Adsorption of Pyruvic and Succinic Acid by Amine-Functionalized SBA15 for the Purification of Succinic Acid from Fermentation Broth

Adsorption of Pyruvic and Succinic Acid by Amine-Functionalized SBA-15 for thePurification of Succinic Acid from Fermentation Broth

Young-Si Jun,† Yun Suk Huh,† Ho Seok Park,† Arne Thomas,‡ Sang Jun Jeon,† Eun Zoo Lee,†Hyo Jin Won,† Won Hi Hong,* ,† Sang Yup Lee,† and Yeon Ki Hong§

Department of Chemical and Biomolecular Engineering, Korea AdVanced Institute of Science and Technology,Daejeon, 305-701, Korea, Max-Planck-Institute of Colloids and Interfaces, Department of Colloid Chemistry,Research Campus Golm, D-14424, Potsdam, Germany, and Department of Chemical and BiologicalEngineering, Chungju National UniVersity, Chungbuk, 380-702, Korea

ReceiVed: April 3, 2007; In Final Form: June 13, 2007

In this study, mesoporous silica SBA-15 was functionalized with primary, secondary, and tertiary amino-functional silanes onto the channel walls using a postsynthesis method as a first attempt to purify succinicacid from a fermentation broth. Ordered mesostructures of pristine and functionalized SBA-15 were evaluatedusing small-angle X-ray scattering (SAXS), transmission electron microscopy (TEM), and N2 adsorption/desorption isotherms.13C and29Si magic-angle spinning (MAS) nuclear magnetic resonance (NMR) with1Hcross-polarization (CP-MAS) and thermogravimetric analysis (TGA) revealed that amino-functional silaneswere covalently bound to the active layer of pore walls. The distribution and accessibility of amine groupswere characterized by scanning transmission electron microscopy (STEM), elemental analysis, and conductivitymeasurements. Adsorption isotherms were analyzed using the Sips model, simultaneously obtaining thetemperature dependence of isotherms derived from the isosteric heats of adsorption. Pyruvic acid had higheradsorption capacities than succinic acid on amine-functionalized SBA-15, resulting in the selective adsorptionof pyruvic acid from binary acid solution. In particular, SBA-15 functionalized with primary amino silaneobtained higher selectivity on pyruvic acid compared to that of other amine-functionalized SBA-15. Theadsorption capacities of pyruvic acid at equilibrium are dependent on the basicity and distribution of aminosilanes. The isosteric heats between 10 and 100 kJ/mol and desorption energy between 1 and 10 kJ/molrevealed that the adsorption of pyruvic and succinic acid originated from the formation of an acid-aminecomplex via hydrogen bonding. It is proposed that the amine functionalization of ordered mesoporous solidsprovides a simple and effective method of separating or purifying useful carboxylic acids.

Introduction

Over the past decade, ordered mesoporous materials such asMCM-n, HMS-n, and SBA-n have been a primary focus inmaterial science because of their regularly ordered pore ar-rangement, narrow pore-size distribution, and high specificsurface area.1 These ordered mesoporous materials offer ad-vantages over other porous solids in terms of good texturalproperties,2a hydrothermal stability,2b and tunability of pore sizeranging from 5 to 30 nm2c in applications that involveadsorption,3 catalysis,4 and use as a host material.5 To combinethese attractive properties with a specific chemical reactivityin a single solid, organofunctional groups such as amine, thiol,carboxylic, alkyl chloride, and aromatic have been incorporatedinto the structure of mesoporous material via postsynthesismethods using active silanol groups on the surface of thematerial6a or co-condensation with an organosiloxane during thepreparation of the material,6b making them an ideal host materialfor, for example, biomolecules.7 These organic-inorganic hostmaterials immobilize the biomolecules using methods such aschemical bonding,8 physical interactions,9 and layer-by-layer

encapsulation by a polymer electrolyte.10 The immobilizedbiomolecules have enhanced thermal and pH stability in aqueoussolutions and organic solvents.11 They retain their characteristicfunctionality after immobilization, as demonstrated by theirpotential for the biocatalytic process.12 In addition, a suitablechoice of grafting agent and support material makes it possibleto control both the adsorption and release of biomolecules fromthe inorganic host material, which is important in the fields ofsensing13 and drug delivery.14 These hybrid materials are alsostrong candidates for the adsorption of heavy metal ions,15a

dyes,15band toxic oxyanions15c from aqueous solutions; however,despite this wide range of applications, there are few reports,to the best of our knowledge, on the separation and purificationof carboxylic acids using mesoporous material as a potentialadsorbent.

Succinic acid, a four-carbon dicarboxylic acid produced as ametabolite of the tricarboxylic acid cycle and also as one ofthe fermentation products of anaerobic metabolism,16 hasattracted great interest because of its applications ranging frompharmaceuticals to food processing, cosmetics, biodegradablepolymers, and synthetic resins.17 Although succinic acid ismainly produced from petrochemical hydrocarbons, fermentativeproduction using ruminal bacterial species such asAnaerobio-spirillum succiniciproducens, Actinobacillus succinogenes,andMannheimia succiniciproducensis considered as a prospective

* To whom correspondence should be addressed. E-mail: [email protected].

† Korea Advanced Institute of Science and Technology.‡ Max-Planck-Institute of Colloids and Interfaces.§ Chungju National University.

13076 J. Phys. Chem. C2007,111,13076-13086

10.1021/jp072606g CCC: $37.00 © 2007 American Chemical SocietyPublished on Web 08/15/2007

alternative in line with the global demand for sustainabledevelopment.18 Biochemical production generally requires aseparation process for the recovery of succinic acid from thefermentation broth because contaminant acids such as acetic,fumaric, and pyruvic acids are also produced, in addition toimpurities such as proteins, metabolites, and media compo-nents.19 Contaminant acids normally induce product inhibitionand have an adverse effect on the recovery of succinic acid fromthe fermentation broth.20 Many separation techniques arecurrently available for the recovery of succinic acid fromaqueous solutions, including precipitation, distillation, solventextraction, membrane process, and electrodialysis;21 however,most of these techniques are inappropriate for practical processesas they consume large amounts of energy and chemicals andproduce waste during the regeneration of separation agents.

