Chemical Engineering Journal 197 (2012) 250–260

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Development of novel in situ nickel-doped, phenolic resin-based

micro–nano-activated carbon adsorbents for the removal of vitamin B-12
Rishabh Saraswat a, Neetu Talreja a, Dinesh Deva b, Nalini Sankararamakrishnan a, Ashutosh Sharma b,c,⇑,
Nishith Verma a,c,⇑
a
Center for Environmental Science and Engineering, Indian Institute of Technology Kanpur, Kanpur 208 016, India
b
DST Unit of Soft Nanofabrication, Indian Institute of Technology Kanpur, Kanpur 208 016, India
c
Department of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur 208 016, India

h i g h l i g h t s g r a p h i c a l a b s t r a c t

" Synthesis of Ni-doped activated A specially designed packed column was developed for applying micro–nanoparticles based adsorbents
carbon spheres and carbon to the aqueous flow without entrainment or channeling in the column. The breakthrough analysis
micro–nanoparticles. revealed that the VB12 uptake in the bed was comparable to the equilibrium adsorption data obtained
" Synthesized materials used as from the batch study. This indicates an effective bed design allowing a uniform flow distribution and
efficient adsorbents for vitamin minimal particle entrainment and loss by flow. The simple methodology for developing micro–nanopar-
B-12. ticles based adsorbents and applying them in the column under dynamic conditions has potential for a
" A specially designed column for wide range of bio-separation and bio-purification applications.
breakthrough study on
micro–nanoparticles.

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

Article history: Activated carbon spheres (ACSs) and carbon micro–nanoparticles derived from nickel (Ni)-doped pheno-
Received 10 March 2012 lic beads (0.8 mm) were synthesized as efficient adsorbents for vitamin B-12 (VB12). Ni was incorpo-
Received in revised form 13 May 2012 rated in an intermediate step during a suspension polymerization. ACSs were synthesized by the
Accepted 14 May 2012
carbonization and activation of Ni-doped polymeric beads. Carbon micro–nanoparticles (average size
Available online 23 May 2012
200 nm) were synthesized by milling polymeric beads followed by carbonization and activation.
Adsorption tests were carried out under both batch and dynamic (flow) conditions. The latter tests were
Keywords:
carried out in a specially designed and fabricated micro-column packed with carbon micro–nanoparti-
Vitamin B-12
Adsorption
cles. The adsorption loading of VB12 on the adsorbents was determined to be 300 mg/g, corresponding
Carbon micro–nanoparticles to an aqueous phase concentration of 500 ppm, which is comparable to or larger than the literature data.
Biochemical engineering The method of in situ synthesis of metal incorporated carbon micro–nanoadsorbents and their use in a
Environment micro-column under the flow conditions has much potential for bio- and pharmaceutical separation
Packed bed and purification applications.
Polymerization Ó 2012 Elsevier B.V. All rights reserved.

⇑ Corresponding authors. Fax: +91 512 2590104.

E-mail addresses: ashutos@iitk.ac.in (A. Sharma), nishith@iitk.ac.in (N. Verma).

