In each value of yield, weight loss, and percentage of removal.
experiment, the same sample was used and the initial characteristics
were determined in order to take consideration of the changes in 2.6.1. Yield and weight loss calculation
composition and to reduce the background interferences. The yield of activated carbon is dened as the ratio of the weight of
the resultant activated carbon to that of the original coffee grounds
2.5. Batch adsorption experiments (dry basis). The percentage yield of activated carbon was calculated as
follows:
This study was divided into four parameters with four different
factors. Different impregnation ratios of 0.5, 1.5 and 2.5 were studied for Yield % = w1 = w0 100 1
each of the parameters with four different factors and the treatment
performance was evaluated and reported as percent removal. Moreover, where w0 is the weight of original coffee grounds (dry basis) and w1 is
the study was concentrated on the identication of the condition that the weight of the resulted activated carbon.
provides the best treatment performance. Four different aspects of the The weight loss or the weight ratio of the nal product to that of
experimental effect factors such as contact time, pH, particle size and raw materials was calculated from:
adsorbent dosage were evaluated based on the range as follows:
Weightloss % = w0 w1 = w0 100 2

2.5.1. Contact time where w0 is the weight of original coffee grounds (dry basis) and w1 is
Adsorption experiments were carried out by adding a xed the weight of the resulted activated carbon.
amount of sorbent into a number of 250 mL conical asks containing
leachate, at different impregnation ratios of coffee ground activated 2.6.2. Percentage of removal calculation
carbons at pH 8.1 and temperature 30 C. The volume of leachate The removal percentage of PO4 and Fe can be calculated as follows:
samples was 100 mL in each case while a xed amount of 10 g CGAC
was used in each experiment. Already impregnated CGAC was added Removal % = C0 Ce = C0 100 3
to the conical asks.
The asks were placed in an orbital shaker (PROTECH Model 720) where C0 and Ce (mg/L) are the liquid-phase concentrations of total
and agitation was provided at 300 rpm for 150 min. The leachate iron and PO4P at initial and equilibrium, respectively.
samples were taken at preset time intervals after 30, 60, 90, 120, 150,
180, 210, 240 and 270 min. The concentrations of total iron and PO4P 3. Results and discussion
were similarly measured.
3.1. Characterization of activated carbon
2.5.2. pH
Adsorption experiments were carried out by adding a xed Characterization of activated carbon included weight loss and
amount of sorbents (10 g) into a number of 250 mL conical asks yield, iodine number, specic surface area, micropore volume and
containing leachate (100 mL in each case), at different impregnation FTIR.
ratios of coffee ground activated carbons at 30 C. The asks were
placed in an orbital shaker (PROTECH Model 720) and agitation was 3.1.1. Weight loss and yield
provided at 300 rpm for 150 min. The pH was varied at 2, 5, 7, 11 and Table 2 illustrates that an increase in sulphuric acid resulted in an
13 by using either 0.1 M H2SO4 or NaOH. The concentrations of total increase in the weight loss and hence a reduced yield of the carbon
iron and PO4P were similarly measured. product. The weight loss was 49, 48.15 and 57.23% for IR 0.5, IR 1.5 and
IR 2.5, respectively. The yield was 51, 51.85 and 42.77% for IR 0.5, IR
2.5.3. Adsorbent dosage 1.5 and IR 2.5, respectively. H2SO4 reacted with char and volatile
Adsorption experiments were carried out by adding various amount matter and diffused quickly out of the surface of particles during the
of sorbents (5, 10, 15, 20 and 25 g) into a number of 250 mL conical asks activation process. Therefore, with high H2SO4 content, the gasica-
containing leachate (100 mL in each case), at different impregnation tion of surface carbon atoms became predominant leading to an
ratios of coffee ground activated carbons at pH 8.1 and temperature increase in the weight loss and a low carbon yield.
30 C. The asks were placed in an orbital shaker (PROTECH Model 720)
and agitation was provided at 300 rpm for 150 min. The concentrations 3.1.2. Iodine number
of total iron and PO4P were similarly measured. The equilibrium time Iodine numbers are indications of the adsorption capacity in
was taken as 150 min for each IR by xing the CGAC dosage at 10 g for micropores, and therefore are often employed to examine the
each case of different impregnation ratios and kinetics of adsorption adsorption capacity of the activated carbon. Table 2 demonstrates
were measured for both total iron and PO4P. that the iodine number increased with an increase in the impregna-
tion ratio. The iodine number was 71.33, 93.67 and 96.5 for IR 0.5, IR
2.5.4. Particle size 1.5 and IR 2.5, respectively. As acid content increased, more reaction
Adsorption experiments were carried out by adding a xed took place and consequently, more surface area was developed. A
amount of sorbents (10 g) into a number of 250 mL conical asks higher acid content generally lead to formation of the product with a
228 S.L. Ching et al. / Desalination 279 (2011) 225234