Recently, polymeric sorbents, which separate carboxylic acidsusing reversible complexation with the amine group, have shown

potential for the purification of succinic acid from aqueoussolutions.22 The strong basicity of these materials makes itpossible to sustain their separation capacity at high pH values,thereby reducing the consumption of energy and chemicalsduring separation; however, the extraction selectivity, texturalproperties, toxicological safety, and thermal and chemicalstability of these materials are somewhat unsatisfying.23 It istherefore necessary to develop new adsorbents for the efficientadsorption of contaminant acids from aqueous solutions.

Here, we report on the first attempt to apply SBA-15functionalized with primary, secondary, and tertiary aminegroups as a potential adsorbent in the separation and purificationprocesses of succinic acid. The morphology and textural andphysical properties of amine-functional groups of modified SBA-15 were characterized using transmission electron microscopy(TEM)/scanning transmission electron microscopy (STEM),small-angle X-ray scattering (SAXS), N2 adsorption/desorptionisotherms, elemental analysis, conductivity measurement, andnuclear magnetic resonance (NMR); subsequently, the adsorp-tion isotherm and separation mechanism of pyruvic and succinicacids were analyzed using the Sips model, isosteric heats ofadsorption, and thermogravimetric analysis (TGA).

Materials and Methods

Chemical Reagents.HCl (Aldrich, 37%), toluene (Sigma-Aldrich, 99%), succinic acid (Sigma-Aldrich, 99.9%), pyruvicacid (Across, 98%), phosphoric acid (Junsei), tetraethyl ortho-silicate (TEOS, Aldrich, 98%), 3-aminopropyltriethoxysilane(Aldrich, 99%),N-methylaminopropyltrimethoxysilane (Gelest),3-(N,N-dimethylaminopropyl) trimethoxysilane (Aldrich, 96%),Dowex MWA-1 (Dow Chemical Co.), and Amberlite IRA-400(Rohm and Haas Co.) were used without further purification.Pluronic P123 (EO20PO70EO20) was purchased from Aldrich,and 0.2µm PVDF syringe filter was obtained from Whatman.Distilled water (18Ω) was produced by Q tech system.

Synthesis of SBA-15. Mesoporous silica SBA-15 wassynthesized with TEOS as a silica source and P123 as astructure-directing agent according to a previously reportedmethod.24 TEOS and P123 were mixed in 1.6 M HCl solutionand were stirred at 308 K for 24 h. After filtration and dryingat 373 K, the resulting powder was calcined at 673 K for 4 h.

Functionalization of SBA-15. SBA-15 was functionalizedwith primary-, secondary-, and tertiary-amino silanes by postsyn-thesis methods.24 Briefly, SBA-15 was treated with 3-amino-propyltriethoxysilane (primary),N-methylaminopropyltrimethox-

Figure 1. TEM micrographs of pristine SBA-15 mesoporous silicaimaged in side view (a) and top view (b).

Figure 2. X-ray diffraction patterns of SBA-15 prior to and followingfunctionalization with primary-amino silane.

SBA-15 and the Purification of Succinic Acid J. Phys. Chem. C, Vol. 111, No. 35, 200713077

ysilane (secondary), and 3-(N,N-dimethylaminopropyl) trimetho-xysilane (tertiary) solution in dry toluene. The mixture washeated under reflux for 15 h. The amine-functionalized SBA-15 was filtered, was washed with acetone, and was dried at353 K under vacuum. The products were named primary-,secondary-, and tertiary-SBA-15.

Batch Adsorption of Carboxylic Acids. Amino-SBA-15 orpolymeric sorbent (0.1 g) was mixed with carboxylic acidsolutions (10 mL) in 20 mL glass scintillation vials with Teflon-lined screw caps. All experiments were conducted at controlledtemperatures for 24 h, which is sufficient for reaching a steadyequilibrium concentration. After equilibration, 1 mL samples

were withdrawn from the mixtures, were centrifuged at 6000rpm for 10 min, and were filtered with a 0.2µm PVDF syringefilter. The concentration of carboxylic acid was determined byhigh-performance liquid chromatography (HPLC) with an ionexchange column (Supelcogel C-610H, 300× 7.8 mm, SU-PELCO) using 0.1 vol % H3PO4 aqueous solution as a mobilephase. The flow rate of the mobile phase was 0.6 mL/min, andthe absorbance was measured using a UV-vis detector (Waters2487).

Characterization of Amino-SBA-15. TEM images wereobtained using a JEOL 200CX field emission transmissionelectron microscope (FE TEM, 200 kV). Finely ground powders

Figure 3. (a) Nitrogen adsorption-desorption isotherms and (b) pore-size distribution prior to and following functionalization with primary-aminosilane.

TABLE 1: Textural and Physicochemical Properties of Amino-SBA-15

materialsurface areaSBET (m2/g)

pore volume(cm3/g)

pore sizedBJH (nm) d100 (Å)

wall thickness(nm)

N content(mmol/g)a

N content(mmol/g)b

SBA-15 838.52 0.94 5.45 89.61 4.90primary 318.66 0.47 4.79 91.47 5.77 2.18 1.82secondary 283.28 0.41 4.34 90.54 6.12 2.31 1.20tertiary 332.76 0.46 4.53 91.00 5.98 1.72 0.90

a N content determined by elemental analysis.b N content determined by conductivity measurement.

13078 J. Phys. Chem. C, Vol. 111, No. 35, 2007 Jun et al.

were suspended in acetone and were deposited on Formvar-coated copper grids. STEM was operated with a probe focusedto 0.2 nm and camera length of 20 cm. The scan raster was512 × 512 points, with a dwell time of 8.5 s per scan.