1385-8947/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cej.2012.05.046
 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260
1. Introduction
Vitamin B-12 (VB12), which is a highly valued pharmaceutical
compound, is the largest and most complex water soluble vitamin
– essential for the normal growth of human health [1,2]. VB12 is
industrially produced by fermentation driven by micro-organisms.
The recovery of pharmaceutical compounds by adsorption from
crude sources or downstream from the fermentation process is
considered to be more energy efficient compared to membrane-fil-
tration and conventional separation techniques such as chroma-
tography and extraction [3].
Polymeric resin-based adsorbents have been extensively used
for the removal of pharmaceutical compounds from water. Some
common examples of precursors for such polymeric adsorbents
are styrene, methyl acrylate, and phenol [4–13]. However, these
adsorbents generally have low adsorption capacities for large mol-
ecules and compounds such as vitamins and proteins [3–5].
Polymers doped with metals, metal oxides, and inorganic ion
exchangers have also found various applications. Polymer/inor-
ganic hybrid composites have been developed as adsorbents for
pharmaceutical compounds through the dispersion of inorganic
nanoparticles within a polymer-matrix [14]. Styrene/divinylben-
zene-based gel-like materials (150 lm) doped with iron have
been prepared for cell labeling [15]. Iron particles have also been
incorporated into polystyrene-beads (average size 80 lm) and
this powdery material produced has been used as a scavenger resin
[16]. Similar micron sized styrene-based particles (7 lm) that
incorporate Fe-particles have been synthesized by suspension
polymerizations for bio-applications, including protein purification
and recovery of enzymes [17].
In recent times, owing to it highly developed porous structure
and large surface area, activated carbon (AC) derived from poly-
meric resins has been used as a potential adsorbent for VB12
[18,19]. Being chemically inert and stable, AC has also shown
excellent biocompatibility relative to its inorganic or organic coun-
terparts without requiring special coatings [20,21]. There have also
been studies on the modification of AC as a way to improve its
adsorption capacity for VB12, for example, by coating it with poly-
methyl methacrylate [22], blending it with ferrocene [19], and
treating it with supercritical water [23]. Recently, carbon nano-
tubes (CNTs) have been suggested as potential adsorbents for
VB12 removal [24].
The method of in situ incorporation of metals during a polymer-
ization step to produce metal-doped polymeric beads (0.8 mm) is
relatively new [25,26]. By this study, iron- and aluminum-doped
polymeric beads could be produced. Carbonization and activation
of the metal-doped polymeric beads produced activated carbon
spheres (ACSs), which were used as adsorbents for the removal
of arsenic and fluoride ions from wastewater. Such carbon-based
materials had metal nanoparticles uniformly dispersed within
their micro-meso porous structures. Ball-milling of the iron/alumi-
num-doped polymeric beads, followed by carbonization and acti-
vation produced carbon micro–nanoparticles (100 nm) whose
adsorption performance with fluoride and arsenic ions were even
more encouraging. Inspired by these studies, we have carried out
the present study on the synthesis of nickel (Ni)-doped, carbon-
based adsorbents for the recovery or removal of VB12. ACSs and mi-
cro/nanocarbon particles are derived from Ni-doped phenolic
beads. ACSs are synthesized by carbonization and activation of
the beads, whereas carbon micro–nanoparticles are synthesized
by milling the beads followed by carbonization and activation.
The present study has three novelties. (1) Ni-salts are success-
fully incorporated in situ in the phenolic precursor during a poly-
merization step to produce spherical beads (0.8 mm). (2) A
special packed bed column is designed and fabricated for using car-
bon micro–nanomaterials (200 nm) as adsorbents under flow
251
(dynamic) conditions, and (3) Ni impregnated carbon micro/nan-
obeads are shown to be effective for the removal of VB12.
2. Materials and methods
2.1. Chemicals
All reagents used in the experiment were high purity grade.
Phenol, hexamethylene tetraamine (HMTA), formaldehyde (37–
41%), triethylamine (TEA), 95% hydrolyzed poly vinyl-alcohol
(PVA) (MW = 95,000), nickel nitrate (99%), 1-heptane sulfonic acid
sodium monohydrate, and cyanocobalamin as VB12 were pur-
chased from Merck (Germany) and used as received. All solutions
were prepared in water obtained from a Milli-Q system (Millipore,
Bedford, MA, USA).
2.2. Synthesis of Ni-doped phenolic beads
Ni-doped phenolic resin beads were synthesized by a suspen-
sion polymerization. A 2 l round bottom flask made of glass was
used as a reactor for polymerization. The three-neck reactor was
equipped with a condenser–reflux assembly, a thermometer and
a half round stirrer.
Phenol was used as a monomer and 50 g of it was mixed with
60 ml of formaldehyde in the flask. To this mixture, 1.5 ml of TEA
was added as a catalyst. The reaction mixture was stirred at a con-
stant speed of 250 rpm for 3 h at room temperature (30 °C ± 1 °C)
to prepare a homogeneous solution. Next, 200 ml of distilled water
was added to the solution. Subsequently, the solution turned milky
white. The speed of the stirrer was increased to 380 rpm. After
30 min, 3.5 g of HMTA was added as a cross linking agent and
water flow in the reflux–condenser was initiated. The flask was
simultaneously heated at a rate of 3 °C per min until the tempera-
ture reached 100 °C. Subsequently, 3 g of PVA was added as a sus-
pension stabilizing agent. The time that PVA was added was found
to be critical to the quality (shape and yield) of the synthesized
beads. After approximately 35 min, a yellowish gel-like substance
began to appear with the simultaneous formation of froth in the
reactor, which was the indication of the incipience of polymeriza-
tion. At this stage, 7.5 g of NiNO3-crystals were added to the reac-
tion mixture, and the color of the suspension mixture turned light
green. Stirring was continued. The formation of solid beads started
after approximately 1 h. After another 2 h, heating was stopped
and the reactor was allowed to cool with continued stirring. The
stirring was stopped when the temperature of the reactor reached
room temperature.
The product was filtered and washed three times with distilled
water, followed by methanol and acetone. The materials produced
were dried in still air for 12 h and then sieved into beads of differ-
ent sizes. In a typical experiment, the yield was approximately 30 g
(50% w/w), with an average size of the synthesized beads of
0.8 mm.
2.3. Synthesis of activated carbon spheres and micro–nanoparticles
The Ni-doped phenolic resin spherical beads were carbonized at
950 °C for 1 h under nitrogen flow in a horizontal tube furnace. The
carbonized beads were then activated at 900 °C for 1 h using steam
(50 sccm) to produce ACSs. Upon carbonization and activation,
approximately 40% weight-loss was observed, and the average par-
ticle size decreased to 0.5 mm. Carbon micro–nanoparticles were
prepared by milling the Ni-doped polymeric beads in a nano-ball
mill (Retsch planetary ball mill, PM100, Germany). Ball-milling
was carried out at 350 rpm for 24 h using 10 mm-Tungsten carbide
grinding balls (25 nos) in a 250 ml grinding jar. The milled particles
252 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260
were then carbonized and activated under similar conditions to
those used for producing carbon spheres.
Fig. 1 provides a flow-chart representing the production of the
Ni-incorporated polymer-derived activated carbon spheres and
micro–nanoparticles. The materials have been labeled for reference
purposes in this article. For example, PhB represents the phenolic
beads and Ni_PhB represents the phenolic beads doped with Ni-
salts, whereas PhB_A and Ni_PhB_A represent activated carbon
spheres derived from the phenolic beads and the Ni-doped pheno-
lic beads, respectively. PhB_BM_A and Ni_PhB_BM_A represent the
carbon micro–nanoparticles produced upon milling, followed by
carbonization and activation of phenolic beads and Ni-doped phe-
nolic beads, respectively.
2.4. Batch adsorption study
A 1 l stock solution of 500 ppm VB12 in Milli-Q water was pre-
pared for the adsorption tests. From the stock solution, 50 cc of test
solutions having different concentrations (10–500 ppm) were pre-
pared in conical flasks. A small amount (0.04 g) of the prepared
adsorbents was transferred to the conical flasks. The flasks contain-
ing the test solutions were kept in a mechanical shaker (125 rpm) at
30 °C. All test samples were prepared in triplicate for reproducibil-
ity. Equilibrium was attained in approximately 6 h, and the solutions
were then filtered on Whatman filter paper to separate the liquid
from the adsorbents. The liquid was then centrifuged at 1000 rpm
for 30 min to precipitate the suspended particles out of the solution.
The centrifuged solution was filtered using 0.22 lm-filter papers.
Fig. 1. Process flow for the synthesis of Ni-doped phenolic precursor based adsorbents.
The concentration of VB12 in the solution was measured using
high performance liquid chromatography (HPLC) (Varian Prostar
325, Australia) equipped with a UV/VIS single WL detector, a bin-
ary pumping system, a main column (Pursuit 5 C18, 250 mm
L  4.6 mm ID) and a guard column (Metaguard Pursuit 5 C18,
4.6 mm ID). The mobile phase (flow rate = 1 ml/min) consisted of
methanol and an aqueous phosphate buffer (pH = 3.5) of 1-heptane
sulfonic acid sodium monohydrate in a ratio of 25:75 (v/v). The
wavelength of the detector was set to 220 nm. The retention time
for VB12 was observed to be approximately 6.5 min following the
injection of samples in the injector-loop (volume = 2.5 ll).
The adsorbed amounts of VB12 by the prepared adsorbents were
calculated from the species balance equation: q = V(Ci  Ce)/W,
where q is the amount (mg/g) of VB12 adsorbed, Ci and Ce are the
initial and final (equilibrium) concentrations (mg/L), respectively,
V is the volume (L) of the solution, and W is the weight (g) of the
adsorbent.
2.5. Micro-column for breakthrough study
In general, micro–nanoparticles (100–500 nm) have relatively
large adsorption capacities because of their large specific surface
areas. It is extremely difficult to use micro–nanoparticles in a
packed bed column under flow conditions because such particles
are likely to be entrained in the flowing stream. Using a binder
phase to produce relatively large sized particles or beads from
micro–nanoparticles may result in the loss of active sties for
adsorption. There are several recent studies on using nanoparticles
 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260 253

in a column under fluidized conditions for different applications 3. Surface characterization

