Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: Being plentiful and readily available in many parts of the world, brown coal is probably one of the most
Received 10 December 2015 potential substitutes for activated carbon. Many reports include the brown coal drying stage prior to
Received in revised form 10 May 2016 adsorption despite the fact that it may lead to coal shrinkage and reduction of its pore volume, thus
Accepted 7 June 2016
diminishing its adsorption capacity. In this study, various types of raw (wet) brown coal were investi-
Available online 20 June 2016
gated as adsorbents for anionic and cationic dyes in aqueous solution. The adsorption capacities of these
raw brown coals were compared with those of oven dried brown coals and also with activated carbon. It
Keywords:
was found that raw brown coals exhibit high adsorption capacities for cationic dye, but very low capac-
Brown coal
Cationic dye
ities for anionic dyes, suggesting the involvement of a cation-exchange mechanism. Additionally, the
Adsorption capacity cation exchange capacities correlate positively with its cation adsorption capacity highlighting the major
role of strongly acidic groups in the adsorption process. Drying the brown coal significantly reduces its
adsorption capacity for cationic dye, supporting the hypothesis of coal shrinkage and the consequent
reduced pore volume upon drying. These findings suggest that raw (wet) brown coal may perform as
well as activated carbon for the removal of positively charged species in aqueous solution. Regeneration
of the used adsorbent may also not be necessary, since the used brown coal can be simply burnt as fuel
or composted.
© 2016 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jwpe.2016.06.004
2214-7144/© 2016 Elsevier Ltd. All rights reserved.
44 G. Yuliani et al. / Journal of Water Process Engineering 15 (2017) 43–48
The FTIR analysis was conducted to identify the lignite’s func- The adsorption capacity (qe ) represents the amount of dye
tional groups. All lignite samples were dried at 105 ◦ C under N2 molecules adsorbed per gram of adsorbent (mg/g) and can be cal-
for 5 h to reduce the intensity of OH bands. The KBr pellets were culated using equation 1.
obtained using 1 part of coal and 200 parts of KBr. The FTIR spectra
were collected using a Perkin-Elmer Spectrum 100 FT-IR spectrom- V
qe = (C0 − Ce )( ) (1)
eter. The analyses for all samples were performed at the same time m
(a) (b)
Fig. 1. Chemical structure of model dyes: (a) SO (cationic) (b) AR (anionic).
G. Yuliani et al. / Journal of Water Process Engineering 15 (2017) 43–48 45
Fig. 2. FTIR spectra of brown coals. Fig. 3 shows the equilibrium adsorption of SO, a cationic dye,
on brown coal and coal-derived activated carbon. LYNA coal and
500 YL exhibited higher adsorption capacities than the coal-based acti-
450 vated carbon. MV coal had lower adsorption capacity than activated
400 carbon and LY coal had the lowest adsorption capacity for the
cationic dye. The initial adsorption slope was higher for activated
350
carbon than for coals, indicating faster uptake of dye molecules. For
qe (mg g-1)
300
LYNA LYNA, YL and MV, the initial dye uptake rates were similar.
250 The upper limits of the adsorption capacities of the brown coal
YL
200 samples for safranin dye were mostly above 150 mg/g. This is higher
AC
150 than some alternative adsorbents reported in the literature for the
MV
100 same dye. For example, the adsorption capacity of adsorbent pre-
LY pared from rice husk for safranin adsorption when calculated with
50
the Langmuir adsorption isotherm was only 35.8 mg/g[18], while
0
adsorption capacities of kaolinite and alumina for the similar dye
0 20 40 60 80 100
Ce (mg L-1)
were reported to be equal to 53.7 mg/g and 0.01 mg/g, respectively
[19].
Fig. 3. Adsorption of SO dye on brown coal and activated carbon as a function of The adsorption data were fitted with the Langmuir and Fre-
equilibrium concentration in solution. undlich isotherm models (Fig. 4) to provide insight on the nature
of the interaction between dye and adsorbent. All adsorption data
whereC0 is the initial concentration of the dye solution (mg/L), Ce is fit the Langmuir isotherm better than the Freundlich isotherm
the equilibrium concentration of the dye solution (mg/L), m is the (Table 3), as indicated by the correlation coefficients (R2 ) greater
amount of adsorbent used (dry basis, g) and V is the volume of dye than 0.99. The Langmuir model characterises the adsorption in
solution (L). terms of a monolayer interaction between the adsorbate molecules
Some models are available to correlate the experimental adsorp- and the adsorbent surface. The agreement between experimental
tion data and may provide useful information on adsorption data (for all brown coal samples) and the Langmuir isotherm sup-
behavior and the heterogeneity of adsorbent surface. The isother- ports the concept of a strong monolayer interaction; this would
mal equilibrium sorption data obtained from the batch experiment be expected from a cation-exchange mechanism. The maximum
was processed according to Langmuir and Freundlich models. The adsorption capacities (Q 0 ) of LYNA and YL coal for safranin were
Langmuir equation is represented as follow [16,17]: similar when calculated using the Langmuir equation (Table 2).