X-ray diffraction patterns were recorded using a RigakuD/max-2500 (18 kW) with an image plate system equipped witha Cu KR radiation generator. Diffraction data were plotted inthe range of 0.25-10° at intervals of 0.01. The interplanarspacing of the SBA-15 structure (d100) was calculated from thefirst-order Bragg reflections using PW 1877 automated powder-diffraction software.

Elemental analysis was performed using a Vario EL 3elemental analyzer manufactured by Elementar.

Thermogravimetric analysis (TGA) was carried out using anSDT-Q600 thermal analyzer produced by TA Instrument.

Temperatures were increased in a linear ramp from 50 to1000°C at heating rates ranging from 10 to 40°C/min under anitrogen atmosphere.

All of the amino-SBA-15 samples were characterized bynitrogen gas adsorption/desorption isotherms at 77 K measuredusing a Micromeritics Tristar 3000 physisorption surface areaand pore size analyzer. Samples were pretreated by heating at200 °C for 2 h under vacuum. The average pore size wasdetermined using the Barrett-Joyner-Halenda (BJH) methodusing the desorption branch of the isotherm;25 the specificsurface area was calculated using the Brunauer-Emmett-Teller(BET) method.26

13C and29Si magic-angle spinning (MAS) nuclear magneticresonance (NMR) spectra with1H cross-polarization (CP) werecollected on a solid-state FT-NMR spectrometer (Bruker, DSX

Figure 4. (a) STEM image of secondary-SBA-15 loaded with HAuCl4 and mass mapping of (b) Si, (c) Au, and (d) Cl.

TABLE 2: Peak Assignments for 29Si MAS NMR Spectra of Amino-SBA-15

sample T2 (int%)

sampleT2

(int%)T3

(int%)Q2

(int%)Q3

(int%)Q4

(int%)Rcontenta

(mol % Si)

primary -57.8 (11.4) -67.6 (14.1) -89.1 (5.9) -100.6 (16.2) -109.8 (52.5) 25.5secondary -52.7 (2.4) -65.0 (27.8) -101.5 (21.0) -110.8 (48.8) 30.2tertiary -57.2 (18.8) -75.0 (5.0) -91.2 (4.9) -100.8 (14.7) -110.3 (56.5) 23.9

a T/(T + Q) × 100.


400 MHz) with a 4 mm MASprobe. Both13C and29Si weretaken at a spinning speed of 6 kHz.

Electrical conductivity was measured using an Accumet AR-50 conductivity meter to measure the amount of amine groupattached to the surface of SBA-15.

Results and Discussion

Characterization of Amino-Functional SBA-15 Mesopo-rous Silica.SBA-15 and amino-SBA-15 possess 2-D hexagonalarrays of silica pore channels with uniform diameter and wallthickness, as shown in Figure 1 and S1. TEM images revealthat the mesostructure of SBA-15 was intact after chemicalmodification of amino-functional silanes. These observationsin small domains are in agreement with the small-angle X-raydiffraction patterns shown in Figure 2 and S2. Three well-

resolved peaks can be attributed to (100), (110), and (200)reflections related to the hexagonal symmetry.27 The reductionin the intensity of the (100), (110), and (200) reflections forsurface-modified SBA-15 is assigned to the change in homo-geneity and space correlation of pores caused by the function-alized monolayer of amino silanes or the partial collapse of theordered structure because of the local hydrolysis by amine groupwith water molecules.28

The nitrogen adsorption/desorption isotherms in Figure 3aand S3a and c at 77 K show type IV with H1-type hysteresis,which is a typical result for hexagonal mesoporous SBA-15.29

After the chemical modification of SBA-15, a decrease insurface area, pore volume, and pore diameter was observedcompared to the original SBA-15, as shown in Table 1;30

however, it retains the mesoporosity derived from the presenceof capillary condensation with the hysteresis of open-pore andnarrow-pore size distribution, as shown in Figure 3b and S3band d. Table 1 lists the textural properties of amino-SBA-15calculated from the nitrogen physisorption and X-ray diffraction(XRD).

The quantification of amine groups on SBA-15 was conductedusing elemental analysis. The concentration of amino silaneson SBA-15 was calculated from the mass fraction of nitrogen.Since not all of the amine groups are included in the interactionbetween carboxylic acids and amine groups because of reasonssuch as steric hindrance, pore blockage, and inhomogeneousdistribution of amino silanes,7 the concentration of active sitesfor carboxylic acid was evaluated by the protonation of surfaceamine group, which is considered as a suitable method toanalyze the accessibility of adsorbate to binding sites in themesoporous material.31b Taking into consideration the fact thatprotons of sulfuric acid solutions are consumed by the proto-nation of amine group on SBA-15,31 the concentration of activeamine group was derived from the difference in electricalconductivity. As shown in Table 1, about 80% of surface aminegroups act as active sites for primary-SBA-15, while about 50%act as sites for secondary- and tertiary-SBA-15. As previouslyreported,31b the inhibition of diffusion of target material (i.e.,H+, SO4

2- in this case) is one of the main factors affecting theaccessibility to surface amine group in mesoporous systemduring the protonation. At a certain level of adsorption, the localelectric field induced from the accumulated counteranion(SO4

2-), which is introduced into the mesopore to counterbal-ance the propylammonium ion, limits the diffusion of ion speciesto the amine groups deep inside the cylindrical mesopores.Especially in this case, amino-SBA-15s had great charge effectcompared to previous results for the same system because ofthe larger counteranion of sulfuric acid than that of hydrochloricacid. The value of 50% for secondary- and tertiary-SBA-15means there was also inhibition from the steric hindrance ofmethyl side group and inhomogeneous distribution of surfaceamino silane. Further discussion about inhomogeneous distribu-tion will be given in29Si MAS NMR.