[27–29]. Packed bed of sand particles has been used recently as a
contactor between an aqueous suspension of Fe/Cu nanoparticles
and a nitrate solution for nitrate reduction applications [30]. Most
recently, a packed bed column of silica nanoparticles was used for
the separation of gallium and germanium [31]. The authors used
glass wool as a support for the nanoparticles in the column. This
arrangement can considerably reduce the possibility of entrain-
ment, but the risk of the support-materials themselves being en-
trained with the flow may still exist.
In the present study, a packing arrangement consisting of a thin
film composite (TFC) reverse osmosis (RO) membrane and stacked
layers of polymeric materials were used as a support for the micro–
nanoparticles in the column. Such an arrangement not only pre-
vented the entrainment of the adsorbent materials and support-
particles, but also minimized channeling and maldistribution of
flow through the column.
Fig. 2 provides a schematic of a Perspex-column (length =
60 mm, I.D. = 6 mm) specially designed and fabricated for the
breakthrough analysis. As shown, the column contains the adsor-
bents (micro–nanoparticles) sandwiched between stacked layers
of polymeric materials having graded (decreasing) sizes. The bot-
tom layers consist of PhB (0.8 mm), PhB (0.5 mm-average size),
and PhB_BM (0.05 mm-average size), which were produced by ball
milling for 3 h. The top layers consist of 0.05 mm-PhB_BM and
0.5 mm-PhB. This way, there exists a gradient of different bed
porosities in the stacked layers and the flow distribution becomes
uniform in the column. At both ends, a short Perspex stub is used
as an end-cap and is thread-fitted onto the main column. All of
the packing materials in the column rest on the TFC-RO membrane
made of polyamide (Permionics Membranes Pvt. Ltd., India). The
membrane is held firmly between Neoprene O-rings that were
press-fitted between the side-walls of the main column and the
stub to prevent any leakage of the solution. A thin stainless steel
(SS) mesh that was placed across the polyamide membrane helps
provide structural support.
3.1. Particle size distribution
Fig. 3 describes the particle size distribution for PhB_BM_A and
Ni_PhB_BM_A nanoparticles. The measurements were carried out
by a particle size analyzer (model: DelsaNano, Beckman Coulter,
US) based on the photon correlation spectroscopy technique. As
shown in the figure, the average particle size of PhB_BM_A and
Ni_PhB_BM_A were 500 nm and 200 nm, respectively. These
data also indicate that the Ni-doped beads produced micro–nano-
particles having narrower particle size distributions compared to
those derived from the beads without Ni. The average particle sizes
of the bead-materials, namely: PhB, PhB_A, Ni_PhB, and Ni_PhB_A,
were measured by sieve analysis and found to be approximately
1 mm, 0.8 mm, 0.8 mm and 0.5 mm, respectively. The analysis
indicates that incorporation of the metal caused a reduction in
the particle size and carbonization and activation of the beads re-
sulted in an approximately 20–30% reduction in their size. Further,
weight loss of the materials due to burn-off during carbonization
and activation was measured to be approximately 40% for PhB
and Ni_PhB, and 50% for Ni_PhB_BM.
3.2. Surface area, pore volume, pore size distribution (PSD)
The specific Brunauer, Emmett, and Teller (BET) surface area,
pore volume, and PSD of the adsorbents synthesized in this study
were determined from nitrogen adsorption/desorption isotherms
using an Autosorb 1C (Quantachrome, USA) instrument. The total
pore volume was measured from the amount of N2 adsorbed at a
relative saturation pressure close to unity (0.9994). The PSD was
calculated from desorption isotherms by the Barrett, Joyner, and
Halenda (BJH) method, applicable for mesopores (2–40 nm), and
from density functional theory (DFT) for micropores (<2 nm).
Fig. 4 shows the adsorption isotherms of the different adsor-
bents prepared in this study. The adsorption/desorption isotherm

3 4

5
2 6
11
1 7

9
12 10

Fig. 2. Schematic of column for dynamic (flow) study on VB12 adsorption by micro–nanoparticles: 1. Peristaltic pump, 2. VB12 in container, 3. Perpsex column, 4.
Ni_PhB_BM_A (200 nm), 5. Ni_PhB_BM (0.05 mm), 6. Ni_PhB (0.5 mm), 7. Ni_PhB (0.8 mm), 8. TFC RO-membrane, 9. O-rings, 10. SS mesh, 11. Thread-fitting, 12. End-stub.
254 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260

Fig. 3. Particle size distribution of (a) PhB_BM_A nanoparticles (avg size = 500 nm), and (b) Ni_PhB_BM_A nanoparticles (avg size = 200 nm).

600 Table 1 summarizes the BET surface area, total pore volume,
PhB_BM_A
Ni_PhB_BM-A
and PSD of the different materials. As observed from the table,
500 PhB_A the BET surface area of the phenolic beads with or without Ni
Ni_PhB_A was small (13 m2/g for the undoped phenolic beads and 2.8 m2/g
Ni_PhB, PhB
Volume adsorbed (cc/g)

