Microporous and Mesoporous Materials 132 (2010) 470–479

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

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

Metal (M = Co2+, Ni2+, and Cu2+) grafted mesoporous SBA-15: Effect

of transition metal incorporation and pH conditions on the adsorption
of Naproxen from water
Sindia M. Rivera-Jiménez, Sandra Méndez-González, Arturo Hernández-Maldonado *
Department of Chemical Engineering, University of Puerto Rico-Mayagüez Campus, Mayagüez 00681-9000, Puerto Rico

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

Article history: Incorporation of transition metals onto the surface of mesoporous silica via grafting techniques could
Received 8 January 2010 become a potential strategy for the bottom-up design of sorbents for the removal of pharmaceutical
Received in revised form 25 March 2010 drugs and other emerging contaminants from water sources. Here we report the incorporation of Co2+,
Accepted 25 March 2010
Ni2+, or Cu2+ onto the surface of mesoporous SBA-15 via an amino-organic grafting method and the
Available online 31 March 2010
screening of the materials as sorbents for the removal of Naproxen from water. The functionalization pro-
cedure involved the reaction of 3-aminopropyltriethoxysilane with OH groups available on SBA-15 sur-
Keywords:
faces, followed by the anchoring of the desired metal. Removal of traces of sulfates from the metal
Mesoporous silica
Grafting
sources and further metal anchoring were achieved after post-synthesis treatment at alkaline conditions.
Transition metals The resulting metal-modified materials were fully characterized using X-ray powder diffraction, N2
Adsorption adsorption/desorption isotherms gathered at 196 °C, SEM/EDX, FT-IR, and TGA. Aqueous phase single
PPCPs point adsorption experiments showed that Cu-based materials are good sorbents for the uptake of
low-concentration Naproxen at 25 °C and approaching alkaline conditions. The materials uptake capac-
ities decreased as follows: CuNH2_g_SBA-15 > NiNH2_g_SBA-15 > CoNH2_g_SBA-15. In general, the
adsorption results correlate well with the observed structural and textural properties of the materials,
suggesting also that the metal amino-organic loading is highly dependent on the nature of grafting pro-
cess and choice of metal.
Ó 2010 Elsevier Inc. All rights reserved.

1. Introduction potable water and aquatic systems is an important goal for world-
wide government agencies and researchers [3,16].
Recent reports have recognized that the occurrence, fate, and Most PPCPs are polar compounds with acidic or basic functional-
treatment of Pharmaceutical and Personal Care Products (PPCPs) ities that occur at low-concentration levels in water sources. These
in the environment are of increasing global concern [1–5]. These characteristics generate unique challenges for their removal using
so called ‘‘emerging contaminants” were first identified in the early traditional water treatment methods [17]. Various sorbents have
1990s and include a wide rage of chemicals such as analgesics (Na- been proposed for the removal of a selected list of PPCPs at low con-
proxen, acetaminophen, and ibuprofen), antibiotics (ciproflaxin, centrations. Examples of these include activated carbon, membranes,
and triclosan), antihyperlipidemics (clofibric acid), and other com- and polymers [6,18–22]. In general, the corresponding uptake capac-
pounds (bisphenol, salicylic acid, and carbamazepine) [6–8]. Dur- ities of many of these materials showed to be highly dependent on pH
ing decades, they have shown to be persistent in the conditions and specific interactions such as van der Waals, electro-
environment probably due to long-shelf life formulations associ- static, and dipole–dipole. Another study showed that the adsorption
ated with product demand of a continuously growing population performance of silica, c-alumina and Parapak-P (non-polar organic
[9–11]. Moreover, traditional water treatment methods are not medium) on selected PPCPs was also highly dependent on pH condi-
suitable for the removal of PPCPs resulting in surface water con- tions, ionic level of the pharmaceutical, and the nature of the surface
centrations in the parts-per-trillion (ppt) range [9,12–15]. charge of the sorbents [23]. To achieve better selectivity at the pres-
Although the ultimate impact of PPCPs on population health is still ent and near future concentration levels it is critical to start consider-
not fully understood, the removal of these contaminants from our ing the bottom-up design of surfaces with better specific interactions,
tailored to remove specific families of PPCPs.
Mesoporous silica are good candidates for the design of PPCP
* Corresponding author. Tel.: +1 787 832 4040x3748; fax: +1 787 834 3655. selective sorbents due to their large surface areas, tunable pore
E-mail address: [email protected] (A. Hernández-Maldonado). structure and well understood surface chemistry [24,25]. Since