The distribution of active amine groups was analyzed to assesstheir effect on adsorption, as discussed in previous studies.15c

Steric congestion of amino silanes was observed by massmapping of Au and Cl atoms in n STEM image. HAuCl4 wasused to stain the amine groups as an indirect means of examiningthe distribution of silanes over the SBA-15 channel.32 Massmapping of Au and Cl atoms confirmed that primary aminosilanes had a generally uniform distribution on the surface ofSBA-15 without significant steric congestion (Figure 4). Inaddition, amine-functionalized SBA-15s were investigated by13C and29Si CP/MAS and29Si MAS NMR.33 Figure 5a shows

Figure 5. (a) 13C and (b)29Si CP/MAS NMR spectra of amino-SBA-15.

TABLE 3: Optimal Parameters of the Sips Equation forPyruvic and Succinic Acid Adsorption

adsorbate adsorbent qm (mmol/g) b (L/mmol) n

pyruvic acid primary 1.58 0.09 0.27secondary 0.56 0.11 0.35tertiary 0.85 0.09 0.18

succinic acid primary 0.51 0.14 0.36secondary 0.41 0.07 0.32tertiary 0.55 0.10 0.37


13C CP/MAS NMR spectra of the organic monolayer. Eachcarbon atom of the propylene group in amino silanes wasdetected and assigned to a specific peak. The presence of newpeaks at 52.2 ppm for secondary-SBA-15 and at 60.1 ppm fortertiary-SBA-15 was attributed to methylene in the amine group,indicating that three kinds of amine groups had interactedchemically with the surface silanol group of SBA-15. Figure5b shows29Si CP/MAS NMR spectra of amino-functionalsilanes. Peaks at-90,-100, and-110 ppm from SBA-15 wereassigned to Si(OH)2(SiO)2(Q2), Si(OH)(SiO)3(Q3), and Si(SiO)4-(Q4), respectively. Three additional peaks that appeared aftermodification were attributed to Si(OH)R(OCH3)(OSi) (T1), Si-(OH)R(SiO)2(T2), and SiR(SiO)3(T3), resulting from the anhy-drous deposition of silanes on SBA-15. The intensities of theQ2 andQ3 peaks were reduced with the consumption of silanolsbecause of the postfunctionalization, whereas the intensity oftheTn peaks increased during this process. Secondary-SBA-15had more closely packed monolayers than primary- and tertiary-SBA-15, as indicated by the intensity of theT3 peak; this resultwas in agreement with the results of elemental analysis and29-Si MAS NMR. The assignments of peaks and relative intensitiesfor 29Si MAS NMR are listed in Table 2.

Adsorption Isotherms of Pyruvic and Succinic Acid.Figure6 shows the adsorption isotherms of pyruvic and succinic acidson primary-, secondary-, and tertiary-SBA-15 at 25°C. Theisotherms were in good agreement with the Sips equation,34

which has the following form:

where b is the equilibrium constant (L/mmol),qm is themaximum adsorption capacity of acid (mmol/g),q is theadsorption capacity of acid at equilibrium (mmol/g),C is theinitial concentration of acid in the aqueous solution (mmol/L),andn is the exponential constant that describes the heterogeneityof the system. The values ofb, qm, andn, which were estimatedaccording to the least-squares method, are listed in Table 3.

It appears that the amounts of adsorbed acid increased withincreasing concentration of acid in the aqueous solution,becoming saturated at 20 mmol/L. In all cases, functionalizedmesoporous adsorbents had better adsorption capacity of pyruvicacid than succinic acid, reflecting the difference in acidity.Primary-SBA-15 showed the best adsorption capacity in terms

Figure 6. Adsorption isotherms of (a) succinic and (b) pyruvic acid on SBA-15 and amino-SBA-15.

q ) qm

(b‚C)1/n

1 + (b‚C)1/n(1)


of the adsorption of pyruvic acid, which exceeded the adsorptionof succinic acid by a factor of 3. Secondary- and tertiary-SBA-15 also showed increased adsorption capacity in the adsorptionof pyruvic acid, being approximately 1.5 times that of succinicacid. SBA-15 showed negligible selectivity on pyruvic acid asa consequence of only minor differences in the adsorptioncapacity of the two acids.

To compare the adsorption capability of amino-SBA-15s,distribution coefficients were calculated at various surfacecoverages, as shown in Tables 4 and 5. Primary-SBA-15exhibited the best distribution of pyruvic and succinic acid atcoverages of 0.5 and 0.7, thereby demonstrating an enhancementof 6.5, relative to succinic acid, in the distribution of pyruvicacid at a coverage of 0.7. In contrast, secondary-SBA-15 showedan enhancement of 3.8 at a coverage of 0.5, while the figurefor tertiary-SBA-15 was 2. These results indicate that primary-SBA-15 is an adsorbent of good capability for pyruvic andsuccinic acid, while pyruvic acid generally showed a betterdistribution than succinic acid in functionalized mesoporousadsorbents.

Effect of Temperature on the Adsorption of CarboxylicAcid. Isosteric Heat of Adsorption.The isosteric heat ofadsorption (-∆Hads) is an important thermodynamic variablein the design of practical adsorption processes. The adsorptionprocess is governed by the molecular cooling of the adsorbateand heat transfer from or through the adsorbent during theadsorption process; this governs the local adsorption of adsorbatemolecules and has an influence on the equilibria and kineticsof adsorption.35 The isosteric heat of adsorption, which is derivedfrom the van’t Hoff equation, is given by

where (-∆Hads) represents the isosteric heat of adsorption (kJ/mol), Q is the measurable adsorption heat (kJ/mol),R is constantparameter,Rg is the ideal gas constant (8.3145 J/mol‚K), andT0 is the reference temperature (K). These parameters wereobtained from the temperature-dependence form of adsorptionisotherms for the Sips model, which are described by thefollowing equations:34

where b0 represents the equilibrium constant at referencetemperatureT0 (L/mmol), n0 is the exponential constant atreferenceT0, qm,0 is the maximum adsorption capacity atreference temperatureT0, andø is a constant parameter. Theadsorption of pyruvic acid on amino-SBA-15 was conductedat temperatures ranging from 298 to 318 K; the derivedparameters for the Sips equation are listed in Table 6.