for the Ni-doped phenolic beads). The addition of Ni-salts resulted

400
300
200
100
0
0.0 0.2 0.4
Relative pressure p/p o
Fig. 4. BET adsorption isotherm of different adsorbents.
data were obtained experimentally using N2 as an adsorbate at
77 K. Prior to the measurement, samples were degassed at 300 °C
for 8 h to remove any adsorbed or trapped gases in the samples.
As observed, the volume of adsorbed N2 gradually increased with
pressure, as measured relative to the saturation pressure (P/Po).
The amount of adsorbed N2 eventually became constant, which
indicates that the samples were equilibrated with N2. All adsorp-
tion isotherms were Type I according to BET classification. It should
be mentioned that none of the isotherms showed any significant
hysteresis loops, which is indicative of microporosity in the mate-
rials. A sharp increase in the amount of adsorbed N2 was observed
near the saturation pressure for PhB_BM_A. This increase sug-
gested the presence of relatively narrow pores in the sample [32].
in the partial blockage of pores. Milling followed by carbonization
and activation resulted in a significant increase in the BET area of
the prepared adsorbents. Further, the BET areas of nanoparticles,
PhB_BM_A and Ni_PhB_BM_A (1228 m2/g and 1175 m2/g) were
significantly larger than those of the ACSs, PhB_A and Ni_PhB_A
(858 m2/g and 768 m2/g). The BET areas (768 m2/g and 1175 m2/
g) of the ACSs and the carbon micro–nanoparticles derived from
the Ni-doped beads were slightly (10%) smaller than those (858
and 1228 m2/g) of the materials derived from the beads without
0.6 0.8 Ni, corroborating the previous observation that the in situ incorpo-
ration of Ni marginally affected the potential internal surface area
of the materials prepared in this study.
Table 1 also describes the volume-contents of micro, meso, and
macro pores of different adsorbents, as per the IUPAC rules. The
micro and mesopores contents of Ni_PhB were observed to be
insignificant. Fine pores develop after carbonization/activation.
Ni_PhB_BM_A, prepared by ball-milling followed by carboniza-
tion/activation has the majority of its volume consisting of microp-
ores. Similar PSD may be observed for the beads without Ni. The
variation observed for the total pore volume of the different mate-
rials is consistent with that of the BET surface area.
3.3. Scanning electron microscopy (SEM) and energy-dispersive X-ray
spectroscopy (EDX)
The surface morphology of the samples (PhB_A, Ni_PhB_A,
PhB_BM_A, Ni_PhB_BM_A) was studied using images obtained
 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260 255

Table 1
BET surface area and PSD of the synthesized adsorbents.

Adsorbent Average size (mm)* BET surface area (m2/g) Total pore volume (cc/g) Meso pore volume (cc/g) Micro pore volume (cc/g)
Ni_PhB 0.8 3 0.01 0.002 0.0004
Ni_PhB_A 0.5 768 0.328 0.010 0.287
Ni_PhB_BM_A 200 nm* 1175 0.709 0.160 0.351
PhB 1.0 13 0.02 0.004 0.005
PhB_A 0.8 858 0.461 0.021 0.397
PbB_BM_A 500 nm* 1228 1.107 0.030 0.404
*
The unit of the average size entered in column 2 of the table is ‘‘mm’’, except for the 3rd and 6th entries which are in ‘‘nm’’.

Fig. 5. SEM images of activated phenolic beads: (a) PhB_A (0.8 mm), (a0 ) Ni_PhB_A (0.5 mm), (b) PhB_A and (b0 ) Ni_PhB_A at high magnifications, (c) PhB_A and (c0 ) Ni_PhB_A,
after adsorption with VB12. EDX mapping (Co: blue and Ni: green): (d) Ph_B_A and (d0 ) Ni_PhB_A, after adsorption with VB12. (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)

from a field-emission SEM (Supra 40 VP, Zeiss, Germany). The ele- on the surface of the specimen to confirm the presence of different
mental composition at certain locations on the sample was deter- elements, including Ni and cobalt (Co). The latter is a constituent of
mined by EDX. EDX mapping was performed at a number of areas VB12 molecules.
256 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260

Fig. 5 shows representative SEM images and EDX mapping of
ACSs: PhB_A (left) and Ni_PhB_A (right). Fig. 5a and a0 shows that
both beads (with and without Ni) were approximately spherical
and porous. At higher magnification (Fig. 5b and b0 ), pores on the
external surface are visible, with lesser porosity observed on the
external surface of Ni_PhB_A. As discussed previously in this article
(Table 1), the BET surface area of Ni_PhB_A was determined to be
smaller than that of PhB_A, which is consistent with the observa-
tions from SEM. Fig. 5c and c0 shows SEM images of the adsorbents
equilibrated with VB12. Changes in the (external) surface morphol-
ogy of PhB_A and Ni_PhB_A after VB12 adsorption are clearly visi-
ble. Fig. 5d and d0 shows EDX maps of the VB12 adsorbed PhB_A
and Ni_PhB_A. The presence of Co (blue) in the former and that
of Co (blue) and Ni (green) in the latter are confirmed. Further,
the larger amount of Co observed on the EDX map for Ni_PhB_A
than on the PhB_A map indicates a larger adsorption capacity for
VB12 of the former than the latter, which is confirmed by the
adsorption study discussed later in this article.
Fig. 6 provides SEM images and EDX mapping of micro–nano-
adsorbents (with and without Ni: PhB_BM_A (left) and Ni_PhB_B-
M_A (right). Similar to the observation made from the SEM
images of the ACSs, a distinct difference in the surface morphol-
ogies between the fresh and spent (equilibrated with VB12)
adsorbents is visible (Fig. 6a–a0 and b–b0 ). The surface of the
adsorbents is partially covered with the solute following adsorp-
tion. Further, the extent of agglomeration of the adsorbed phase
is more extensive on the nanoadsorbent particles than on the
ACSs beads. Fig. 6c and c0 shows EDX mapping of the adsorbed
samples. If the number-density of Co shown on the maps is used
as a qualitative indicator of the adsorption of VB12, Ni_PhB_BM_A
may be considered to have a larger adsorption capacity than
PhB_BM_A; this is also corroborated by the adsorption data dis-
cussed later in this article. The Ni-incorporated nanoadsorbents
(Ni_PhB_BM_A) may also be considered superior to their Ni-ACS
counterpart, which is similarly confirmed from the adsorption
data.

