Separation and Purification Technology 48 (2006) 197–201
Short communication
Amino acid adsorption onto mesoporous silica molecular sieves Andrea J. O’Connor a,∗ , Akiko Hokura b , Jenny M. Kisler a , Shogo Shimazu c , Geoffrey W. Stevens a , Yu Komatsu d a Department of Chemical and Biomolecular Engineering, University of Melbourne, Vic. 3010, Australia Department of Applied Chemistry, Tokyo University of Science, 1-3 Kagurazaka, Shinjuku, Tokyo 162-8601, Japan c Department of Materials Science, Chiba University, 1-33, Yayoi-cho, Inage-ku, Chiba-shi, Chiba 263-8522, Japan Department of Environmental Systems Engineering, Kanazawa Institute of Technology, 7-1 Ohgigaoka, Nonoichi, Ishikawa 921-8501, Japan b
d
Abstract Mesoporous molecular sieves are promising as adsorbents for purification of biological molecules, such as amino acids, due to their tuneable mesopore sizes and high surface area. In this study, the adsorption of the basic amino acid, lysine, onto MCM-41, a siliceous mesoporous molecular sieve, has been investigated under a range of solution conditions. It was found to adsorb according to a Langmuir-type isotherm with a maximum capacity at pH 6 of 0.21 mmol/g. The extent of adsorption depends strongly on the pH and ionic strength of the adsorbate solution, due to a combination of ion exchange and electrostatic interactions governing the adsorption process. © 2005 Elsevier B.V. All rights reserved. Keywords: Amino acid; Adsorption; Mesoporous molecular sieve; MCM-41; Lysine
1. Introduction Amino acids are significant in the food industry as nutritional supplements and ingredients in parenteral solutions, as well as being used as building blocks for production of pharmaceutical and agrochemical compounds. These applications require amino acids to be supplied at high purities by processes such as ion exchange chromatography from natural or synthetic sources. Amino acids are also interesting as model adsorbates due to their molecular size and zwitterionic nature. Mesoporous molecular sieves have been identified as promising adsorbents for biochemical molecules due to their tuneable pore sizes with narrow pore size distributions in the mesopore range (i.e., 2–50 nm) of relevance for molecules such as amino acids, peptides and proteins, as well as their high surface areas and large pore volumes [1–10]. To date, relatively few studies have investigated this potential, particularly for amino acids, which are small enough to fit easily
within the pores of a wide range of mesoporous molecular sieves [1,11,12]. MCM-41 is a mesoporous molecular sieve, developed by scientists at Mobil Corporation [13,14], with a regular hexagonal array of mesopores, the size of which can be tailored between 2 and 10 nm by altering the synthesis conditions. It also possesses high surface area (>1000 m2 /g) and can be synthesised as a silicate or aluminosilicate matrix. It has been identified as an excellent model adsorbent as well as having potential for use in industrial separation processes [1,15]. In this study, the adsorption of the divalent basic amino acid, lysine, on siliceous MCM-41 was studied under a range of solution conditions to provide insight into the adsorption kinetics, capacity and mechanisms. Lysine is a significant product of current industrial fermentation followed by ion exchange purification processes [16].
2. Experimental ∗
Corresponding author. Tel.: +61 3 8344 8962; fax: +61 3 8344 4153. E-mail addresses:
[email protected] (A.J. O’Connor),
[email protected] (A. Hokura),
[email protected] (S. Shimazu),
[email protected] (G.W. Stevens),
[email protected] (Y. Komatsu). 1383-5866/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2005.07.007
2.1. Materials Sodium silicate solution (27% SiO2 and 14% NaOH) and cetyltrimethylammonium bromide (CTAB) were obtained
A.J. O’Connor et al. / Separation and Purification Technology 48 (2006) 197–201
198
from Aldrich. dl-Lysine monohydrochloride was purchased from Wako Pure Chemical Industries. Analytical grade sulfuric acid, sodium chloride, hydrochloric acid and sodium hydroxide were also used. Distilled and de-ionized water used throughout the experiments was prepared using a Milli-Q system (Nihon Millipore Kogyo, Tokyo). All reagents were used as received. 2.2. Synthesis Siliceous MCM-41 was prepared by a hydrothermal method similar to that reported in the literature [13,17]. CTAB was used as the structure directing agent with sodium silicate as the silica source. For a typical batch, CTAB (16.4 g) was dissolved in distilled water (69.2 g) with vigorous stirring and heating at 80 ◦ C. Sodium silicate solution (19.0 g) and distilled water (40.6 g) were then combined with the CTAB solution. The solution was mixed and heated thoroughly for 10 min. Then, the pH of the mixture was adjusted to 11.0 by dropwise addition of a 6N H2 SO4 solution. The solution was stirred for a further 30 min, and then moved to a closed PTFE vessel with stainless steel jacket. The vessel was held at 100 ◦ C for 144 h. After cooling the vessel to room temperature, the precipitated white powder was filtered, washed, and dried under vacuum. The surfactant was removed from the silicate framework by calcination; the sample was heated up to 550 ◦ C and held at this temperature for 1 h under flowing nitrogen, followed by 6 h of flowing air. All samples were stored in a desiccator where the relative humidity was less than 20%, because the porous structure can be damaged by exposure to water [18,19].