1387-1811/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.micromeso.2010.03.029
 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479 471

their discovery in 1992 [26] mesoporous silicas from the M41S with copious amounts of water. The recovered solid cake was dried
family have demonstrated their potential in applications like heavy in a forced convection oven overnight at 60 °C. The surfactant tem-
metal and organic contaminants removal, catalysis, chromato- plate molecules inside the pores of the mesoporous material were
graphic separation, and controlled drug delivery [27–34]. A recent removed via calcination with air at 500 °C for 6 h. For nomencla-
report documents the use of nickel grafted on mesoporous support ture purposes, the resulting material was simply denominated
as a potential sorbent for the removal of Naproxen from water at SBA-15.
ppm concentration levels [35]. However, it was concluded that
supports with better stability should be considered in order to in- 2.3. Preparation of metal modified mesoporous silica
crease the amount of metal incorporated onto the surface without
compromising the structural and porosimetry properties. In addi- The modification of SBA-15 was done in a step-wise fashion
tion, based on ab initio calculations, the report suggested that the using grafting techniques as showed in Scheme 1 [39]. First, for
adsorption process could be described as a complexation mecha- the chemical incorporation of amino-organic moieties, 2 g of
nism between the metal center and the Naproxen molecule. SBA-15 (BET SA  800 m2/g) were pretreated at 150 °C for 2 h in a
One possible way of reducing the grafting effects on the overall nitrogen atmosphere to remove physisorbed water. After the solid
materials textural properties of mesoporous silica is by employing was brought to room temperature under nitrogen, it was sus-
substrates that display dense pore walls. SBA-15 mesoporous silica pended in 400 ml of dry toluene and stirred for approximately
are known to have thick pore walls (3.1–6.4 nm) and considerable 30 min. An excess amount of 3-aminopropyltriethoxysilane in
pore diameters (4.6–30 nm) [36], characteristics that are essential 80 ml toluene was added drop-wise to the slurry and stirred under
to facilitate the grafting while providing the required pore volume reflux at 100 °C for 24 h. The material was filtered, washed with
for the removal of large sorbates. It should be mentioned that Bui dry toluene followed by ethanol, and Soxhlet-extracted using a
and Choi have recently reported the use of unmodified mesoporous mixture of 100 ml of diethyl ether and 100 ml dichloromethane
SBA-15 for the removal of a set of PPCPs [37]. Although they for 24 h. Finally, for the immobilization of metal complexes 1.5 g
showed that the materials performance was greatly dependent of amino-organic functionalized silica was mixed with a 0.02 M
on pH, no attempt to modify the solid surface chemistry was made. alcoholic solution of the corresponding M2+ sulfate (M = Co2+,
In this report, we present a detailed account on the modification Ni2+, and Cu2+) for 18 h at 60 °C also under nitrogen atmosphere.
of SBA-15 mesoporous supports with cobalt(II), nickel(II), and cop- The mixture was filtered, washed with copious amounts of ethanol
per(II) amine complexes, respectively. The materials chemical and and then Soxhlet-extracted with ethanol for 12 h. The materials
textural properties have been thoroughly evaluated and carefully were dried overnight in a forced convection oven at 60 °C.
considered during the analysis of single point adsorption tests for All metal modified materials were pretreated before character-
the removal of Naproxen from water. We also report the effect of ization and adsorption experiments to remove traces of sulfates
solution pH on the Naproxen uptake as well as its use as plausible and to promote final anchoring of the metal to the amino-organic
sorbent regeneration strategy. moieties. This pretreatment was achieved by washing 1 g of the me-
tal-modified materials with 500 ml of a NaOH solution (pH 12). The
materials were let to dry at 60 °C for at least 1 h and were denomi-
2. Experimental section
nated CoNH2_g_SBA-15, NiNH2_g_SBA-15, and CuNH2_g_SBA-15,
respectively. It is important to note that each metal modified mate-
2.1. Reagent and materials
rial showed a characteristic color depending on the type of metal
incorporated. It is common for metal ions of d-block family to form
The following reagents were used for the synthesis and surface
complexes that are associated with a change in color and sometimes
modification of SBA-15: Pluronic P123 Surfactant (O-BASF Cortra-
a change in the intensity of color [40]. Table 1 summarizes the color
tin), hydrochloric acid (HCl, 37 wt%, ACS Reagent, Sigma–Aldrich),
changes of the metal-modified materials and the suggested struc-
tetraethyl orthosilicate (TEOS 98 wt%, Sigma–Aldrich), toluene
tured based on the color changes after pretreatment.
(anhydrous, 99.8%, Sigma–Aldrich), ethanol (denaturized Sigma–
Aldrich), cobalt(II) sulfate heptahydrate (99%, Sigma–Aldrich),
2.4. Characterization
nickel(II) sulfate hexahydrate (Puris Reagent ACS, Riedel-deHaën),
copper(II) sulfate pentahydrate (98%, ACS reagent, Sigma–Aldrich),
X-ray powder diffraction patterns were obtained using a Rigaku
dimethyldichlorosilane (DMDCS, 99%, Sigma–Aldrich), methanol
Ultima III diffractometer with Cu ja radiation (k = 1.5418 Å) and
(spectrophotometric grade, 99%, Sigma). In addition, sodium
operating at 40 kV and 44 mA. All XRD patterns were collected in
hydroxide (Sigma ultra, 98%, Sigma–Aldrich) and Naproxen (98%,
the 2h range between 0.5° and 3° with a scanning rate of 0.01°/
Sigma–Aldrich) were used for the adsorption experiments. Prior
min. The textural properties of the samples were measured via
to use, all water was thoroughly distilled and deionized.
nitrogen adsorption isotherms at 196 °C (77 K) using a Micromer-
itics ASAP 2020 volumetric porosimetry test instrument equipped
2.2. Synthesis of pure siliceous mesoporous SBA-15 with turbo-molecular drag pumps. Prior to each adsorption mea-
surement, the SBA-15 and grafted samples were each degassed un-
Pure siliceous SBA-15 was synthesized using a scaled up version der vacuum for 5 h at 200 and 100 °C, respectively. The total
of the recipe previously reported by Stucky and co-workers surface area St, external surface area Sext, and primary mesopore
[36,38]. Relative large amounts of the SBA-15 were necessary for volume Vp of the samples were calculated using the as-plot method
the modification, characterization and adsorption routines. In a using a macroporous silica gel LiChrospher Si-1000 as a Ref. [41].
typical synthesis, 8.0 g of Pluronic P123 were dissolved at 35 °C Pore size distributions (PSD) were calculated from the desorption
in 60 ml of water and 240 g of a 2 M HCl solution under moderate branch of the N2 isotherm data using the Barret–Joyner–Halenda
stirring. After the polymer was dissolved, 18.2 ml of TEOS were (BJH) method applying the Kruk–Jaroniec–Sayari correction [42].
added drop-wise into the stirring solution and left to stir at 35 °C Images of the materials particles were obtained using a JEOL-
for 20 h. The resulting slurry mixture was then transferred into a JSM-6930 scanning electron microscope (SEM) operating with a
Teflon lined stainless-steel reactor and then sealed and heated to secondary electron detector and an accelerating voltage of 15 kV.
100 °C for 24 h under static and autogenous pressure conditions. Energy Dispersive X-ray Analysis (EDX) was used to determine me-
The final solid material was recovered by filtration and washed tal atomic ratios with respect to silica (gSi/gmetal) and nitrogen (gN/
472 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479