Fig. 6. SEM images of phenolic precursor based activated nanoparticles: (a) PhB_A_BM (500 nm) and (a0 ) Ni_PhB_A_B (200 nm) before and (b) PhB_A_BM and (b0 )
Ni_PhB_A_BM, after adsorption with VB12. EDX mapping (Co: blue and Ni: green): (c) PhB_A_BM and (c0 ) Ni_PhB_BM_A, after adsorption with VB12. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260
4. Results and discussions
4.1. Atomic absorption spectroscopy (AAS) analysis of Ni-content in
the adsorbents
The amount of Ni incorporated into various adsorbents, or Ni-
loading on the adsorbents, was determined using AAS (Varian AA
240, USA). The measurements were carried out using an air-acety-
lene flame. The purpose of the AAS analysis was to determine if the
incorporated Ni particles were firmly held onto the surface of the
prepared adsorbents or if they were not removed from the materi-
als in solution during an adsorption study. AAS was also carried out
on the spent solutions obtained after contacting them with the Ni-
doped adsorbents under the severe conditions of leaching and
acid-digestion.
Leaching was carried out by mixing different adsorbents
(0.04 g) to an aqua-regia mixture (50 cc) contained in beakers.
The beakers were placed on a shaker (125 rpm) at 30 °C. Ni-leach-
ing was carried out for 6 h. Acid-digestion was carried out under
even more severe conditions. The beakers containing adsorbents
mixed with the aqua-regia solution were heated at 40 °C until all
the acid contents were evaporated. The remaining ash in the beak-
ers was mixed with distilled water (50 cc). The spent solutions
obtained after all three tests (regular adsorption, leaching, and
acid-digestion) were analyzed for Ni-content.
The AAS analysis revealed that no or an insignificant amount of
Ni was removed during the regular adsorption tests, indicating the
suitability of the adsorbents in the removal of VB12. The AAS data
obtained from the other two tests are reported in Table 2. As ob-
served, the Ni-content in all samples obtained from acid-digestion
tests are larger (approximately 1.5–2 times) than in the corre-
sponding leaching tests. These results are consistent with the se-
vere conditions used in the former.
Assuming that all or most of the Ni was removed during the
acid-digestion tests, the Ni-loading obtained for Ni_PhB may be
estimated as 17 mg/g, which is observed to significantly increase
to 27.5 mg/g in the Ni_PhB_BM_A nanoparticles. Ball-milling of
the metal doped beads followed by carbonization and activation
results in the production of 200 nm-particles. When these nano-
particles are leached or digested, relatively larger amounts of Ni
are removed from the adsorbents. On the other hand, the extent
of Ni removed from the micron-size (0.8 mm) beads by leaching
or digestion is relatively less. The differences between the Ni-load-
ings obtained for Ni_PhB, Ni_PhB_A, and Ni_PhB_BM are insignifi-
cant and within experimental error.
There is another important aspect of the development of an effi-
cient adsorbent. The spent adsorbents must be regenerated before
recycling them for reuse. After regeneration, there should not be
significant change in the adsorption characteristics of the materi-
als. In other words, the active surface sites of the materials should
be recoverable on desorption of the adsorbate (VB12 in the present
case). The incorporated metals should also remain adhered to the
surface during regeneration. The tests to address the regeneration
and stability aspects of the prepared materials are extensive and
will be addressed separately in a future study.
Table 2
AAS analysis of Ni-contents in adsorbents.
Sample Nickel-loading in adsorbent (mg/g)
Leaching-test Digestion-test
Ni_PhB 8.5 17.0
Ni_PhB_BM 7.5 19.5
Ni_PhB_A 9.25 18.7
Ni_PhB_BM_A 15.5 27.5
257
4.2. pH variation during adsorption
Experiments were carried out under batch conditions to ascer-
tain the pH variations in VB12 solutions during adsorption by dif-
ferent materials (PhB, Ni_PhB, PhB_A, Ni_PhB_A, PhB_BM_A, and
Ni_PhB_BM_A). The pH variation was monitored over an adsorp-
tion equilibrium time of approximately 6.5 h under the experimen-
tal conditions (shaker speed = 150 rpm, temperature = 30 °C). Fig. 7
shows the pH of the solutions before and after adsorption, after the
equilibrium time for different materials. The initial pH of the test
solutions varied from approximately 6.1 to 6.9 when increasing
VB12 concentrations from 5 ppm to 500 ppm. There was a marginal
decrease in pH after adsorption for all materials. Except for the
phenolic beads for which the pH decreased by approximately 1.1,
a maximum decrease of 0.25 was observed for all materials.
The relatively large decrease observed for phenolic beads is attrib-
uted to a possible modification of the polymeric surface containing
functional groups. We revisit this aspect later in this paper. The
above pH variation results have significance from the perspective
of pre- or post-treatment of water. In most cases, the pH of ground
or potable water is in the vicinity of 7 (usually between 6 and 7.5).
The marginal variation in pH during adsorption suggests that pre-
or post-treatment of water may not be required using the adsor-
bents prepared in this study.
Batch experiments were also carried out to ascertain the effect
of pH over the range 6–7 on the amount of VB12 removed from
solution by the different adsorbents. From the data (not presented
here for brevity), no appreciable change (less than 1%) in the
amount of VB12 was observed. Based on the results obtained in this
section, all experiments for VB12 adsorption were henceforth car-
ried out at pH = 7.
4.3. Batch adsorption study
Fig. 8 presents the solid phase equilibrium concentrations of
VB12 as a function of the aqueous phase concentrations for differ-
ent adsorbents (PhB, Ni_PhB, PhB_A, Ni_PhB_A, PhB_BM_A and
Ni_PhB_BM_A). Each curve essentially represents the equilibrium
isotherm of VB12 at 30 °C over an aqueous phase concentration
7.0
PhB_BM_A
Ni_PhB_BM_A
PhB_A
Ni_PhB_A
PhB
6.5 Ni_PhB
Final pH
6.0
5.5
5.5 6.0 6.5 7.0
Initial pH
Fig. 7. pH variation of solution before and after VB12 adsorption.
258 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260

300 1.0

250
Solid phase concentration, q (mg/g)

0.8
Ni_PhB_BM_A Initial Concentration
(mg/l)
200 PhB_BM_A
200

Cexit /C inlet
Ni_PhB_A 0.6
PhB_A 150
150 Ni_PhB 100
PhB 50
0.4
100

0.2
50

0 0.0
0 50 100 150 200 250 300 0 2000 4000 6000 8000 10000 12000
Aqueous phase concentration, Ce (ppm) Time (min)

Fig. 8. Adsorption isotherms of VB12 on phenolic precursor based adsorbents. Fig. 9. Column breakthrough curves for VB12 on Ni_PhB_BM_A.