higher level of volume and surface area which could then provide a 3.2. Effect of adsorbent dosage
higher adsorption capacity; however, it was difcult to establish this
concept with only three impregnation ratios studied in this work. The removals of total iron and PO4P as a function of adsorbent
dosage at different IR of 0.5, 1.5 and 2.5 were calculated. Initial
concentration of total iron was 4.57 mg/L while 260 mg/L for PO4P.
3.1.3. Specic surface area and pore structure
Fig. 2a shows the variations in total iron adsorption at different
An important property of the activated carbon is its adsorptive
dosages of CGAC while Fig. 2b shows the variations in PO4P
capacity, which is related to the specic surface area. Table 2 shows
adsorption at different CGAC dosages. These results indicate that the
the BET surface area obtained for each impregnation ratio. The specic
removal percentage of total iron increased with an increase in the
surface areas for IR 0.5, IR 1.5 and IR 2.5 were 146.1, 55.43 and
adsorbent dosage from 5 to 25 g (Fig. 2a). A similar trend was
121.5 m 2 g 1, respectively. Using H2SO4 as the activating reagent, the
observed for PO4P removal (Fig. 2b). These results were expected as
maximum surface area was obtained for IR 0.5. Interestingly, with
with increased adsorbent dosage, the available sorption surface and
increasing impregnation ratio from 0.5 to 2.5, surface area decreased
availability of more adsorption sites increase which results in more
from 146.1 to 121.5 m 2 g 1. There were also signicant reductions in
adsorbate attached to the surfaces [21,22]. Although an increase in
micropore and total pore volumes with increasing impregnation ratio.
adsorption occurred with respect to increase in dosage, the trend of
In a study of the preparation of the activated carbon from cattle-
amount adsorbed per unit mass decreased by increasing dosage as
manure by ZnCl2 activation [20], it was found that when the
shown in Fig. 2 (a, b).
impregnation ratio was higher than 1.5 in the activation process, a
The removal percentage of total iron increased in the order of IR of
decrease in both BET surface area and micropore volume of the
0.5, 1.5 and 2.5 (Fig. 3a). The removal percentage of total iron
activated carbon was observed. Experimental results suggested that
improved by 19.5, 18.6 and 12.5% for IR 0.5, IR 1.5 and IR 2.5,
the impregnation ratio strongly affects characteristics of the activated
respectively when the adsorbent dosages were increased from 5 to
carbon; hence an appropriate impregnation ratio should be selected
25 g. A higher increment in removal percentage was seen for IR 0.5
for the production of the activated carbon with the highest surface
compared to other impregnation ratios. The best removal of 72.6% for
area and pore volume. As a comparison, micropore volume obtained
total iron was achieved for IR 2.5 by using 25 g of adsorbent. The PO4
in this study for IR 0.5 was smaller in comparison to that obtained by
P removal percentage for IR 1.5 decreased by 27% and 3.8% as the
using H3PO4 activated CGAC in a recent study conducted by Reffas et
adsorbent dosage increased from 10 to 15 g and 15 to 20 g,
al. [1]. It was not possible to draw a comparison for IR 1.5 and IR 2.5
respectively. Meanwhile, the removal percentage for IR 1.5 increased
used in this study because they [1] used different impregnation ratios.
from 50% to 57.7% when the dosage increased from 20 to 25 g. For IR
2.5, the percentage removal decreased by 11.54 and 19.23% when the
3.1.4. FTIR of CGACs adsorbent dosage increased from 5 to 10 g and 10 to 15 g,
The infrared spectra of CGACs before and after adsorption are respectively. As mentioned earlier, it was rather difcult to establish
illustrated in Fig. 1. The more intense bands are reported in Table 3. It an optimum impregnation ratio in this study; however, a higher
is clear that CGACs has considerable numbers of heterogeneous layers impregnation ration generally leads to higher ash content due to
of pores where there was a good possibility for total iron and PO4P to consumption of more carbon atoms in activation reaction [23]. This
be adsorbed. The FTIR spectrum of CGACs (Fig. 1 and Table 3) shows could be veried from more uctuations that occurred in the PO4
that some peaks were shifted or disappeared but no new peak was removal for IR 1.5 and 2.5. The unstable trend that occurred for IR 1.5
detected. The changes observed in the spectrum indicate the possible and 2.5 could have been due to the reactions of various anions present
involvement of those functional groups on the surface of the CGACs in in leachate consequently leading to reduced removal performance.
the adsorption process. However, no uctuation was observed for IR 0.5 curves which refer to
Fig. 1 indicates that functional groups OH, NH and CH stretching increased conductivity of activated carbon.
in alkyl or aryl existed in the CGACs at wavenumber 32323839 cm 1 Although percent adsorption increased with increase in adsorbent
(before adsorption) and 32463839 cm 1 (after adsorption). Amino dosage, the amount adsorbed per unit mass decreased. This statement
groups were also detected in CGACs at wavenumber 2299 can be supported from the results plotted in Fig. 3 (a, b). The decrease
2395 cm 1 (before adsorption) and 2347 cm 1 (after adsorption), in adsorption density with increase in adsorbent dose is mainly owing
while C_O groups were found at wavenumber 1521 cm 1 (before to unsaturation sites through the adsorption process [24,25]. Another
adsorption) and 1543 cm 1 (after adsorption). The OH and CO reason might be the particle interactions such as aggregation,
bending (in-plane) was detected at 1433 cm 1 (before adsorption) resulting from high sorbent concentration. Such aggregation would
and 1415 cm 1 (after adsorption). POC strongest band highest lead to a decrease in the total surface area of the adsorbent [26].
frequencies for aliphatic amines (1053 cm 1) and CH out-of-plane
deformation (825 cm 1) were also detected before and after 3.3. Effect of contact time
adsorption.
The removal of total iron and PO4P as a function of contact time at
different impregnation ratios of 0.5, 1.5 and 2.5 was observed. The
Table 2
results are presented in Fig. 4 (a, b). As expected, the removal
Physical properties of activated carbon prepared at various impregnation ratios. percentage increased with time until the equilibrium was attained
between the amounts of adsorbate adsorbed on the adsorbent and
Impregnation ratio H2SO4/coffee
that remained in solution.
Parameter 0.5 1.5 2.5 In the total iron adsorption curve shown in Fig. 4 (a), the uptake of
Particle size (mm) 1.182 1.182 1.182 total iron for IR 0.5 was slow in the rst 90 min and the removal
Weight loss (%) 49 48.15 57.23 increased rapidly by 18% as the contact time increased from 180 to
Yield (%) 51 51.85 42.77
270 min. Among the three different impregnation ratio of adsorbents,
Iodine number (mg/g) 71.33 93.67 96.5
BET surface area (m2/g) 146.1 55.43 121.5 IR 2.5 performed a constant removal percentage of around 73% as
Pore specic volume Vtot (cm3/g) 0.0705 0.01938 0.05925 different stage of increment was not seen in the adsorption curve. In
Micropore area (m2/g) 224.7 86.64 187.2 contrast, adsorption curve for IR 0.5 increased rapidly after contact
Micropore volume (cm3/g) 0.07986 0.03079 0.06653 time 180 min and the removal percentage kept increasing instead of
Average pore diameter (nm) 1.6 1.398 1.951
slowing down. Such trend indicates that the adsorption process took
 S.L. Ching et al. / Desalination 279 (2011) 225234 229