Figure 7 shows plots of the isosteric heats of adsorption onprimary-, secondary-, and tertiary-SBA-15 with adsorbed amountsof pyruvic acid. Surface heterogeneity is observed from thedecrease in the heats of adsorption with increasing amounts ofpyruvic acid adsorbed, simultaneously demonstrating thatsurface amine groups acted as specific adsorption sites forpyruvic acids in the aqueous solutions.36 The observed negativeenthalpies support the fact that the adsorption of pyruvic acidon amino-functional SBA-15 is an exothermic process. Thevalues between 10 and 100 kJ/mol furthermore support theinterpretation that the interaction between amino-functionalSBA-15 and pyruvic acid predominantly originates fromhydrogen bonding between amine and carboxylic groups, asreported previously.37

The heats of adsorption are higher for secondary-SBA-15 thanprimary- and tertiary-SBA-15 at lower loadings, indicating thatin comparison with other functionalized SBA-15, secondary-SBA-15 interacted more strongly with pyruvic acid. Accordingto the hard-soft acid base (HSAB) theory, the recovery ofcarboxylic acid from aqueous solution depends on the basicityof the adsorbent.22 In the case of amino silanes with identicalsilane groups, the relative basicity of the amine group depends

TABLE 4: Adsorption of Pyruvic Acid on Amino-SBA-15

concentration of PA specific ads.

adsorbentinitial

(mmol/L)final

(mmol/L) (mmol/g) coverage Kda

primary 2.49 2.27 0.02 0.01 9.7513.73 2.37 1.14 0.72 478.5219.04 4.18 1.49 0.94 355.22

secondary 3.16 3.01 0.01 0.03 4.816.74 5.06 0.17 0.30 33.08

12.97 9.06 0.39 0.70 43.0918.49 13.43 0.51 0.90 37.68

tertiary 4.00 3.56 0.04 0.05 12.4010.03 7.57 0.25 0.29 32.4313.50 7.16 0.63 0.75 88.5917.48 9.95 0.75 0.89 75.76

a Kd ) the adsorbed amount of acid (mmol/g)/[the final aqueous-phase concentration of acid (mmol/L)]× 1000.

TABLE 5: Adsorption of Succinic Acid on Amino-SBA-15

concentration of SA specific ads.

adsorbentinitial

(mmol/L)final

(mmol/L) (mmol/g) coverage Kd

primary 11.17 5.29 0.39 0.77 73.9916.11 9.17 0.46 0.91 50.45

secondary 5.68 4.80 0.06 0.14 12.0921.16 17.79 0.22 0.55 12.6431.22 25.69 0.37 0.90 14.37

tertiary 5.48 4.12 0.09 0.16 21.9310.61 6.04 0.30 0.55 50.4715.51 9.23 0.42 0.76 45.3121.05 13.63 0.49 0.90 36.30

TABLE 6: Optical Parameters for theTemperature-dependent Sips Equation

primary secondary tertiary

qm,0 (mmol/g) 1.54 0.58 0.51b0 (L/mmol) 0.09 0.11 0.07Q/RgT0 9.23 15.86 10.03n0 0.27 0.38 0.23R 11.26 19.23 55.75ø -8.57 -9.30 -8.06

TABLE 7: Kinetic Parameters Determined from TGAAnalysis

adsorbentTmax

a

(°C)Ed

(J/mol)A

(S-1)

primary 182.94 8660 863.91secondary 215.69 3180 61.74tertiary 200.13 510 5.49

a Tmax at heating rate of 20 (°C/min).

(-∆Hads) ) Q - (R‚Rg‚T0)n2 ln( q

qm - q) (2)

b ) b0 exp[ QRgT0

(T0

T- 1)] (3)

1n

) 1n0

+ R(1 -T0

T) (4)

qm ) qm,0 exp[ø(1 - TT0

)] (5)


on the organic substituents attached to the nitrogen. The pKa

values for ammonium ions are as follows: methylamine (10.66),diethylamine (10.73), and trimethylamine (9.81).38 Therefore,primary- and secondary-SBA-15 should have a stronger interac-tion with pyruvic acid than tertiary-SBA-15; however, the resultfor primary-SBA-15 was not consistent with this reasoning. Thisoutcome can be explained by steric hindrance originating fromthe more closely packed arrays of the monolayer, as shownalready in Figure 5b and Table 2. The minor decline in theisosteric heat of primary-SBA-15 also supports this explanation.

TGA Data.Thermogravimetric analysis is a useful methodfor describing the characteristic state of pyruvic acid adsorbedonto the mesoporous adsorbent. Assuming a first-order desorp-tion process, as already used in studies of the adsorption ofalkanes, alcohols, and aromatic compounds,39 the activationenergy of desorption and preexponential coefficients wasobtained using a simple model that relates the heating rate ofTGA to the temperature (Tmax) at which the desorption rate ismaximized:

where â is the heating rate of TGA (K/min),A is thepreexponential coefficient (min-1), and Ed is the activation

energy of desorption (kJ/mol). The values of the activationenergy of desorption,Ed, and preexponential coefficient,A, arelisted in Table 7.