Scheme 1. Step-wise modification of SBA-15 surface using grafting technique.

gmetal). The experiments were performed using the aforementioned uptake experiment was performed by combining a predetermined
SEM unit at an accelerating voltage of 10 kV and an energy disper- amount of the adsorbent with 15 ml solution containing an initial
sive X-ray (EDX) spectrometer. Calculations to analyze surface concentration of 14 ppm Naproxen in 25 ml Teflon bottles. All
composition were done using the standardless quantification. Re- glassware used was previously deactivated using a 10% (v/v) solu-
sults are reported based on the average of six different randomly tion of DMDCS in toluene, washed twice with methanol, and dried
selected areas to ensure that the data collected was representative at room temperature for at least 1 h. The pH of the initial solution
of the bulk sample. used for the adsorption experiments was adjusted to 2, 7 or13
FT-IR spectra were obtained in a Nicolet 6700 FT-IR instrument using 1 M HCL and 1 M NaOH solutions without further adjust-
loaded with a Praying Matins attachment operated at ambient con- ment after adding the sorbents. The mixtures were then placed
ditions. Powder samples were analyzed at 4 cm1 resolution and in a constant temperature oscillation shaker (operated at
averaged over 400 scans in the absorption band range of 4000– 250 rpm) and allowed to equilibrate at room temperature for ca.
400 cm1. All spectra were corrected for presence of humidity 24 h. At the end of the adsorption experiment, the final slurry
and carbon dioxide. Weigh loss profiles were obtained using a high was centrifuged at 3500 rpm and for 5 min. Triplicate measure-
resolution TA-Q500 thermogravimetric analyzer to quantify the ments were taken of the remaining liquid phase to determine the
amount of amino-organic moieties present on the metal final concentration, which was measured using a Shimadzu UV-
(M = Cu2+, Ni2+, and Co2+) grafted mesoporous SBA-15. These 2401 PC UV/Vis spectrophotometer at the corresponding charac-
experimental results were compared to theoretical values obtained teristic wavelength of Naproxen. The adsorbed amount of Naprox-
assuming monolayer formation over the support surface. TGA mea- en was calculated by concentration difference using the equation:
surements were performed under constant flow of air at 60 ml/min
ðC i  CÞ  V i
using platinum sample holders and a heating rate of 5 °C/min from Q¼  103
W s  MW Napro
25 to 800 °C. Air was pretreated using moisture and hydrocarbons
traps to avoid sample contamination. where Q (lmol/g) is the adsorbed amount of Naproxen, Ci (mg/l) is
the initial concentration of Naproxen used, C (mg/l) is the remaining
2.5. Adsorption experiments concentration after ca. 24 h, Vi (l) is the liquid volume of the batch,
Ws (g) is the mass of the adsorbent added, and MWNapro (mg/mmol)
Single point adsorption experiments were carried out using a is the molecular weight of Naproxen. Whenever possible, fractional
batch equilibrium technique described elsewhere [35]. A typical uptake data was gathered and used to estimate diffusion related
 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479 473

Table 1
Suggested metal coordination structure based on observed color changes.