31 3232
3839
30 (a) Before 441
29 1433
28
568
27 3504 2395 2299 825
26
3616 1521 1244
25 1053
24 3736
%T
23 (b) After
22
3650
21 442
20
19 569
3246 2347 825
18 3839 3736 1543 1415 1241 1053
17
16.
0 4000. 360 320 280 240 200 180 160 140 120 100 80 60 400.
0 0 0 0 0 0 0 0 0 0 0 0 0 0
cm-1

Fig. 1. FTIR of CGAC: (a) before adsorption and (b) after adsorption.

more than 270 min to reach equilibrium. In the case of IR 1.5, the IR 1.5, the uptake of PO4P increased gradually with increasing
removal of total iron increased gradually with increasing contact time contact time, then slightly dropped at contact time 150 min, and
at the early stage of adsorption until the contact time reached thereafter reached equilibrium with 69% removal at 270 min.
210 min. The increase in removal rate from 76 to 84% towards the end Optimum removal efciency for both IR 0.5 and IR 2.5 was 84.6 and
of the experiments indicates that IR 1.5 took more than 270 min to 69.2%, respectively at contact time of 150 min. Hence, the contact time
reach the equilibrium. of 150 min was considered as the optimum.
The adsorption curve for PO4P as shown in Fig. 4 (b) uctuated
more in comparison to total iron. The tremendous uctuation trend 3.4. Effect of pH
was seen both in the studies of effect of contact time and adsorbent
dosage for IR 2.5 on PO4 removal. These results proved that IR 2.5 was The removal of total iron and PO4P as a function of pH at different
not suitable to be used for PO4P adsorption. The efciency of PO4P impregnation ratios of 0.5, 1.5 and 2.5 was studied. The results are
reduction for the three different impregnation ratio of adsorbents was presented in Fig. 5 (a, b). Adsorption capacity for total iron at various
of the order IR 0.5 N IR 1.5 N IR 2.5. In order to explain this behavior, the
dependency of the removal capacity on the physical structure of the
sorbent was considered. The removal capacity was clearly dependent
on the total pore volume and the mean pore diameter as shown in
Table 2. As IR 0.5 had higher pore volume and the mean pore
diameter, phosphate uptake by the sorbent was higher. The
enhancement of the removal capacity seemed to be due to the
contribution of the wide pores created, which maintained the
diffusional transport of PO4P from the liquid phase into the internal
surface of the sorbent. A slight uctuation was observed in the
adsorption curves of IR 0.5 and IR 1.5. This phenomenon might have
been due to the effect of foreign anions such as SO42, Cl and NO3
which reduced the percent removal of PO4P [27]. In case of IR 0.5 and

Table 3
FTIR of CGACs.

Main observed bands

(cm1)

Before After Differences Assignment

adsorption adsorption

3839 3839 0 OH, NH, CH stretching in alkyl or aryl

3736 3736 0
3616 3650 34
3232 3246 14
2395 2347 48 Amino groups
2299
1521 1543 22 C_O stretching
1433 1415 18 in-plane OH bending and CO stretch of
dimmers
1244 1241 3 CO stretching
1053 1053 0 POC strongest band highest frequencies
for aliphatic amines
Fig. 2. Effect of different adsorbent dosage on the amount adsorbed of (a) total iron, and
825 825 0 CH out-of-plane deformation
(b) PO4P using CGAC.
230 S.L. Ching et al. / Desalination 279 (2011) 225234

degree of ionization of the adsorbate during reaction [28]. Thus, the

role of hydrogen ion concentration in the removal of total iron was
examined. The maximum removal efciency achieved by IR 0.5, IR 1.5
and IR 2.5 was 96, 86 and 100%, respectively at alkaline pH range 11
13. The removal efciency increased signicantly with increasing pH
for all adsorbents. It could be observed that increasing pH of the
solution from 2 to 11 caused linear increase in total iron removal
efciency for IR 0.5 and IR 1.5. The removal efciency increased
rapidly above pH 11. However, different trend was observed for IR 2.5.
The adsorption increased rapidly from pH 5 to 11, and thereafter
maintained at 100% at pH 11 and 13. Better adsorption performed by
IR 2.5 could be due to increasing number of micropores which are
proved by the highest iodine number for IR 2.5 compared to IR 0.5 and
IR 1.5, making total iron easier to penetrate into the micropores [29]. It
could be concluded that the adsorbent achieved a good total iron
removal at pH 13 with a minimum value of 80% (Fig. 5a). These results
could be attributed to the positive charge of total iron which would
react and attach onto adsorbent surface's functional groups. There are
four major functional groups on the surface of activated carbon which
are carboxyl, carbonyl, hydroxyl, and lactonized carboxyl [29].
However, the actual chemical reaction between the metal ion and
functional groups on the activated carbon surface was complex and
difcult to understand. The removal efciency of metal ions generally
increased when the initial pH of solution was increased. This was
because the adsorbents were highly selective for H3O + ions when the
concentration of H3O + ions was high at low pH value. Thus, at lower
pH values, H3O + ions competed with metal ions for the exchange sites
in adsorbents [30,31]. A higher pH value also leads to the formation of
Fig. 3. Percent removal of (a) total iron, and (b) PO4P using CGAC.
iron precipitates resulting in enhanced total iron removal.
Maximum removal of PO4P occurred at pH 5 for IR 1.5 with
pH values is presented in Fig. 5 (a). The pH of the aqueous solution is removal efciency of 58% as shown in Fig. 5 (b). In contrast, IR 0.5 and
an important operational parameter in the adsorption process IR 2.5 reached the maximum removal efciency at pH 13 with the
because it affects the solubility of the metal ions, concentration of removal of 73 and 77%, respectively. At pH 2, IR 0.5 had the lowest
the counter ions on the functional groups of the adsorbent and the removal efciency of 34.6%. IR 1.5 and IR 2.5 had the lowest removal
efciency in pH range 713 and 711, respectively. Phosphate exists
as H3PO4, H2PO4, HPO42, and PO43, depending on the pH of the

Fig. 4. Effect of different contact time on the removal performance of (a) total iron, and
(b) PO4P using CGAC. Fig. 5. Effect of pH on removal performance of (a) total iron, and (b) PO4P using CGAC.
 S.L. Ching et al. / Desalination 279 (2011) 225234 231

system. The predominant species of phosphate is H3PO4 at lower pH.