Fig. 1. Typical XRD spectra of MCM-41 samples before and after calcination.
2.3. Characterisation Powder X-ray diffraction (XRD) data were collected on Rigaku and Philips diffractometers using Cu K␣ radiation. The spectra before and after calcination (Fig. 1) could be indexed to a hexagonal lattice as expected for MCM-41 [20], and showed some contraction of the structure upon calcination, indicated by the decrease in the unit cell parameter, a0 (Table 1). Nitrogen adsorption and desorption isotherms were measured for the calcined samples at 77 K using BELSORP 28SA and Micromeritics ASAP 2000 systems. The samples were heated to about 200 ◦ C under vacuum for at least 2 h to remove any adsorbed water before the isotherms were recorded. The shape of the isotherms, shown in Fig. 2, was also as expected for MCM-41 with a step at a relative pressure of around 0.3 [20]. The surface area (SBET ) was cal-
Table 1 Typical structural properties of MCM-41 Sample
a0 (nm)
SBET (m2 /g)
Vpore (cm3 /g)
Dpore (nm)
Before calcination After calcination
4.9 4.4
– 918
– 1.04
– 2.74
Fig. 2. (a) Typical nitrogen adsorption and desorption isotherms for MCM41 at 77 K and (b) DH pore size distribution.
A.J. O’Connor et al. / Separation and Purification Technology 48 (2006) 197–201
199
culated using the BET model [21,22] and was greater than 800 m2 /g for all samples. The pore diameter and pore volume were estimated using the Dollimore and Heal (DH) method [23,24]. Elemental analysis of the MCM-41 was performed using an inductively coupled plasma atomic emission spectrophotometer (Seiko Instruments, Chiba, Japan). The amount of Si was determined to be 37.3% (w/w) and that of Na was 1.6% (w/w). The amount of carbon was determined to be 0.3% (w/w) using a total carbon analyser (Leco, USA), indicating that the calcination process effectively removed the surfactant. 2.4. Amino acid adsorption The extents of adsorption of lysine onto MCM-41 under a range of solution concentrations, pH values, ionic strengths and contact times were determined using batch adsorption tests at 25 ± 1 ◦ C and solution depletion analysis. MCM41 samples (∼40 mg) and solutions of dl-lysine monohydrochloride in water (10 mL) were mixed in sealed plastic vials using a rotating mixer (Heto Rotamix RK) operating at 40 rpm. The pH of the solutions was adjusted prior to mixing using 0.1 M NaOH or 0.1 M HCl. After a chosen contact time the solution pH was measured and the suspensions were separated by centrifugation prior to analysis of the supernatant using a Total Organic Carbon (TOC) Analyser (Shimadzu TOC-5000A). Standard curves were generated for lysine over appropriate concentration ranges for each set of samples, and standard solutions were used to check the analyser performance at regular intervals during analyses. The concentrations determined were corrected for the (small) effect of residual carbon in the MCM-41 on the TOC analysis, determined from a control batch test run under the same experimental conditions but with no lysine in solution.
3. Results and discussion Lysine (Fig. 3) was selected for this study because it is a small (MW = 146.2) basic amino acid (pI = 9.7). Thus, it is positively charged in solutions around neutral pH, whereas silica carries a net negative charge (pI ≈ 2.0) [16,25].
Fig. 3. The monovalent cationic form of lysine, predominant at pH 4–6.
Fig. 4. Amount of lysine adsorbed as a function of contact time (initial pH 5.7, initial [lysine] = 0.33 mmol/L, 25 ◦ C).
3.1. Adsorption isotherm The time dependence of lysine adsorption on MCM-41 was investigated (Fig. 4) to determine the time required for equilibrium to be reached between the solid and solution. The adsorbed amount increased with contact time, reaching equilibrium within about 20 h. The amount adsorbed was monitored over extended periods and no change was seen over contact times less than 200 h. Beyond this, a slow decrease in adsorbed amounts began, which was attributed to a loss of stability in the MCM-41. This is expected as the porous structure of MCM-41 has been shown to degrade upon extended exposure to water or water vapour [18,19]. Thus, further adsorption experiments were conducted for contact times of between 20 and 190 h to ensure equilibrium was reached without significant deterioration of the MCM-41 structure. Fig. 5 shows the adsorption isotherm for lysine on MCM41 at 25 ◦ C with an initial solution pH of 6.0. The data fitted the Langmuir adsorption isotherm equation well, with a maximum adsorption capacity of 0.21 mmol/g. This is comparable to a previous report [12] showing ca. 0.22 mmol/g of
Fig. 5. Adsorption isotherm for lysine onto MCM-41 (initial pH 6.0, 25 ◦ C).