The overall trend of the adsorbents’ performance in removing
Ce Ce 1
= 0 + 0 (2) cationic dye from aqueous solution was: LYNA > YL > Activated car-
qe Q Q KL
bon > MV > LY.
whereCe is the concentration of the dye remaining in the solution The adsorption capacity of all brown coals for the anionic dye
(mg/L), qe is the amount of dye adsorbed per gram of adsorbent (alizarin red) decreased significantly (Fig. 5). In contrast, activated
(mg/g), Q 0 is the maximum surface coverage (formation of mono- carbon maintained high capacities for both dyes. While the adsorp-
0.8 3
0.7 2.5
0.6
LYNA 2
Ce/qe (g L-1 )
0.5
AC
log qe
YL
0.4 1.5
AC LYNA
0.3
1 YL
0.2 MV
0.5 MV
0.1 LY
LY
0 0
0 50 100 150 -1 0 1 2 3
Ce (mg L-1) log Ce
(a) (b)
Fig. 4. (a) Langmuir and (b) Freundlich isotherms for SO adsorption on activated carbon and brown coal.
46 G. Yuliani et al. / Journal of Water Process Engineering 15 (2017) 43–48
Table 2
Surface area of brown coals and coal-based activated carbon.
AC LY Brown coal
LYNA MV YL
1200 were used to predict the van der Waals surface of the dyes
[22]. The volumes occupied by the safranin and alizarin red ions
1000 averaged 305.4 Å3 and 243.3 Å3 , respectively at 298 K. Therefore,
AC
activated carbon can accommodate more alizarin red molecules
800 YL within its micropores than safranin. It is also likely that the smaller
qe(mg g-1)
MV alizarin red molecules can penetrate smaller pores from which the
600 safranin would be excluded. This suggests physisorption as domi-
LYNA nant adsorption mechanism for activated carbon. In physisorption,
400 the surface area and pore volume typically govern the adsorption
LY
efficiency. The activated carbon has high surface area and pore
200 volumes for both micro and mesopores (Table 2).
The surface areas of the brown coal samples are roughly one
0
quarter of the activated carbon (Table 2). Despite their lower
0 50 100 150
Ce (mg L-1) surface areas, brown coals have considerably higher adsorption
capacities for the cationic dye (Table 3). This suggests that an addi-
Fig. 5. Adsorption of AR dye on brown coal and activated carbon as a function of tional mechanism, such as cation-exchange, is enhancing cationic
equilibrium concentration in solution. dye adsorption. Chemisorption appears to be the dominant mech-
anism for safranin adsorption.
tion capacity of activated carbon for anionic dye was higher than The adsorption capacities of brown coal for anionic dye were
that for cationic dye, the adsorption capacity of the brown coal for considerably less than a quarter of those of activated carbon,
anionic dye was less than 15% of that for cationic dye. The adsorp- as expected from pure physisorption (proportional to surface
tion of the anionic dye adsorption onto brown coals did not fit the area) (Table 2). The negatively-charged adsorbate molecules were
Freundlich model at all. Therefore, only the Langmuir model was unable to access the available pores of the brown coal surface.
used to plot the adsorption data (Fig. 6). The negatively-charged surface might prevent the adsorbates from
The maximum adsorption capacities (Q 0 ) of brown coals for adsorbing onto the available pores on brown coal structure. This
alizarin red dye adsorption were insignificant when compared to also suggests the predominant role of chemisorption in the adsorp-
activated carbon (Table 3). The Q 0 values for all coals were less tion process using brown coal.
than 4% of that for coal-based activated carbon. Also the Langmuir
isotherm fits are relatively poor and the K values are very low. How-
ever, the overall trend of the brown coals’ performance in removing 3.2. Effect of brown coal drying
anionic dye from aqueous solution was: LY > YL > MV > LYNA.
Most activated carbons have a very large specific surface area, Raw brown coal having a moisture content between 55 and 57%
typically above 800 m2 /g, with micropores contributing over 80% was reported to have Langmuir isotherm saturation capacity of
of the total surface area groups [16]. This provides ample adsorp- 286–370 mg/g for methylene blue adsorption [15]. On the other
tion sites for the dye molecules to establish physical interactions hand, the adsorption capacities of dry and pyrolysed brown coal
with the adsorbent’s surface (physisorption). Additionally, acti- for methylene blue were only 45 mg/g and 25 mg/g, respectively
vated carbons possess functional groups with both basic and acidic [23,24]. The dry and pyrolysed coals exhibited lower adsorption
properties. The acid functional groups are carboxylic, lactonic, and efficiency and effectiveness presumably due to surface shrinkage
phenolic, while the basic functional groups include ketonic and and possible preoxidation and crosslinking during the drying pro-
pyronic oxygen-containing species [20]. These functional groups cess.