These protonated species are weakly attached to the sites of the
adsorbent. With increased pH, other species of phosphate such as
H2PO4 and HPO42 were prevalent; hence their adsorption occurred
at high pH. At pH value more than 8, the concentration of OH in
solution, which competes with the PO43 species, became higher. The
highly negatively charged adsorbent sites did not favor the adsorption
of oxyanions due to the electrostatic repulsion [32,33]. Thus, it could
be concluded that optimum pH suitable for PO4P removal was in pH
range 57 and pH more than 13.

3.5. Effect of particle size

The removal of total iron and PO4P as a function of particle size at

different impregnation ratios of 0.5, 1.5 and 2.5 was determined. The
results are presented in Fig. 6 (a, b).
The adsorption of total iron as shown in Fig. 6 (a) gradually
decreased with a decrease in particle size. The percentage removal of
total iron at particle size of 0.153 mm was the minimum which was
around 40, 54 and 53% for IR 0.5, IR 1.5 and IR 2.5, respectively. The
optimum size range for total iron adsorption was 0.4250.6 mm, 0.6
1.18 mm, and 1.182 mm for IR 0.5, IR 1.5 and IR 2.5, respectively. The
efciency of Fe reduction for the three different impregnation ratios of
adsorbents was of the order IR 2.5 N IR 1.5 N IR 0.5. The maximum
removal percentages of 47, 61 and 72% were achieved for IR 2.5, IR 1.5
and IR 0.5, respectively. There was a linear decrease in removal
efciency for IR 2.5 from size range 1.182 to 0.4250.6 mm but it
remained constant for size range 0.30.425 and nally dropped by
8.5% at 0.150.3 mm. For IR 1.5, the adsorption curve decreased by
6.3% in the beginning and thereafter maintained at 54% from size Fig. 7. Langmuir (a) and Freundlich isotherm (b) for total iron.
range 0.4250.6 mm to 0.150.3 mm. While for IR 0.5, the uptake only
dropped by 11.6% from size range 0.4256 mm to 0.30.425 mm. The
results indicated that the total iron adsorption was more inuenced
by other elements than its particle size. It possibly shows that the total iron adsorption occurred by chemical adsorption rather than physical
adsorption (outer sphere complex) [34].
The similar trend was observed in PO4P adsorption performance
as shown in Fig. 6 (b). In the case of IR 2.5, PO4 uptake decreased
linearly from size range 0.61.18 mm to 0.150.3 mm. The maximum
removal efciency of 65.4% and minimum removal efciency of 19.2%
occurred at size range 0.61.18 mm and 0.150.3 mm, respectively.
For IR 1.5, PO4P uptake decreased slightly from size range 0.6
1.18 mm to 0.30.425 mm. The maximum removal efciency of 76.9%
and minimum removal efciency of 57.7% occurred at size range 0.15
0.3 mm and 0.30.425 mm, respectively. The maximum removal at
smallest size range could be attributed to the dilution error during the
preparation of samples for spectrophotometer analysis. While for IR
0.5, the removal efciency was the minimum at low size range 0.15
0.3 mm which removed only 19% of PO4P. The PO4P uptake capacity
decreased with the decrease in particle size. The PO4P uptake
capacity was a function of iron content that was bound in the
adsorbent adsorption sites. Since the uptake of iron increased with
increasing particle size, it could possibly result in the larger particle
size providing a greater amount of iron for phosphate removal from
solution. Similar theory was proved in a study conducted by Eberhardt
and Min [35].