range of 10–300 mg/l or ppm. As observed, the solid phase concen-

The quantity of the solute retained in the column was deter-
trations monotonically increased with increasing aqueous phase
mined by calculating the area above the breakthrough curves and
concentration and asymptotically reached a maximum value for
applying this to the mass balance equation [25]. The uptake capac-
each type of adsorbent. The maximum value corresponded to the
ity of the adsorbent was evaluated by dividing the mass of the ad-
saturation capacity of the adsorbent.
sorbed VB12 by the mass of the adsorbent. The VB12 uptakes
There are three salient observations. (1) Carbonized/activated
corresponding to the four concentrations were calculated to be
beads had larger adsorption capacities than their polymeric bead
148 mg/g, 210 mg/g, 231 mg/g, and 235 mg/g for 50, 100, 150,
counterparts. (2) The saturation adsorption capacities of Ni_PhB_B-
and 200 ppm, respectively. These numbers may be compared to
M_A and Ni_PhB_A are approximately 10% and 35% larger than their
the batch adsorption data of 168 mg/g, 225 mg/g, 260 mg/g, and
counterparts (without metals), PhB_BM_A and PhB_A, respectively.
280 mg/g, respectively, at the same concentrations. Therefore, the
On the other hand, the adsorption capacity of Ni_PhB is approxi-
difference between the adsorption capacity of the micro–nanoad-
mately three times larger than that of PhB. Therefore, the in situ
sorbent particles obtained under batch and flow conditions was a
addition of Ni to polymeric beads improves the adsorption capacity,
although slightly (10%) for nanoadsorbents. Based on the FT-IR
spectra discussed later in this paper, the improved adsorption per-
formance of Ni-incorporated materials is attributed to the interac-
tion of Ni with the functional groups of VB12, and (3) a maximum
adsorption capacity of approximately 300 mg/g was obtained for
Ni_PhB_BM_A, the micro–nanoparticle-based adsorbents which
were prepared by ball-milling of the polymeric beads followed by
carbonization and activation. Here, it is important to mention that
the equilibrium loadings of VB12 obtained for Ni_PhB_BM_A is much
higher than most of the reported literature data [3,24,33,34] and
comparable in a few cases [1,22,35]. The superior performance of
the micro–nanoparticle-based adsorbent (Ni_PhB_BM_A) is attrib-
uted to the combined effects of Ni as well as large number of active
sites available for adsorption, as reflected by its significantly large
BET area (1175 m2/g). In other words, there are advantages in
using Ni_PhB_BM_A as adsorbents because of relatively faster kinet- Fig. 10. Reaction scheme for the formation of Ni doped phenolic beads.
ics of uptake and large adsorption capacity for VB12.

4.4. Column study

PhB
PhB (adsorbed)
The purpose of the column experiment was to ascertain the per- Ni_PhB
formance of the specially designed column for the prepared micro– Ni_PhB (adsorbed)
Transmittance