Figure 8 shows derivative TGA plots of tertiary-SBA-15 withpyruvic acid at various heating rates. Peaks derived fromadsorbed pyruvic acid exist in the range of 170-220°C, whichis higher than the normal boiling point of pyruvic acid(165°C) and which shifts toward the higher temperature regionas the heating rate is increased. Hydrogen bonding betweenpyruvic acid and amino silanes reduced the volatility of pyruvicacid adsorbed on amino-SBA-15. The values of the activationenergy (1-10 kJ/mol) of desorption for all adsorbents supportthe fact that the primary mechanism of adsorption is hydrogenbonding between the carboxylic and amine groups. The valueof the preexponential coefficient in the range of 1013 (s-1)indicates that the thermal desorption of pyruvic acid from amino-SBA-15 involves elementary reactions without readsorption.39

Effect of pH on the Adsorption of Carboxylic Acid. Figure9 shows the pH dependence of the adsorption process. The initialconcentration of the acid solutions was 20 mmol/L, and the pHvalues of the pyruvic and succinic acid solutions were 2.43 and2.95, respectively. The concentrations of undissociated anddissociated succinic and pyruvic acid at various pH values werecalculated from the definition of the apparent equilibriumconstant via eqs (7-9):40

Figure 7. Isosteric heats of adsorption for pyruvic acid on amino-SBA-15 at 25°C.

TABLE 8: Competitive Adsorption of Fumaric and Succinic Acid on Amino-SBA-15

concentration of FA specific ads. concentration of SA specific ads.

adsorbentinitial

(mmol/L)final

(mmol/L) (mmol/g) Kd

initial(mmol/L)

final(mmol/L) (mmol/g) Kd

primary 12.42 11.29 0.11 10.01 337.55 336.34 0.12 0.36secondary 12.50 11.53 0.10 8.42 339.80 339.73 0.01 0.02tertiary 12.50 11.73 0.08 6.57 340.23 340.22 0.00a 0.00b

a 9.32× 10 -4. b 2.73× 10 -3.

TABLE 9: Competitive Adsorption of Pyruvic and Succinic Acid on Amino-SBA-15

concentration of PA specific ads. concentration of SA specific ads.

adsorbentinitial

(mmol/L)final

(mmol/L) (mmol/g) Kd

initial(mmol/L)

final(mmol/L) (mmol/g) Kd

primary 26.92 13.94 1.30 93.09 361.01 360.67 0.03 0.09secondary 26.92 22.22 0.47 21.14 361.01 351.80 0.92 2.62tertiary 26.06 23.23 0.28 12.20 353.12 351.24 0.19 0.53

âRgTmax

2) A

Edexp(-

Ed

RgTmax) (6)


Figure 9. pH dependence of adsorption of (a) pyruvic acid on amino-SBA-15 and concentration of pyruvic acid and pyruvate anion and (b)succinic acid on amino-SBA-15 and concentration of succinic acid and bisuccinate anion at various pH values.

Figure 8. Derivative thermogravimetry of tertiary-SBA-15 saturated with pyruvic acid.


where SA is succinic acid, BA is bisuccinate anion, and PA ispyruvic acid.

The uptake of pyruvic and succinic acid was significantlyaffected by varying the pH value. The similar profile of theuptake to that of the concentration of undissociated carboxylicacid confirms the adsorption results from the hydrogen bondsbetween the amine and carboxylic groups. At pH values higherthan pKa, carboxylic acid exists as a carboxylate anion, resultingin a significant reduction in the uptake by the amine group.Although there are still bisuccinate anions with one carboxylicgroup at pH values higher than 6, the uptake of succinic acid isclose to zero in these regions. The same trend was observed inthe adsorption of pyruvic acid; however, the uptake did not fallto zero at pH values higher than pKa in the case of primary-and tertiary-SBA-15. This can be explained by the hydrogenbond between the oxygen of the carbonyl group and thehydrogen of the amine group. Low uptake at pH values belowthe ranges of pKa is explained by the consumption of adsorptionsites that accompanies the formation of ammonium salt withH+ from HCl.

Competitive Adsorption of Contaminant and SuccinicAcid. The competitive adsorption of contaminant and succinicacid was conducted on the basis of the composition of thefermentation broth. Initial concentration ratios of succinic tocontaminant acids were about 27 for fumaric acid and 13 forpyruvic acid as shown in Tables 8 and 9. The degree ofselectivity on contaminant acids over succinic acid was quanti-fied by the selectivity coefficient for adsorption,k, which isdefined as

where Kd,i is the distribution coefficient of speciesi in themesoporous adsorbent. Despite recording a value more than 10-fold the amount of succinic acid, secondary- and tertiary-SBA-15 exhibited selectivities of 429.94 and 2399.54 for fumaricacid, respectively, while primary-SBA-15 recorded a selectivityof 988.59 for pyruvic acid, as shown in Figure 10. This meansthat secondary- and tertiary-SBA-15 showed preferential ad-sorption of fumaric acid and that primary-SBA-15 showedpreferential adsorption of pyruvic acid. The adsorption selectiv-ity of mesoporous adsorbents on contaminant acids was muchhigher than that of polymeric sorbents such as Dowex MWA-1and Amberlite IRA-400. As previsouly reported,40 these poly-meric sorbents with strong basicity showed selectivity only inthe region between 1 and 20 for adsorption of succinic acidover lactic acid and formic acid over acetic acid in equalmolarity solution. In particular, the enhanced selectivity isobtained at low pH values, while the maximum selectivity ofpolymeric sorbents occurs at a pH of about 6 because of thedifference in dissociation of acid with different pKa values; thisarises because the pH of the aqueous solution is maintained atpH values between 2 and 3 after the primary purification stepthat uses reactive extraction for the effective separation ofsuccinic acid from the fermentation broth.42 The selectivity ofthe primary-, secondary-, and tertiary-SBA-15 for contaminantacids at low pH values should be exploited during the secondarypurification or polishing step22 and makes amino-SBA-15attractive as adsorbents for the effective recovery of succinicacid from fermentation broth with greater than 99% purity.