3.6. Adsorption isotherms

Equilibrium isotherms were analyzed using the Langmuir and

Freundlich isotherms. Both Langmuir and Freundlich isotherms are
the most common isotherm models used to describe the adsorption
characteristics of the adsorbents used in water and wastewater
treatments [36]. The Langmuir isotherm is a typical model for
Fig. 6. Effect of particle size on removal performance of (a) total iron, and (b) PO4P monolayer adsorption based on assumption that sorption takes
using CGAC. place at specic homogenous sites within the adsorbent [37].
232 S.L. Ching et al. / Desalination 279 (2011) 225234

The Langmuir isotherm is expressed as:

 
0 0
1 = qe = 1 = Q + 1 = Q b Ce : 4

The Freundlich isotherm is an empirical equation based on the

assumption that adsorption takes place on heterogeneous surfaces.
The Freundlich isotherm model is described as [38]:

logqe = logK + 1 = n logCe 5

where qe is the amount of solute adsorbed per unit weight of

adsorbent (mg/g), Ce is the equilibrium concentration of adsorbate in
solution after adsorption (mg/L) and K(mg/g(L/mg)1/n), n, Q 0 (mg/g)
and b (L/mg) are the constants.
The results of the total iron adsorption obtained by each isotherm are
shown in Fig. 7 (a, b). According to the results shown in Fig. 7 (a) for
Langmuir isotherm and Fig. 7 (b) for Freundlich isotherm, both these
isotherms showed very good ts for adsorbate with corresponding
coefcient (R2) values above 0.9 except for IR 0.5 in case of Langmuir
isotherm. Thus, it could be concluded that both isotherms described the
adsorption mechanism of total iron using the CGAC as adsorbent
indicating that the adsorption of total iron on CGAC took place as
monolayer adsorption on the surface homogenous in adsorption afnity.
In case of PO4P (Fig. 8a), Langmuir isotherm t the data better for
IR 2.5 as indicated by the value of R 2 greater than 0.9. However, the
Langmuir isotherm for IR 0.5 and 1.5 showed values of R 2 lower than
0.9 representing that the data was not described by the isotherm. A
similar behavior was observed for Freundlich isotherm with lower Fig. 9. Pseudo rst (a) and second order kinetic model (b) for total iron.
values of R 2 (Fig. 8b).
3.7. Adsorption kinetics

Several kinetic models may be used to examine the controlling

mechanism of adsorption process and the potential rate limiting steps
such as transfer and chemical reaction [39]. In this study, the adsorption
kinetics was predicted by two common models which were pseudo-
rst-order and pseudo-second-order model. The pseudo-rst-order
kinetic equation is based on the sorption capacity of the solid phase. It
predicts the behavior over the whole range of adsorption which is in
agreement with chemical adsorption mechanism being the rate
controlling step which may involve valency forces through sharing or
exchange of electrons between adsorbate anions and adsorbent [39].
The pseudo-rst-order model is generally expressed as:
0
dq = dt = k1 qe q: 6

The pseudo-second-order model is expressed as:

0
dq = dt = k2 qe q 7

where qe is the amount of solute adsorbed at equilibrium per unit

weight of adsorbent (mg/g), q the amount of solute adsorbed at time t
per unit weight of adsorbent (mg/g) while k1 (min 1) and k2 (g/
mg min) are the constants.
It was observed that total iron did now t very well in pseudo-rst-
order model (Fig. 9a), however it tted very well in pseudo-second-
order model with R2 values more than 0.9 (Fig. 9b). The high correlation
coefcient (R2) values indicated that the model successfully described
the adsorption kinetics of the parameters. Similar was true in the case of
PO4P where the values of R2 were less than 0.9 for pseudo-rst-order
model while they were higher than 0.9 for pseudo-second-order model
Fig. 8. Langmuir (a) and Freundlich isotherm (b) for PO4P. except IR 2.4 (Fig. 10). Overall, it implied that the adsorption of both total
 S.L. Ching et al. / Desalination 279 (2011) 225234 233

[4] M. Hirata, N. Kawasaki, T. Nakamura, K. Matsumoto, M. Kabayama, T. Tamura, S.