Sample Observed colora Suggested structure

Before pretreatment After pretreatment After adsorption experiment
SBA-15 White White White –
CoNH2_g_SBA-15 Mix of yellowish brow and light blue No noticeable changes No noticeable changes

Yellowish-brown

Light blue
NiNH2_g_SBA-15 Light green Intense light green No noticeable changes

CuNH2_g_SBA-15 Light blue Intense light blue Royal blue

a
Color observed under white fluoresce light and white background.

information. Transient uptake data was estimated using equation fraction peak intensities was observed after the incorporation of
shown above, with C measured as a function of time. the metals. Previous studies suggest that this behavior could be
attributed to a reduction in scattering power of the mesoporous
3. Results and discussion walls due to the incorporation of amino-organic moieties inside
the SBA-15 pores [28,44]. Table 2 shows d-spacing values for the
3.1. Structure, textural, and morphological properties of metal grafted SBA-15 and metal-modified materials. In general, the incorpora-
mesoporous SBA-15 tion of the metal and amino-organic moieties does not seem to
considerably affect the interplanar spacing, with the exception of
Three transition metals were used for the modification of SBA- only a minor shift observed for the Ni-modified materials probably
15 using the grafting technique. Fig. 1 compares the low-angle XRD suggesting a pore size reduction.
patters of SBA-15 samples before and after the incorporation of the The nitrogen adsorption–desorption isotherms for all samples
metal. Unmodified SBA-15 exhibits three well-defined X-ray dif- are gathered in Fig. 2 and a summary of structural parameters
fraction peaks, which can be indexed to the (1 0 0), (1 1 0), and are presented in Table 2. Most of the observed isotherms are of
(2 0 0) planes. These characteristic peaks suggest a p6mm symme- type IV with H2 hysteresis loops, which is typical of SBA-15 mate-
try, typical of uniform mesopore structure with hexagonal long- rials [38,45,46]. The adsorption isotherm data for SBA-15 shows a
range order [36,43]. The corresponding d100-spacing for the relatively sharp increase at around P/Po = 0.77, suggesting capillary
unmodified SBA-15 sample is 9.2 nm, which agrees well with the condensation of nitrogen inside a uniform mesopore structure and
value reported elsewhere [36]. Three characteristic diffraction in agreement with the low-angle XRD pattern observations. Struc-
peaks are still present for the grafted NiNH2_g_SBA-15 and Cu tural properties such as pore uniformity, high surface area
NH2_g_SBA-15 samples, suggesting that the long-range hexagonal (870 m2/g), pore volume (1.06 cm3/g), and interplanar spacing
structure was preserved. However, an overall decrease in the dif- associated to the pore channels (9.2 nm) demonstrate that a neat
474 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479

uted to the partial blocking of the pores [28,44]. In spite of the ob-
served variations in the textural properties, the changes in the
overall shape of the adsorption/desorption isotherms and the pres-
ence of the inflection points in all samples suggest that the metal
and amino-organic groups were incorporated inside the pores
and that most of the sorbents original structure order is still re-
tained after modification [47]. These observations correlate well
with the general XRD findings.
Fig. 3 shows the SEM micrographs for all the samples at two dif-
ferent magnifications. In agreement with the results reported else-
where, SBA-15 has the typical wheat-like morphology and consists
of aggregates of uniform rope-like particles [47] with a mean size
of ca. 1 lm. After the metal grafting (Fig. 3b–d), it can be observed
that the macroscopic aggregates do resemble the typical wheat-
like morphology, but with larger external area when compared to
that of the unmodified SBA-15 sample. In addition, an overall de-
crease in the mean particle size is apparent at higher magnification
possibly associated to the increase in external surface areas pre-
sented in Fig. 2. SEM images also show that both cobalt (Fig. 3b)
and copper (Fig. 3d) modified materials still resemble rope-like
particles. However, as shown in Fig. 3c, the NiNH2_g_SBA-15 sam-
ple seems to be affected the most.