nanoadsorbents (Ni_PhB_BM_A) in this study. During a typical flow

experiment, maldistribution or channeling of the liquid in the
packed bed adsorber can cause underutilization of the adsorbents.
In such a case, the adsorption performance may be evaluated by
examining the breakthrough response and determining the corre-
sponding uptake of the solutes.
Fig. 9 describes the breakthrough curves obtained for four dif-
ferent VB12 inlet concentrations (50 ppm, 100 ppm, 150 ppm, and
200 ppm). The amount of adsorbents, the liquid flow rate, and
the temperature were maintained constant for each case. The ob- 4000 3500 3000 2500 2000 1500 1000 500
served breakthrough response was consistent with the operating Wavenumber (c.m.)
conditions, with a decrease in the breakthrough and saturation
times with increasing inlet concentrations. Fig. 11. FT-IR spectra of synthesized adsorbents before and after VB12 adsorption.
 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260
Table 3
Assignment of FTIR bands.
Chemical group/vibration band Range (cm1)
mOH (OAH water + OAH intra- and inter-H bonds) + (mNH) 3200–3600
mCO (conj dNH) (amide I) 1620–1675
mNH (conj. dC@N) (amide II) 1520–1560
mPO3 950–980
1060–1065
maximum of 15%, and this is attributed to possibly small channel-
ing or entrainment of the flow in the packed bed column of the mi-
cro–nanoparticles under the flow conditions. We can thus
conclude that the bed configuration was effective in preventing
the entrainment of the micro–nanoparticles in the flowing fluid
and also provided a uniform flow in the packed bed. This is a sali-
ent feature of the column developed in this study. Further
improvement in the design of the column and its potential applica-
tions to bio-separation on relatively large scales are currently
underway and will be addressed in a future study.
Prior to the column tests on Ni_PhB_BM_A, a blank experiment
was carried out to ascertain the contribution of the combined poly-
meric beads and TFC membrane, used as support for the nanoad-
sorbents (Ni_PhB_BM_A), to the adsorption of VB12 during a
typical column experiment. The blank experiment was carried
out without the Ni_PhB_BM_A adsorbents. The other operating
conditions were kept unchanged. Total quantity of polymeric
beads (PhB) in the column was approximately 1.2 g. Total weight
of the TFC sheet was insignificant (less than 0.02 g). The break-
through data for the blank experiment (inlet concentra-
tion = 200 mg/l) showed insignificant adsorption of the solute,
with the solute-breakthrough immediately observed following
switching of the column’s inlet to the stream laden with VB12.
The results of the blank column-experiment are consistent with
the batch adsorption data depicted in Fig. 8. The saturation adsorp-
tion capacity of polymeric beads, PhB is 15 mg/g (approximately
20 times smaller) in comparison to 300 mg/g for the adsorbents,
Ni_PhB_BM_A. Also, the BET surface area of PhB is 13 m2/g
(approximately 100 times smaller) in comparison to 1228 m2/g
for Ni_PhB_BM_A, as shown in Table 1. Thus, it may be concluded
that the total uptake of VB12 calculated from the breakthrough
curves (Fig. 9) is predominantly due to the nanoadsorbents
(Ni_PhB_BM_A) used in the column. The contribution of the poly-
meric materials (combined beads and TFC sheet) to the total up-
take is negligible.
4.5. Proposed mechanism of VB12 removal by the adsorbents
Fig. 10 describes a tentative molecular structure of nickel doped
synthesized polymeric adsorbents in this study. A step growth
polymerization occurs between phenol and formaldehyde. Phenol
initially reacts with formaldehyde to form hydroxymethyl phenol.
The hydroxymethyl group subsequently reacts with either free
ortho or para site of the phenol molecule or another hydroxy-
methyl group. This is followed by the incorporation nickel as
shown in Fig. 10.
FT-IR spectra were taken for the various adsorbents (PhB and
Ni_PhB) before and after adsorption and the main bands were as-
signed (Fig. 11 and Table 3). The spectra were recorded on a Tensor
27 (Bruker, Germany) instrument over the wave number range of
4004000 cm1 using the attenuated total reflectance (ATR) with
a germanium (Ge) crystal. The spectra and the corresponding data
acquisition were automatically obtained using an interfaced com-
puter and a standard software package.
The interpretation of the spectra was found to be difficult
because of the superimposition of several signals in close wave
259
PhB (cm1) PhB_Ads (cm1) NiPhB (cm1) NiPhB_Ads (cm1)
3600 3675 3600 3677
– 1649 – 1650
1527 1540, 1516 1517 1514, 1543
– 980, 1060 – 975, 1064
number range. For example, a first broad and poorly resolved band
around 3600 cm1 may correspond to the contribution of OAH
stretching from water molecule or from intra- and intermolecular
hydrogen bonds or NAH stretching.
The band corresponding to the amide I C@O stretching mode of
the propionamide side chains of the corrin ring of VB12 was ob-
served for the adsorbed samples of PhB and Ni_PhB at 1649 cm1
and 1645 cm1, respectively. The slight difference observed in
the absorption bands may be attributed to the bonding of the car-
bonyl group with Ni. There are also new bands near 1540 cm1 in
the spectra of the VB12 saturated adsorbents. These additional
bands are attributed to the NAH amide II stretching vibrations of
the corrin ring of VB12[36]. The 1060 cm1 and 1064 cm1 bands
for the saturated samples of PhB and Ni_PhB, respectively are as-
signed to the PO3 degenerated stretching of the phosphate moiety
in VB12, whereas the bands over 975–979 cm1 are assigned to PO 3
symmetric stretching [37]. These two absorption bands at
1060 cm1 and 975 cm1 are sensitive to direct metal–phosphate
coordination. Further, the infrared spectra show that the bands
due to the phosphate groups in VB12 have not only moved from
998 cm1 in virgin VB12 to 975 cm1, the intensity of the band
has also increased for the Ni doped adsorbent indicating the direct
interactions between the phosphate groups of VB12 and Ni ions.
From the above discussions it may be concluded that Ni plays a
major role in the adsorption of VB12 and the groups such as amino,
hydroxyl, carbonyl and phosphate groups of VB12 are involved in
the adsorption.
5. Conclusions
Activated carbon spheres (ACSs) of approximately 0.8 mm and
micro–nanoparticles of approximately 200 nm, derived from nick-
el-doped phenolic beads, were successfully applied as adsorbents
for VB12 removal in aqueous solutions over the concentration range
of 5–300 mg/l. The adsorption loading of VB12 on micro–nanopar-
ticle adsorbents was determined to be as high as 300 mg/g, which
is comparable to or greater than that of the other adsorbents re-
ported in the literature. An especially designed column was devel-
oped in this study for applying micro–nanoparticles based
adsorbents to the aqueous flow without entrainment or channeling
in the packed column. The breakthrough analysis revealed that the
VB12 uptake in the bed was comparable to the equilibrium adsorp-
tion data obtained from the batch study. This indicates an effective
bed design allowing a uniform flow distribution and minimal par-
ticle entrainment and loss by flow. The simple methodology for
developing micro–nanoparticles based adsorbents and applying
them in the column under dynamic conditions has potential for a
wide range of bio-separation and bio-purification applications.
Acknowledgments
The authors acknowledge support from DST (New Delhi, India)
in the form of a research grant and through its DST Unit of Excel-