Tanada, Adsorption of dyes onto carbonaceous materials produced from coffee
grounds by microwave treatment, J. Colloid Interface Sci. 254 (2002) 1722.
[5] USDA, United States Department of Agriculture, Tropical Products: World Markets
and Trade [online]. Available from: 2009 www.fas.usda.gov/htp/coffee/2009/
2009/2009_Coffee_June.pdf [Accessed 10 April 2010].
[6] O. Ioannidou, A. Zabaniotou, Agricultural residues as precursors for activated
carbon production: a review, Renewable Sustainable Energy Rev. 11 (2007)
19662005.
[7] D. Mohan, C.U. Pittman Jr., Activated carbons and low cost adsorbents for
remediation of tri- and hexavalent chromium from water, J. Hazard. Mater. B137
(2006) 762811.
[8] D. Kalderis, D. Koutoulakis, P. Paraskeva, E. Diamadopoulos, E. Otal, J.O.D. Valle, C.
Fernandez-Pereira, Adsorption of polluting substances on activated carbons
prepared from rice husk and sugarcane bagasse, Chem. Eng. J. 144 (2008) 4250.
[9] Y.N. Lim, G. Shaaban, C.Y. Yin, Treatment of landll leachate using palm shell
activated carbon column: axial dispersion modeling and treatment prole, Chem.
Eng. J. 146 (2009) 8689.
[10] M. Heavey, Low-cost treatment of landll leachate using peat, Waste Manage. 23
(2003) 447454.
[11] M. Umar, H.A. Aziz, M.S. Yusoff, Trends in the use of Fenton, electro-Fenton and
photo-Fenton for the treatment of landll leachate, Waste Manage. 30 (2010)
21132121.
[12] M. Umar, H.A. Aziz, M.S. Yusoff, Variability of parameters involved in leachate
pollution index and determination of LPI from four landlls in Malaysia, Int. J.
Chem. Eng. Article ID: 747953 (2010), doi:10.1155/2010/747953.
[13] M. Umar, H.A. Aziz, M.S. Yusoff, Assessing the chlorine disinfection of landll
leachate and optimization by response surface methodology (RSM), Desalination
274 (2011) 278283.
[14] M.V. Lopez-Ramon, F. Stoeckli, C. Moreno-Castilla, F. Carrasco-Marin, On the
characterization of acidic and basic surface sites on carbons by various techniques,
Carbon 37 (1999) 12151221.
[15] B. Bestani, N. Benderdouche, B. Benstaali, M. Belhakem, A. Addou, Methylene blue
and iodine adsorption onto activated desert plant, Bioresour. Technol. 99 (2008)
84418444.
[16] ASTM D4607 - 94, Standard Test Method for Determination of Iodine Number of
Activated Carbon Available online at, , 2006 http://www.astm.org/Standards/
D4607.htm Accessed on 14-05-11.
[17] V. Gmez-Serrano, F. Piriz-Almeida, C.J. Durn-Valle, J. Pastor-Villegas, Formation
of oxygen structures by air activation. A study by FT-IR spectroscopy, Carbon 37
(1999) 15171528.
[18] A. Omran, A.O. El-Amrouni, L.K. Suliman, A.H. Pakir, M. Ramli, H.A. Aziz, Solid
waste management practices in Penang State: a review of current practices and
the way forward, Environ. Eng. Manage. J. 8 (2009) 97106.
[19] M.J.K. Bashir, M.H. Isa, S.R.M. Kutty, Z.B. Awang, H.A. Aziz, S. Mohajeri, I.H. Farooqi,
Fig. 10. Pseudo rst (a) and second order kinetic model (b) for PO4P. Landll leachate treatment by electrochemical oxidation, Waste Manage. 29
(2009) 25342541.
[20] Q. Qian, M. Machida, H. Tatsumoto, Preparation of activated carbon from cattle-
manure compost by zinc chloride activation, Bioresour. Technol. 98 (2007)
iron and PO4P followed second-order kinetics suggesting that the 353360.
adsorption was controlled by chemisorption [37]. [21] R. Malik, D.S. Ramteke, S.R. Wate, Adsorption of malachite green on groundnut
shell waste based powdered activated carbon, Waste Manage. 27 (2007)
11291138.
[22] B.H. Hameed, Spent tea leaves: a new non-conventional and low-cost adsorbent