Fig. 1. Powder XRD patterns of (a) SBA-15; (b) CoNH2_g_SBA-15; (c) NiNH2_g_SBA- 3.2. Composition of the metal grafted mesoporous SBA-15
15; and (d) Cu NH2_g_SBA-15. The data were shifted vertically for clarity.
EDX results are summarized in Table 3 and where used to ana-
lyze the composition for SBA-15 and the variants in a semi-quan-
Table 2 titative fashion. The metal loading on the siliceous SBA-15 was
Textural and structural properties obtained from N2 adsorption–desorption at analyzed using the atomic ratio of the siliceous support over that
196 °C and X-ray diffraction data. of the metal. The atomic ratios for CoNH2_g_SBA-15, NiNH2_g_S-
Sample St (m2/g) Vp (cm3/g) Dp (nm) d100 (nm) BA-15, and CuNH2_g_SBA-15 were 6.4, 6.3, and 3.5, respectively.
SBA-15 870 1.06 7.9 9.2
In general, all modified materials appear to have being loaded with
NH2_g_SBA-15 486 0.63 7.1 – their corresponding metal, being CuNH2_g_SBA-15 the one with
CoNH2_g_SBA-15 441 0.31 4.6/8.1 10.2 the highest loading. To have insight on the possible stoichiometry
NiNH2_g_SBA-15 383 0.29 7.0 8.8 between amino-organic groups and the metal loading, a molar ra-
CuNH2_g_SBA-15 283 0.34 7.6 10.2
tio between atomic nitrogen and metal loaded was calculated.
These ratios were 5.2, 5.7, and 4.9 for CoNH2_g_SBA-15,
NiNH2_g_SBA-15, and CuNH2_g_SBA-15, respectively.
The formation of metal complexes with amino-organic moieties
mesoporous SBA-15 was successfully obtained using a scale-up is not necessary following a 1:1 stoichiometry. For instance, the re-
synthesis version of the one reported by Stucky and co-workers. sults shown in Table 3 indicate that each metal center in the
After modification of SBA-15 via the grafting method presented NiNH2_g_SBA-15 sample is probably forming bonds with approxi-
above, the isotherm inflections associated to hysteresis (see Fig. 2; mately six amino groups concurrently. This need of an additional
data for NH2_g_SBA-15 not shown for clarity) shifted to lower rel- amino-organic moiety may result in a partially or fully blocked
ative pressures and the final volume of nitrogen adsorbed de- pore therefore producing a decrease in pore volume (see Table 2).
creased. In addition, the total surface area estimated with the as- Similar observations were made by Rivera-Jimenez and Hernan-
method (Table 2) decreased as follows NH2_g_SBA-15 > CoN- dez-Maldonado during the grafting of MCM-41 with a analogous
H2_g_SBA-15 > NiNH2_g_SBA-15 > CuNH2_g_SBA-15. However, a nickel-organic moiety [35].
comparison of the PSD curves for SBA-15 and the metal modified
samples (Fig. 2) reveals that the grafting process does not have 3.3. Characterization of the inorganic and organic moieties
much influence on the average pore size of the resulting material.
Fig. 2 also shows the distribution of external and pore areas. The Infrared spectroscopy was used to characterize the support and
percentage of pore surface area available normalized by the total the inorganic and organic moieties present in the solid samples
area of the corresponding material follows the order SBA- after modification. Fig. 4 shows FT-IR spectra for all the samples.
15 > CuNH2_g_SBA-15 > CoNH2_g_SBA-15 > NiNH2_g_SBA-15. This For SBA-15, the bands at 1078, 821, and 968 cm1 may be attrib-
observation may be attributed to complete or partial blocking of uted to the asymmetric Si–O–Si vibrations, symmetric vibrations
some of the pores due to amino-organic moieties that may have on the surface, and Si–OH vibrations, respectively. In addition
been hydrolyzed among them instead of forming a single mono- and the broad band in the 3700–3200 cm1 range suggests the
layer on the surface [28,44]. A possible schematic representation presence of silanols associated with adsorbed water. After grafting,
of the pore blockage is showed in Scheme 2. the intensity of the band associated with the single silanol groups
Decrease in textural properties is more evident in the at 975 cm1 and 3740 decreased and shifted to lower wavelength
NiNH2_g_SBA-15 samples (Fig. 2 and Table 2) and this correlates values suggesting the anchoring of the amino-organic moieties.
well with the interplanar changes observed during XRD analysis, In addition, after modification, the FT-IR data shows weak bands
which denotes some structural collapsing. Meanwhile, the ob- at the 1550–1750 cm1 range that may be attributed to the differ-
served CoNH2_g_SBA-15 sample porosimetry data yields a bimodal ent NH2 vibrational modes [27,48]. As reported before, the new
PSD with a minimal pore size of 4.6 nm, which is possibly attrib- bands that appeared between 3000–2800 cm1 and near
 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479 475

Fig. 2. Left: Nitrogen adsorption/desorption equilibrium isotherms at 196 °C and inset with PSD curves. Right: External and pore surface area distribution with values given
in m2/g.