Conclusions

We investigated the adsorption of pyruvic and succinic acidon SBA-15 functionalized with primary-, secondary-, andtertiary-amino silane. It appears that the adsorption originatedfrom the formation of acid-amine complexes via hydrogenbonding, as indicated by the analyzed isosteric heats of

Figure 10. Selectivity coefficient for competitive adsorption of fumaric/succinic acid and pyruvic/succinic acid on amino-SBA-15s and polymericsorbents; some data for amino-SBA-15 are multiplied by 0.1; initial concentrations of each species are [FA]0 ) 12.50 (mmol/L) and [SA]0 )340.23 (mmol/L) for fumaric/succinic acid system and [PA]0 ) 26.01 (mmol/L) and [SA]0 ) 361.01 (mmol/L) for pyruvic/succinic acid system.

[SA] )[SA]t

1 + 10pH-pKA1 + 102pH-pKA1-pKA2(7)

[BA] )[SA]t‚10pH-pKA1

1 + 10pH-pKA1 + 102pH-pKA1-pKA2(8)

[PA] )[PA]t

1 + 10pH-pKA(9)

ki/j )Kd,i

Kd,j(10)


adsorption and thermogravimetric analysis. Primary- and tertiary-SBA-15 showed the maximum adsorption capacity for pyruvicacid (1.58 mmol/g) and succinic acid (0.55 mmol/g), respec-tively. The amine-functionalized SBA-15 adsorbed more pyruvicacid than succinic acid at the same coverage. The presence ofinactive sites on the adsorbents was identified from conductivitymeasurements, which were in good agreement with the resultfrom the adsorption of pyruvic acid. Adsorption was alsoaffected by the basicity and distribution of amino silanes;therefore, it is believed that the isolated amino silanes areeffective in the adsorption of carboxylic acid. Competitiveadsorption revealed that amine-functionalized SBA-15 is suitablefor the removal of contaminant acids from fermentation broth.

Acknowledgment. The authors are grateful to the Centerfor Advanced Bioseparation Technology Research (BSEP,KOSEF) and Brain Korea 21 (BK21) for the funding.

Supporting Information Available: TEM micrographs,X-ray diffraction patterns, and nitrogen adsorption-desorptionisotherms. This material is available free of charge via theInternet at http://pubs.acs.org.

References and Notes

(1) Hartmann, M.Chem. Mater.2005, 17, 4577.(2) (a) Zhao, D.; Sun, J.; Stucky, G. D.Chem. Mater.2000, 12, 275.

(b) Choi, M.; Heo, W.; Klietz, F.; Ryoo, R.Chem. Commun.2003, 1340.(c) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka,B. F.; Stucky, G. D.Science1998, 279, 548.

(3) (a) Kang, T.; Park, Y.; Choi, K; Lee, J. S.; Yi, J.J. Mater. Chem.2004, 14 (6), 1043. (b) Katiyar, A.; Yadav, S.; Smirniotis, P. G.; Pinto, N.G. J. Chromatogr., A2006, 1122, 13. (c) Liu, X.; Zhou, L.; Huang, D.;Zhou, Y. Chem. Phys. Lett.2005, 415, 198.

(4) (a) Yang, L.; Qi, Y.; Yuan, X.; Shen, J.; Kim,J. Mol. Catal. A2005, 229, 199. (b) Byambajav, E.; Ohtsuka, Y.Appl. Catal., A2003, 252(1), 193.

(5) (a) Wang, S.; Choi, D. G.; Yang, S. M.AdV. Mater.2002, 14 (18),1311. (b) Lee, J.; Jin, S.; Hwang, Y.; Park, J. G.; Park, H. M.; Hyeon, T.Carbon2005, 43 (12), 2536. (c) Zhang, L.-X.; Shi, J.-L.; Yu, J.; Hua, Z.-L.; Zhao, X.-G.; Ruan, M.-L.AdV. Mater. 2002, 14 (20), 1510.

(6) (a) Feng, X.; Fryxell, G. E.; Wang, S.-Q.; Kin, A. Y.; Liu, J.;Kemner, K. M.Science1997, 276, 923. (b) Voss, R.; Thomas, A.; Antonietti,M.; Ozin, G. A. J. Mater. Chem.2005, 15, 4010.

(7) Han, Y. J.; Stucky, G. D.; Butler, A.J. Am. Chem. Soc.1999, 121,9897.

(8) Yiu, H. H. P.; Wright, P. A.; Botting, N. P.J. Mol. Catal. B2001,15, 81.

(9) Avnir, D.; Braun, S.; Lev, O.; Ottolenghi, M.Chem. Mater.1994,6, 1605.

(10) Wang, Y.; Caruso, F.Chem. Mater.2005, 17, 953.(11) (a) Wang, P.; Dai, S.; Waezsada, S. D.; Tsao, A. Y.; Davison, b.

H. Biotechnol. Bioeng.2001, 74, 249. (b) Pandya, P. H.; Jasra, R. V.;Newalkar, B. L.; Bhatt, P. N.Microporous Mesoporous Mater.2005, 77,67.

(12) Wang, Y.; Caruso, F.Chem. Mater.2005, 17, 953.(13) Martin, B. D.; Gaber, B. P.; Patterson, C. H.; Turner, D. C.

Langmuir1998, 14, 3971.(14) Inglis, W.; Sanders, G. H.; Williamsan, P. M.; Davies, M. C.;

Roberts, C. J.; Tendler, S. J. B.Langmuir2001, 17, 7402.(15) (a) Wu, G.; Wang, Z.; Wang, J.; He, C.Anal. Chim. Acta2007,

582 (2), 304. (b) Dhodapkar, R.; Rao, N. N.; Pande, S. P.; Kaul, S. N.

Bioresour. Technol.2006, 97, 877. (c) Yoshitake, H.; Yokoi, T.; Tatusmi,T. Chem. Mater.2002, 14, 4603. (d) Lei, C.; Shin, Y. S.; Liu, J.; Ackerman,E. J.J. Am. Chem. Soc.2002, 124, 11242.

(16) Kwon, Y. D.; Lee, S. Y.; Kim, P.J. Microbiol. Biotechnol.2006,16 (9), 1448.