4. Conclusions for removal of basic dye from aqueous solutions, J. Hazard. Mater. 161 (2009)
753759.
[23] J. Zhang, Z. Zhong, D. Shen, J. Zhao, H. Zhang, M. Yang, W. Li, Preparation of
The experimental impregnation ratios strongly affected the physical bamboo-based activated carbon and its application in direct carbon fuel cells,
characteristics of coffee grounds activated carbon. The optimum 2011 dx.doi.org/10.1021/ef200161c
impregnation ratio was found to be 0.5 to obtain higher specic surface [24] C. Namasivayam, D. Sangeetha, Equilibrium and kinetic studies of adsorption of
phosphate onto ZnCl2 activated coir pith carbon, J. Colloid Interface Sci. 280
area (224.7 m2/g) and micropore volume (0.07986 cm3/g). FTIR indicated
(2004) 359365.
that the presence of functional group OH, NH, CH, C_O, OH, CO and [25] L.J. Yu, S.S. Shukla, K.L. Dorris, A. Shukla, J.L. Margrove, Adsorption of chromium
POC in coffee grounds activated carbon. Impregnation ratio of 2.5 and from aqueous solutions by maple sawdust, J. Hazard. Mater. B100 (2003) 5363.
0.5 was optimum for the removal of total iron (77%) and PO4P (84%), [26] M.M. Rao, G.P.C. Rao, K. Seshaiah, N.V. Choudary, M.C. Wang, Activated carbon
from Ceiba pentandra hulls, an agricultural waste, as an adsorbent in the
respectively at 10 g dosages and pH 8.1. The optimum pH was 13 for total removal of lead and zinc from aqueous solutions, Waste Manage. 28 (2008)
iron removal and pHb 5 and pHN 11 for PO4P removal onto the coffee 849858.
grounds. The adsorption capacity of total iron increased with an increase [27] C. Namasivayam, D. Sangeetha, Equilibrium and kinetic studies of adsorption of
phosphate onto ZnCl2 activated coir pith carbon, J. Colloid Interface Sci. 280
of the adsorbent dosage. But it was exceptional for PO4P removal which (2004) 359365.
showed uctuation trend possibly due to the effect of foreign ions present [28] M. zacar, .A. engil, Adsorption of metal complex dyes from aqueous solutions
in the leachate. by pine sawdust, Bioresour. Technol. 96 (2005) 791795.
[29] S.M. Nomanbhay, K. Palanisamy, Removal of heavy metal from industrial
wastewater using chitosan coated oil palm shell charcoal, Electron. J. Biotechnol.
8 (2005) 4353.
References [30] A.B. Jusoh, W.H. Cheng, W.M. Low, A. Nora aini, M.J.M.M. Noor, Study on the
removal of iron and manganese in groundwater by granular activated carbon,
[1] A. Reffas, V. Bernardet, B. David, L. Reinert, M.L. Bencheikh, M. Dubois, N. Batisse, L. Desalination 182 (2005) 347353.
Duclaux, Carbons prepared from coffee grounds by H3PO4 activation: character- [31] E. Alvarez-Ayuso, A. Garcia-Sanchez, X. Querol, Purication of metal electroplating
ization and adsorption of methylene blue and Nylosan Red N-2RBL, J. Hazard. waste waters using zeolites, Water Res. 37 (2003) 48554862.
Mater. 175 (2010) 779788. [32] A.H. ren, A. Kaya, Factors affecting adsorption characteristics of Zn2+ on two
[2] A.J. Adi, Z.M. Noor, Waste recycling: utilization of coffee grounds and kitchen natural zeolites, J. Hazard. Mater. 131 (2006) 5965.
waste in vermicomposting, Bioresour. Technol. 100 (2009) 10271030. [33] L. Ruixia, G. Jinlong, T. Hongxiao, Adsorption of uoride, phosphate, and arsenate
[3] M.A. Silva, S.A. Nebra, M.J. Silva, C.G. Sanchez, The use of biomass residues in the ions on a new type of ion exchanger ber, J. Colloid Interface Sci. 248 (2002)
Brazilian soluble coffee industry, Biomass Bioenergy 14 (1998) 457467. 268271.
234 S.L. Ching et al. / Desalination 279 (2011) 225234