lence on Soft Nanofabrication. The first author also acknowledges
CSIR (New Delhi, India) for the financial research assistantship.
260 R. Saraswat et al. / Chemical Engineering Journal 197 (2012) 250–260
References
[1] Z. Guo, L.F. Wang, Z. Gao, W.W. Zhang, Equilibrium and kinetic studies on the
adsorption of VB12 onto CMK-3, Chin. Chem. Lett. 18 (2007) 233–236.
[2] S.P. Stabler, R.H. Allen, L.P. Fried, M. Pahor, S.J. Kittner, Racial differences in
prevalence of cobalamin and folate deficiencies in disabled elderly women,
Am. J. Clin. Nutr. 70 (1999) 911–919.
[3] A.M. Ramos, M. Otero, A. Rodrigues, Recovery of Vitamin B12 and
cephalosporin-C from aqueous solutions by adsorption on non-ionic
polymeric adsorbents, Sep. Purif. Technol. 38 (2004) 85–98.
[4] P.S. Porter, M.J. Semmens, A review of virus and protein sorption on ion
exchange resins, Environ. Int. 3 (1980) 311–319.
[5] N.M. Khose, B.D.J. Dasare, Studies of the uptake of vitamin B12 and theophylline
by phenol-based porous condensate cation exchangers, J. Chem. Technol.
Biotechnol. 31 (1) (1981) 219–225.
[6] D. Sacco, E. Dellacherie, Liquid chromatography of cephalosporin C on
substituted polystyrene resins, Anal. Chem. 56 (8) (1984) 1521–15244.
[7] M. Hicketier, K. Buchholz, Investigations on cephalosporin C adsorption
kinetics and equilibria, Appl. Microbial. Biotechnol. 32 (1990) 680–685.
[8] I.M. Huxaham, B. Rowatt, Die Makromolekulare Chemie, Characterisation of
macroporous polystyrene and its relationship to the adsorption of
Cephalosporin C, Die Makromol. Chem. 192 (2003) 1695–1703.
[9] M.V. Chaubal, G.F. Payne, C.H. Reynolds, R.L. Albright, Equilibria for the
adsorption of antibiotics onto neutral polymeric sorbents: experimental and
modeling studies, Biotechnol. Bioeng. 47 (2) (2005) 215–226.
[10] I. Bravo-Osuna, C. Ferrero, Jimenez-Castellanos, Water sorption–desorption
behaviour of methyl methacrylate–starch copolymers: effect of hydrophobic
graft and drying method, Eur. J. Pharm. Biopharm. 59 (2005) 537–548.
[11] F. Salto, J.G. Prieto, Interactions of cephalosorins and penicillins with nonpolar
macroporous styrenedivinylbenzene copolymers, J. Pharm. Sci. 7 (9) (1981)
994–998.
[12] W.O. Khunjar, G.L. Nancy, Sorption of carbamazepine, 17a-ethinylestradiol,
iopromide and trimethoprim to biomass involves interactions with exocellular
polymeric substances, Chemosphere 82 (2011) 917–922.
[13] J.R. Domínguez, T. Gonzalez, P. Palo, E.M. Cuerda-Correa, Removal of common
pharmaceuticals present in surface waters by Amberlite XAD-7 acrylic-ester-
resin: Influence of pH and presence of other drugs, Desalination 269 (2011)
231–238.
[14] B. Pan, W. Zhang, Lu Lv, Q. Zhang, S. Zheng, Development of polymeric and
polymer-based hybrid adsorbents for pollutants removal from waters, Chem.
Eng. J. 151 (2009) 19–29.
[15] L.D. Santa Maria, M.C.A.M. Leite, M.A.S. Costa, J.M.S. Ribeiro, L.F. Senna, M.R.
Silva, Preparation of composite materials containing iron in a cross-linked
resin host based on styrene and divinylbenzene, Eur. Polym. J. 39 (2003) 843–
846.
[16] S. Rana, P. White, M. Bradley, Synthesis of magnetic beads for solid phase
synthesis and reaction scavenging, Tetrahedron Lett. 40 (1999) 8137–8140.
[17] X.B. Ding, Z.H. Sun, G.X. Wan, Y.Y. Jiang, Preparation of thermosensitive
magnetic particles by dispersion polymerization, React. Funct. Polym. 38
(1998) 11–15.
[18] M. Dutta, N.N. Dutta, K.G. Bhattacharya, Aqueous phase adsorption of certain
beta-lactam antibiotics onto polymeric resins and activated carbon, Sep. Purif.
Technol. 16 (1999) 213–224.
[19] J.B. Yang, L.-C. Ling, L. Liu, F.-Y. Kang, Z.-H. Huang, H. Wu, Preparation and
properties of phenolic resin-based activated carbon spheres with controlled
pore size distribution, Carbon 40 (2002) 911–916.
[20] S.R. Sandeman, C.A. Howella, G.J. Phillipsa, A.W. Lloyda, J.G. Daviesa, S.V.
Mikhalovskya, S.R. Tennisonb, A.W. Rawlinsonb, O.P. Kozynchenkob, H.L.H.
Owenc, J.D.S. Gaylorc, J.J. Rousec, J.M. Courtneyc, Assessing the in vitro
biocompatibility of a novel carbon device for the treatment of sepsis,
Biomaterials 267 (2005) 124–7131.
[21] J.-B. Yushin, L.-C. Ling, L. Liu, F.-Y. Kang, Z.-H. Huang, H. Wu, Preparation and
properties of phenolic resin-based activated carbon spheres with controlled
pore size distribution, Carbon 40 (2006) 911–916.
[22] Z. Guo, G. Zhu, D. Zhang, G. Tian, Y. Chen, W. Zhang, S. Qiu, Adsorption of
vitamin, B12 on ordered mesoporous carbons coated with PMMA, Carbon 43
(2005) 2344–2351.
[23] F. Salvador, M.J. Sanchez, Martin A. Sanchez, C. Sanchez-Jimenez, Preparation
of active carbon from charcoal by activation with supercritical water.
Eurocarbon, in: 1st World Conference on Carbon, German Carbon Group,
Berlin, 2000, pp. 669–670.
[24] (a) C. Ye, Q.-M. Gong, F.-P. Lu, J. Liang, Adsorption of uraemic toxins on carbon
nanotubes, Sep. Purif. Technol. 58 (2007) 2–6;
(b) C. Ye, Q.-M. Gong, F-P. Lu, J. Liang, Preparation of carbon nanotubes/
phenolic-resin -derived activated carbon spheres for the removal of middle
molecular weight toxins, Sep. Purif. Technol. 61 (2008) 9–14.
[25] A. Sharma, N. Verma, A. Sharma, D. Deva, N. Sankararamakrishnan, Iron doped
phenolic resin based activated carbon micro and nanoparticles by milling:
synthesis, characterization and application in arsenic removal, Chem. Eng. Sci.
65 (2004) 3591–3601.
[26] V. Kumar, N. Telreja, D. Deva, N. Sankararamakrishnan, A. Sharma, N. Verma,
Development of bi-metal doped micro- and nano multi-functional polymeric
adsorbents for the removal of fluoride and arsenic(V) from wastewater,
Desalination 282 (2011) 27–38.
[27] J. Yang, T. Zhout, L. Song, Agglomerating vibro-fluidization behavior of nano-
particles, Adv. Powder Technol. 20 (2009) 158–163.
[28] A. Ajbar, Y. Bakhbakhi, S. Ali, M. Asif, Fluidization of nano-powders: effect of
sound vibration and pre-mixing with group A particles, Powder Technol. 206
(2011) 327–337.
[29] L. Zhou, R. Diao, T. Zhou, H. Wang, H. Kage, Y. Mawatari, Behavior of magnetic
Fe3O4 nano-particles in magnetically assisted gas-fluidized beds, Adv. Powder
Technol. 22 (2011) 427–432.
[30] S.F. Mossa Hosseini, B. Ataie-Ashtiani, M. Kholghi, Nitrate reduction by nano-
Fe/Cu particles in packed column, Desalination 276 (2011) 214–221.
[31] L. Zhang, X. Guo, H. Li, Z. Yuan, X. Liu, T. Xu, Separation of trace amounts of Ga
and Ge in aqueous solution using nano-particles micro-column, Talanta 85
(2011) 2463–2469.
[32] F. Rodrıguez-Reinoso, A.C. Pastor, H. Marsh, Preparation of activated carbon
cloths from viscous rayon. Part II: physical activation processes, Carbon 38
(2000) 397–406.
[33] S.A. Hashsham, D.L. Freedman, Adsorption of Vitamin B12 to alumina,
kaolinite, sand and sandy oil, Water Res. 37 (2003) 3189–3193.
[34] C. Liu, S. Wang, Z. Rong, X. Wang, G. Gu, W. Sun, Synthesis of structurally stable
MCM-48 using mixed surfactants as co-template and adsorption of vitamin
B12 on the mesoporous MCM-48, J. Non-Cryst. Solids 356 (2010) 1246–1251.
[35] S. Xu, G. Qiao, H. Wang, D. Li, T. Lu, Preparation of mesoporous carbon derived
from mixtures of phenol-formaldehyde resin and ethylene glycol, Mater. Lett.
62 (2008) 3716–3718.
[36] K.S. Taraszka, E. Chen, T. Metzger, M.R. Chance, Identification of structural
markers for vitamin B12 and other corrinoid derivatives in solution using FTIR
spectroscopy, Biochemistry 30 (1991) 1222–1227.
[37] M. Tsuboi, Infrared and Raman Spectroscopy in Basic Principles in Nucleic Acid
Chemistry, vol. 1, Academic Press, UK, 1974. p. 399.