(17) Zeikus, J. G.; Jain, M. K.; Elankovan, P.Appl. Microbiol.Biotechnol.1999, 51, 545.

(18) (a) Hong, S. H.; Kim, J. S.; Lee, S. Y.; In, Y. H.; Choi, S. S.; Rih,J. K.; Kim, C. H.; Jeong, H.; Hur, C. G.; Kim, J. J.Nat. Biotechnol.2004,22, 1275. (b) Samuelov, N. S.; Lamed, R.; Lowe, S.; Zeikus, J. G.Appl.EnViron. Microbiol. 1991, 57, 3013.

(19) Davison, B. H.; Nghiem, N. P.; Richardson, L. R.Appl. Biochem.Biotechnol.2004, 113, 653.

(20) (a) Hong, Y. K.; Hong, W. H.; Chang, H. N.Biotechnol. Lett.2000,22, 871. (b) Davison, B. H.; Nghiem, N. P.; Richardson, L. R.Appl.Biochem. Biotechnol.2004, 113, 653.

(21) (a) Bessling, B.; Loening, J. M.; Ohligschlaeger, A.; Schembecker,G.; Sundmacher, L.Chem. Eng. Technol.1998, 21, 393. (b) Han, D. H.;Hong, W. H.Sep. Sci. Technol.1996, 31, 1123. (c) Zeikus, J. G.; Elankovan,P.; Grethlein, A.Chem. Proc.1995, 58 (7), 71. (d) Choi, J.; Hong, W. H.J. Chem. Eng. Jpn. 1999, 32 (2), 184.

(22) (a) Tung, L. A.; King, C. J.Ind. Eng. Chem. Res.1994, 33, 3217.(b) Garcia, A. A.Biotechnol. Prog.1991, 7, 33.

(23) Ju, Y.; Webb, O.; Dai, S.; Lin, J.; Barnes, C.Ind. Eng. Chem. Res.2000, 39, 550.

(24) Salis, A.; Meloni, D.; Ligas, S.; Casula, M. F.; Monduzzi, M.;Solinas, V.; Dumitriu, E.Langmuir2005, 21, 5511.

(25) Linsen, B. G.; van den Heuvel, A.The Solid Gas Interface; MercelDekker; New York, 1967.

(26) Brunauer, S.; Emmett, P. H.; Teller, E.J. Am. Chem. Soc.1938,60, 309.

(27) Beck, J. S.; Vartuli, C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C.T.; Schmitt, K. D.; Chu, C. T.-W.; Olson, D. H.; Sheppard, E. W.; McCullen,S B.; Higgins, J. B.; Schlenker, J. L.J. Am. Chem. Soc.1992, 114, 10834.

(28) (a) Zhao, X. S.; Lu, G. Q.J. Phys. Chem. B1998, 102, 1556. (b)Evans, J.; Zaki, A. B.; El-Sheikh, M. Y.; El-Safty, S. A.J Phys. Chem. B2000, 104, 10271.

(29) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti,R. A.; Rouquerol, J.; Siemienewska, T.Pure Appl. Chem.1985, 57, 603.

(30) (a) Yiu, H.; Wright, P. A.; Botting, N. P.J. Mol. Catal. B2001,15, 81. (b) Antochshuk, V.; Jaroniec, M.Chem. Mater.2000, 12, 2496.

(31) (a) Yoon, J. Y.; Park, H. Y.; Kim, J. H.; Kim, W. S.J. ColloidInterface Sci.1996, 177, 613. (b) Walcarius, A.; Etienne, M.; Lebeau, B.Chem. Mater.2003, 15, 2161. (c) Walcarius, A.; Etienne, M.; Bessiere, J.Chem. Mater.2002, 14, 2757. (d) Yang, J. J.; El-Nahhal, I. M.; Chuang,I.-S.; Maciel, G. E.J. Non-Cryst. Solids1997, 209, 19.

(32) (a) Gu, J. L.; Shi, J. L.; You, G. J.; Xiong, L. M.; Qian, S. X.;Hua, Z. L.; Chen, H. R.AdV. Mater. 2005, 17 (5), 557. (b) Engel, A.;Colliex, C. Curr. Opin. Biotechnol.1993, 4, 403.

(33) Feng, X.; Fryxell, G. E.; Wang, L.-Q.; Kim, A. Y.; Liu, J.; Kemner,K. M. Science1997, 276, 923.

(34) Duong, D. D.Adsorption Analysis, Equilibria and Kinetics; ImperialCollege Press: London, 1998.

(35) (a) Chakraborty, A.; Saha, B. B.; Koyama, S.Appl. Phys. Lett.2006, 89, 171901. (b) Sircar, S.; Cao, D. V.Chem. Eng. Technol.2002,25, 10.

(36) Chin, D.; King, C. J.Ind. Eng. Chem. Res.1999, 38, 3738.(37) Liu, B. J.; Ren, Q. L.J. Colloid Interface Sci.2006, 294, 281.(38) (a) McMurry, J.Organic Chemistry; Thomson Learning; Brooks/

Cole: Pacific Grove, CA, 2000. (b) Yoshida, H.; Kishimoto, N.Chem. Eng.Sci.1995, 50, 2203.

(39) (a) Senf, L.; Frank, H.J. Chromatogr.1990, 520, 131. (b) Suzuki,M.; Misic, D. M.; Joyama, O.; Kawazoe, K.Chem. Eng. Sci.1978, 33,271.

(40) Husson, S. M.; King, C. J.Ind. Eng. Chem. Res.1999, 38, 502.(41) Jun, Y. S.; Huh, Y. S.; Hong, W. H.; Hong, Y. K.Biotechnol. Prog.

2005, 21, 1673.


Adsorption of Pyruvic and Succinic Acid by Amine-Functionalized SBA15 for the Purification of Succinic Acid from Fermentation Broth

Documents