[34] C. Namasivayam, D. Sangeetha, Removal of molybdite from water by adsorption [37] M.H. Isa, L.S. Lang, F.A.H. Asaari, H.A. Aziz, N.A. Ramli, J.P.A. Dhas, Low cost removal
onto ZnCl2 activated coir pith carbon, Bioresour. Technol. 97 (2006) 11941200. of disperse dyes from aqueous solution using palm ash, Dyes Pigm. 74 (2007)
[35] P. Mondal, C.B. Majumder, B. Mohanty, Effects of adsorbent dose, its particle size and 446453.
initial arsenic concentration on the removal of arsenic, iron and manganese from [38] Y.R. Lin, H. Teng, Mesoporous carbons from waste tire char and their application in
simulated ground water by Fe3+ impregnated activated carbon, Hazard. Mater. 150 wastewater discoloration, Microporous Mesoporous Mater. 54 (2002) 167174.
(2008) 695702. [39] C. Duran, D. Ozdes, A. Gundogdu, H.B. Senturk, Kinetics and isotherm analysis of
[36] T.L. Eberhardt, S.H. Min, Biosorbents prepared from wood particles treated with basic dyes adsorption onto almond shell (Prunus dulcis) as a low cost adsorbent,
anionic polymer and iron salt: effect of particle size on phosphate adsorption, J. Chem. Eng. Data 56 (2011) 21362147.
Bioresour. Technol. 99 (2008) 626630.