in the metal sites. Such hypothesis comes from comparison with

the TGA data obtained for SBA-15, in which the peak water desorp-
tion temperature was nearly 20° lower. The two other weight loss
regions observed for the CoNH2_g_SBA-15 samples are attributed
to the decomposition of the amino-organic groups from the pore
surface and the loss of OH groups. As shown in Table 3, the calcu-
lated loadings for the amino-organic moieties based on TGA were
between 1.96 and 2.50 mmol/g. These overall values do not match
with the calculated theoretical monolayer value of 4 mmol/g, indi-
cating this, among other things, that the substrate surface was
probably lacking OH groups. Grafting of the moieties is highly
dependant on the OH concentration on SBA-15 surface and it has
Scheme 2. Representation of complete and partial pore blockage.
been reported before that typical calcined SBA-15 materials have
OH concentrations of ca. 1.2 lmol/m2 [50]. Assuming that such va-
lue is also representative of our SBA-15 sample (SA = 870 m2/g),
1150 cm1 can be assigned to both symmetric and asymmetric
then the expected net amount of OH groups is in the order of ca.
stretching of CH3 and CH2 groups [49]. The effect on the incorpora-
1 mmol/g. Such small surface OH concentration could explain the
tion of the metal groups can be observed by the shift to lower
observed same order, low amino-organic loadings, resulting also
wavenumber values of the bands assigned to amine groups. These
in few anchored and loosely bound species.
results once again confirm the modification of SBA-15 by amino-
organic groups and metals.
The TGA weight loss profiles for SBA-15 and the metal-modified 3.4. Sorbents single-component adsorption performance
materials are shown in Fig. 5. This technique provides information
about thermal stability, weakly adsorbed water, and the amount of Single point adsorption experiments were used to understand
amino-organic immobilized on the surface. For SBA-15, a loss in the effect of metal incorporation and pH conditions on the adsorp-
mass is observed from room temperature to 120 °C corresponding tion of Naproxen at 25 °C from water using an initial concentration
to the loss of weakly physisorbed water (11%). Additional losses of of 14 ppm. It is evident from Fig. 6 that the adsorption capacities
3% (between 120 and 600 °C) and 1% (between 600 and 800 °C) are influenced significantly by the pH of the initial solution and
may be attributed to small traces of the organic template used dur- the type of metal incorporated. For SBA-15, no adsorption was ob-
ing the synthesis procedure and the loss of OH groups due to the served at high pH conditions and the obtained adsorption capaci-
formation of siloxane bridges, respectively. Our SBA-15 materials ties are 0.45 and 0.69 lmol/g for low and neutral pH,
were calcined at 500 °C, which is well below the temperature range respectively. These adsorption capacities can be attributed to the
required for loss of hydroxyl groups and therefore minimizing loss presence of acid sites on SBA-15 due to the surface OH groups [37].
of anchorage points during grafting. For the metal-modified materials, low and neutral pH condi-
The TGA profiles of the metal-modified materials exhibited par- tions results in smaller adsorption capacities towards Naproxen
ticular weigh loss characteristics. For instance, the CoNH2_g_SBA- when compared to the results obtained for SBA-15. However, a
15 profile shows a sharp weight loss of 7% at ca. 137 °C probably general trend can be observed for pH 13 tests where adsorption
attributed to the presence of interatomic water that is complexed performance follow the order Cu NH2_g_SBA-15 > Ni NH2_g_SBA-
476 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479

Fig. 3. SEM images of (a) SBA-15; (b) CoNH2_g_SBA-15; (c) NiNH2_g_SBA-15; and (d) CuNH2_g_SBA-15 materials at different magnifications.

Table 3
tron donor site on Naproxen. It is suggested here that a metal com-
Compositional analysis obtained from TGA and EDX.
plex is created by the donation of available p-electrons from the
Sample NH2 loading (mmol/g) gSi/gmetal gN/gmetal donor site to the electron acceptor r-orbital of the metal ion.
SBA-15 – – – Depending on the strength of the complex, the metal could back-
CoNH2_g_SBA-15 2.12 6.4 5.2 donate d-orbital atoms to Naproxen that is consequently electron
NiNH2_g_SBA-15 1.96 6.3 5.7 deficient due to the initial donation. The probability of back-dona-
CuNH2_g_SBA-15 2.50 3.5 4.9
tion will increase with the higher available number of electrons on
the d-orbital of the metal ions. In other words, following the order
Cu2+ > Ni2+ > Co2+, which matches well with the observed uptake
15 > Co NH2_g_SBA-15. At pH values higher than the pKa of polar data. Moreover, the maximum Naproxen uptake was obtained for
compounds such as Naproxen (pKa  4), the molecule is mostly CuNH2_g_SBA-15 and was accompanied by a sorbent color change
present in solution in ionic form due to the loss of H+, as shown typical of complex formation (see Table 1). In general, these results
by Scheme 3. These electronic changes generate an additional elec- suggest that the adsorption of Naproxen occurs by means of metal
 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479 477

factor in their corresponding adsorption performance. For instance,

results obtained by EDX showed that the highest metal loading on
the siliceous support was obtained for CuNH2_g_SBA-15 followed
by Co and Ni-modified materials. In addition, even though all the
sorbent variants appeared to retain their structural order, the last
two materials were found to have the highest decrease in textural
properties, which also correlates well with their low adsorption
capacities. In general, the incorporation of copper complexes onto
the surface of SBA-15 using grafting technique seems to result in a
potential sorbent for the removal of Naproxen-like PPCPs from
water. However, tests with water samples that include other typi-
cal contaminants is in order and it would be premature to assume
that the observed selectivity will be retained.
Fig. 7 shows the fractional uptake data gathered for Naproxen
adsorption onto CuNH2_g_SBA-15. About 75–80% of the equilib-
rium loading was achieved at ca. 24 h, which indicates that there
is some slight diffusion resistance. Also included in Fig. 7 is the up-
take profile assuming a solid semi-infinite slab particle phenome-
nological model [46]. Given the sorbent particle morphology
shown in the corresponding high magnification CuNH2_g_SBA-15
SEM image (see Fig. 3d), it is evident that neither a spherical nor
a cubic shape approximation was appropriate to model the uptake
data. Fitting of a semi-infinite slab transport model yields a diffu-
sion time constant of ca. 0.019 h1 and, for an average particle size
of 1 lm, this translates to an average diffusion coefficient of
approximately 3  1016 m2/min. Adsorption kinetics tests for
Fig. 4. Fourier transform infrared spectra of (a) SBA-15; (b) CoNH2_g_SBA-15; (c)
NiNH2_g_SBA-15; and (d) Cu NH2_g_SBA-15. the other adsorbents variants were performed, but given the rela-
tively low loading amounts (please refer to Fig. 6), it was rather dif-
ficult to see appreciable changes after 24 h.
complexation based interactions, but further studies are consid- One important aspect of designing sorbents to remediate prob-
ered necessary to evaluate the definite valence state of the an- lems such as the removal of PPCPs is the ease of regeneration.
chored metals (e.g., XPS) to verify such hypothesis. Further Although mesoporous materials are thermally stable, a combustion
characterization of the best sorbent candidate together with process is out of the question here given the presence of metal–or-
Molecular Orbital (MO) theory calculations [51,52] could help to ganic complexes. Based on Fig. 6, however, it is evident that one
elucidate the nature of the adsorption interactions and we will possible way of achieving regeneration would be a pH swing. For
present these data in a separate paper when available. example, the copper-based sorbents excelled adsorbing Naproxen
Compositional analysis obtained for the metal-modified materi- at alkaline conditions but not performed well at lower pH condi-
als revealed that the loading of the metal might be an important tions. Implementation of a pH swing in a process is rather simple,