may play a role in the adsorption mechanism such as increasing the In this study, three Victorian brown coals (LYNA, MV and YL)
interaction of activated carbon with polar adsorbates. However, the were dried under N2 for 3 h and the adsorption capacities for
density of these functional groups is low, typically between 0.54 the cationic dye solution before and after drying were compared
to 14.9 group nm−2 , with the total functional groups concentra- (Fig. 7). It can be seen that the adsorption capacities of all coal sam-
tion of around 1 mmol g−1 [21]. Because of their low concentration, ples were diminished as a result of drying. Additionally, positive
they are probably not the main factor influencing the high capac- correlation between the adsorption capacities (Q 0 ) of raw brown
ity of activated carbon for either anionic or cationic species in the coals and moisture content (Table 1) was observed. Coal sample
wastewater. having the lowest moisture content, LY, had the lowestQ 0 . This
The higher adsorption capacity of activated carbon for alizarin supports the role of water molecules occupying the coal pores in
red compared to safranin may be due to the bulkier molecular maintaining the active site for adsorption and the hypothesis of
structure of safranin (Fig. 1.). Molecular dynamics1 simulations
rials Studio® suite of programs (Accelrys, San Diego). The occupied volume was
1
Molecular dynamics simulations were carried out for each of the adsorbate ions determined from the Connolly surface which uses the van der Waals radii of the
at 298 K for 100 ps using the Forcite module and the Compass forcefield of the Mate- atoms in the structure.
G. Yuliani et al. / Journal of Water Process Engineering 15 (2017) 43–48 47
Table 3
Langmuir and Freundlich isotherm constants and correlation coefficients for SO and AR dyes adsorption on brown coal and activated carbon.
Adsorbent Langmuir Isotherm SO(cationic) dye Freundlich Isotherm SO(cationic) dye Langmuir Isotherm AR (anionic) dye
Q 0 (mg/g) KL (L/mg) Correlation coefficient (R2 ) KF (mg/g) n Correlation Coefficient (R2 ) Q 0 (mg/g) KL (L/mg) Correlation coefficient (R2 )
LYNA 476 0.58 0.9998 167 3.2 0.9033 5.76 0.01 0.8806
YL 476 0.5 0.9982 209 5.3 0.9788 30.4 0.01 0.658
MV 357 0.82 0.9995 157 4.6 0.9411 10.4 0.02 0.9779
LY 139 0.27 0.9829 48 4.1 0.9007 34.7 0.04 0.9902
AC 357 56 0.9998 293 13 0.7318 1000 0.91 0.9091
LY
Brown coal has many of the oxygen containing groups associ-
6
LYNA ated with its structure, present as carboxyl and phenolic groups
4 [29]. When added to water or dye solutions, these acidic groups
lose their hydrogen ions due to ionization that result in lowering
2 the pH of water or solutions. This mechanism forms a negatively
charged layer on the surface that leads to its ability to interact well
0 with positively charged molecules. The presence of these carboxyl
15 65 115 165 groups and, to a lesser extent, the phenolic groups possibly aug-
Ce (mg/L) ments the adsorption of positively charged dye onto brown coal
[30].
Fig. 6. Langmuir isotherm adsorption for AR on brown coals.
Coal-COOH Coal-COO− + H+ (i)
+ −
Dye + nCoal-COO = (Coal-COO-Dye) (ii)
surface shrinkage and pore collapsing of brown coal upon drying
[25]. To confirm the role of oxygen containing groups and cation
The removal of water from brown coal is always accompanied exchange mechanism in brown coal’s adsorption, the cation-
by shrinkage even at low temperature (30 ◦ C) in vacuum [26]. Sig- exchange capacity of the raw coal is plotted versus the adsorption
nificant levels of pore collapsing and crosslinking occur during capacity for cationic dye (Fig. 8). The adsorption capacity of coal
drying at 105 ◦ C, the standard temperature for drying coal [25]. generally increased as the cation-exchange capacity increased. The
During evaporation, water molecules occupying the large pores are low adsorption capacity of LY (Fig. 3 and Table 3), for example,
lost first, followed by water in the smaller pores. Coal shrinkage may be due to its low cation exchange capacity. This result sup-
occurs, but the structural changes are relatively small at this point ports the involvement of oxygen functionalities in cation-exchange
due to the ordered arrangement and the rigidity of the coalified mechanism of positively charged dye adsorption.
500 500
400
qe (mg g-1 db)
400
qe (mg g-1 db)
300 300 YL
MV
200 200 YL dried
MV dried
100 100
0
0
0 50 100
0 50 100
Ce (mg L-1) Ce (mg L-1)
500
qe (mg g-1 db)
400
300
LYNA
200 LYNA dried
100
0
0 10 20 30 40
Ce (mg L-1)
Fig. 7. Equilibrium adsorption of cationic dye (safranin) onto raw and dried brown coal.
48 G. Yuliani et al. / Journal of Water Process Engineering 15 (2017) 43–48
40 [5] C.J. Butler, A.M. Green, A.L. Chaffee, The influence of water quality on the
LYNA reuse of lignite-derived waters in the Latrobe Valley, Australia, Coal Prep.
Effecve caon exchange capacity,