Fig. 5. Thermal gravimetric analyzes data.

478 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479

Fig. 6. Adsorption capacities for the removal of Naproxen (Ci = 14 ppm) from water at 25 °C and after ca. 24 h of equilibration.

elucidate the applicability of this material to real scenarios, includ-

ing regeneration capabilities.

Acknowledgements
Scheme 3. Naproxen molecule in neutral and ionic forms.
This work was supported by the National Science Foundation
under grants CBET-0546370 and CBET-0619349. Support from
the Puerto Rico NASA Space Grant Scholarship program is also
gratefully acknowledged. We wish to also acknowledge Professor
Carlos Rinaldi and Professor Carlos Velázquez (UPRM Chemical
Engineering) for providing access to UV–vis and SEM analysis,
respectively.

References

[1] C.P. Yu, K.H. Chu, Chemosphere 75 (2009) 1281.

[2] M. Burke, Chem. World 5 (2008) 48.
[3] C.G. Daughton, T.A. Ternes, Environ. Health Perspect. 107 (1999) 907.
[4] D.W. Kolpin, E.T. Furlong, M.T. Meyer, E.M. Thurman, S.D. Zaugg, L.B. Barber,
H.T. Buxton, Environ. Sci. Technol. 36 (2002) 1202.
[5] I. Munoz, M.J. Gomez-Ramos, A. Aguera, J.F. Garcia-Reyes, A. Molina-Diaz, A.R.
Fernandez-Alba, TRAC-Trends Anal. Chem. 28 (2009) 676.
[6] T.A. Ternes, M. Meisenheimer, D. McDowell, F. Sacher, H.J. Brauch, B.H. Gulde,
G. Preuss, U. Wilme, N.Z. Seibert, Environ. Sci. Technol. 36 (2002) 3855.
[7] K. Lin, W. Liu, J. Gan, Environ. Sci. Technol. 43 (2009) 3860.
[8] D.S. Aga, Fate of Pharmaceuticals in the Environment and in Water Treatment
Systems, CRC Press, Boca Raton, 2008.
Fig. 7. Naproxen adsorption fractional uptake in Cu NH2_g_SBA-15 at 25 °C. Solid [9] B. Halford, Chem. Eng. News 86 (2008) 13.
line represents fractional uptake model for slab-shape solid particles. [10] O.A.H. Jones, N. Voulvoulis, J.N. Lester, Crit. Rev. Environ. Sci. Technol. 35
(2005) 401.
[11] J.L. Packer, J.J. Werner, D.E. Latch, K. McNeill, W.A. Arnold, Aquat. Sci. 65 (2003)
342.
probably requiring an acid sorbent flush stage to achieve full [12] J.W. Kim, H.S. Jang, J.G. Kim, H. Ishibashi, M. Hirano, K. Nasu, N. Ichikawa, Y.
Takao, R. Shinohara, K. Arizono, J. Health Sci. 55 (2009) 249.
regeneration. [13] M. Carballa, F. Omil, J.M. Lema, Water Res. 39 (2005) 4790.
[14] T.A. Ternes, J. Stuber, N. Herrmann, D. McDowell, A. Ried, M. Kampmann, B.
Teiser, Water Res. 37 (2003) 1976.
4. Conclusions [15] A. Joss, S. Zabczynski, A. Gobel, B. Hoffmann, D. Loffler, C.S. McArdell, T.A.
Ternes, A. Thomsen, H. Siegrist, Water Res. 40 (2006) 1686.
[16] B.W. Schwab, E.P. Hayes, J.M. Fiori, F.J. Mastrocco, N.M. Roden, D. Cragin, R.D.
This study evidences that the incorporation of transition metals
Meyerhoff, V.J. D’Aco, P.D. Anderson, Regul. Toxicol. Pharmacol. 42 (2005) 296.
onto the surface of SBA-15 using grafting techniques could become [17] S.A. Snyder, P. Westerhoff, Y. Yoon, D.L. Sedlak, Environ. Eng. Sci. 20 (2003)
a potential strategy for the bottom-up design of new sorbent for 449.
[18] Y. Yoon, P. Westerhoff, S.A. Snyder, E.C. Wert, J. Membr. Sci. 270 (2006) 88.
the removal of low-concentration of Naproxen-like pharmaceutical
[19] K.A. Robberson, A.B. Waghe, D.A. Sabatini, E.C. Butler, Chemosphere 63 (2006)
drugs from water sources. At alkaline conditions, copper modified 934.
SBA-15 seems to be more effective for the removal of Naproxen [20] F.H. Frimmel, M. Assenmacher, M.U. Kumke, C. Specht, G. Abbt-Braun, G.
when compared to the other tested sorbents. This enhanced Gräbe, Chem. Eng. Process. 41 (2002) 731.
[21] M. Le Noir, A.-S. Lepeuple, B. Guieysse, B. Mattiasson, Water Res. 41 (2007)
adsorption may be attributed to the metal complexation of Na- 2825.
proxen with the copper ions. Further efforts will be necessary to [22] Z.R. Yu, S. Peldszus, P.M. Huck, Water Res. 42 (2008) 2873.
 S.M. Rivera-Jiménez et al. / Microporous and Mesoporous Materials 132 (2010) 470–479 479

[23] O. Lorphensri, J. Intravijit, D.A. Sabatini, T. Kibbey, K. Osathaphan, C. Saiwan, [37] T.X. Bui, H. Choi, J. Hazard. Mater. 168 (2009) 602.
Water Res. 40 (2006) 1481. [38] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120
[24] D. Bruhwiler, G. Calzaferri, Micropor. Mesopor. Mater. 72 (2004) 1. (1998) 6024.
[25] C. Cooper, R. Burch, Water Res. 33 (1999) 3689. [39] S. Shylesh, A.P. Singh, J. Catal. 244 (2006) 52.
[26] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, [40] F.A. Cotton, G. Wilkinson, C.A. Murillo, M. Bochmann, Advanced Inorganic
C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Chemistry, Wiley, New York, 1999.
Schlenker, J. Am. Chem. Soc. 114 (1992) 10834. [41] M. Jaroniec, M. Kruk, Langmuir 15 (1999) 5410.
[27] S. Oh, T. Kang, H. Kim, J. Moon, S. Hong, J. Yi, J. Membr. Sci. 301 (2007) 118. [42] M. Jaroniec, L.A. Solovyov, Langmuir 22 (2006) 6757.
[28] D. Perez-Quintanilla, I. del Hierro, M. Fajardo, I. Sierra, J. Mater. Chem. 16 [43] S. Garg, K. Soni, G.M. Kumaran, M. Kumar, J.K. Gupta, L.D. Sharma, G.M. Dhar,
(2006) 1757. Catal. Today 130 (2008) 302.
[29] Z. Yan, S.Y. Tao, J.X. Yin, G.T. Li, J. Mater. Chem. 16 (2006) 2347. [44] V. Meynen, P. Cool, E.F. Vansant, Micropor. Mesopor. Mater. 104 (2007) 26.
[30] S.B. Wang, H.T. Li, Micropor. Mesopor. Mater. 97 (2006) 21. [45] R.T. Yang, Gas Separation by Adsorption Processes, Imperial College Press,
[31] Z.L. Hua, W.B. Bu, Y.X. Lian, H.R. Chen, L. Li, L.X. Zhang, C. Li, J.L. Shi, J. Mater. River Edge, NJ, 1997.
Chem. 15 (2005) 661. [46] D.D. Do, Adsorption Analysis: Equilibria and Kinetics, Imperial College Press,
[32] Y. Wang, R.T. Yang, J.M. Heinzel, Chem. Eng. Sci. 63 (2008) 356. London, 1998.
[33] L. Pasqua, F. Testa, R. Aiello, S. Cundari, J.B. Nagy, Micropor. Mesopor. Mater. [47] T. Kang, Y. Park, K. Choi, J.S. Lee, J. Yi, J. Mater. Chem. 14 (2004) 1043.
103 (2007) 166. [48] K.F. Lam, K.L. Yeung, G. McKay, Micropor. Mesopor. Mater. 100 (2007) 191.
[34] J.C. Doadrio, E.M.B. Sousa, I. Izquierdo-Barba, A.L. Doadrio, J. Perez-Pariente, M. [49] M. Mureseanu, A. Reiss, I. Stefanescu, E. David, V. Parvulescu, G. Renard, V.
Vallet-Regi, J. Mater. Chem. 16 (2006) 462. Hulea, Chemosphere 73 (2008) 1499.
[35] S.M. Rivera-Jimenez, A.J. Hernandez-Maldonado, Micropor. Mesopor. Mater. [50] J.M. Rosenholm, M. Linden, Chem. Mater. 19 (2007) 5023.
116 (2008) 246. [51] A.J. Hernandez-Maldonado, R.T. Yang, Catal. Rev. 46 (2004) 111.
[36] D.Y. Zhao, J.L. Feng, Q.S. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. [52] R.T. Yang, Adsorbents: Fundamentals and Applications, John Wiley & Sons Inc.,
Stucky, Science 279 (1998) 548. New Jersey, 2003.