200
A.J. O’Connor et al. / Separation and Purification Technology 48 (2006) 197–201
lysine adsorbed onto a siliceous MCM-41 sample prepared with C14 H29 N(CH3 )3 Br as the structure directing surfactant, in place of the C16 H33 N(CH3 )3 Br used in this work. As expected, the pore size of the C14 -MCM-41 was slightly smaller (Dpore = 2.5 nm according to the BJH model) than the C16 -MCM-41 synthesised here, although its BET surface area was comparable (1070 m2 /g). The fact that essentially the same amount of lysine was adsorbed on the two materials indicates that the pore size did not significantly affect their capacity for lysine, which fits easily into the pores. This is in contrast to studies involving larger adsorbates such as proteins, where the amount of a solute adsorbed decreases significantly as the solute size approaches the pore size [4,5,9]. 3.2. Adsorption mechanism The effect of electrostatics on the adsorption of lysine on MCM-41 was tested by measuring the amounts adsorbed under a range of pH and ionic strength conditions. Fig. 6 shows that the amount adsorbed increased strongly with increasing pH over the range 3–6. This range is between the isoelectric points of the silica surface and lysine and so the silica is negatively charged and lysine is positively charged. Lysine exists as a mixture of its monovalent and divalent cationic forms at pH values between 1 and 4, whereas the monovalent form predominates at pH 4–6 [16,26]. The point of zero charge for silica is generally reported as being around 2 ± 0.5 [25] although a slightly higher value of 3.2 was determined for one siliceous MCM-41 sample [12]. Hence, the charge on the silica surface would be close to zero at pH 3 and become increasingly negative as the pH increases, explaining the increasing amounts of lysine adsorbed. The solution pH changed during the lysine uptake process as shown in Fig. 6. At low pH, little lysine was adsorbed and there was a correspondingly small change in pH. However, in the pH range 4.5–6, the pH decreased dur-
Fig. 6. Lysine adsorption onto MCM-41 at different solution pH values and corresponding changes in solution pH during adsorption (initial [lysine] = 0.62 mmol/L, 25 ◦ C).
Fig. 7. Lysine adsorption onto MCM-41 with different added salt concentrations (initial pH 5.5, initial [lysine] = 0.30 mmol/L, 25 ◦ C).
ing uptake, indicating that protons were being released into solution as lysine was adsorbed. This is consistent with an ion exchange uptake mechanism in which protons from the silanol groups on the silica surface are exchanged with the positively charged lysine molecules. However, the pH change is less than would be expected if uptake occurred solely through ion exchange (e.g., at pHfinal = 4.7, pH = −1.3, whereas a change of −1.7 would be expected if ion exchange were the only adsorption mechanism). Thus, some lysine is likely to be adsorbing through electrostatic attraction to negatively charged, deprotonated silanol groups on the surface. The pH change for solutions with pHfinal > 5 was somewhat less than that at pHfinal 4.5–5. This may indicate a shift in the adsorption mechanism with less ion exchange and more electrostatic binding as the silica surface becomes more deprotonated. Further evidence of the important role of electrostatic interactions in the adsorption process came from experiments conducted at a fixed initial pH and increasing ionic strength (Fig. 7). The amount of lysine adsorbed decreased strongly as more sodium chloride was added to the initial lysine solution. This confirms that electrostatics is an important factor in adsorption under these conditions, as added electrolyte results in increases in electrostatic shielding and competition for binding sites on the silica surface. In fact, the amount of lysine adsorbed at the highest added salt concentration (0.1 M) was only 4 mol% of that adsorbed with no added salt. The adsorption capacity of the MCM-41 for lysine under the conditions tested here is lower than that on some other adsorbent materials, such as zeolite Beta (0.4 mmol/g) [27]. This is likely to be due to the relatively low acidity of the siliceous MCM-41 surface [28]. However, the capacity could be increased by rehydrating the adsorbent after calcination to increase the silanol group density [29] or by modifying the surface to incorporate functional groups, such as sulfonic acid groups, targeted for adsorption of the amino acid. The latter approach has the added benefit that surface modifications can be tailored to also enhance the stability of the mesoporous molecular sieve in aqueous solutions [19].
A.J. O’Connor et al. / Separation and Purification Technology 48 (2006) 197–201
4. Conclusion The basic amino acid, lysine, adsorbs on siliceous MCM41 following a Langmuir-type isotherm with a maximum capacity at pH 6 of 0.21 mmol/g. The extent of adsorption depends on the solution pH and ionic strength, with a combination of electrostatic interactions and ion exchange governing the uptake of lysine. MCM-41 shows promise as a high surface area adsorbent for biological molecules such as amino acids, although its stability in the presence of water requires improvement before industrial application can be implemented.
Acknowledgements A.O. gratefully acknowledges support of a Science and Technology Agency of Japan Fellowship and the National Institute for Research into Inorganic Materials, Japan. Authors from the University of Melbourne acknowledge access to infrastructure from the Particulate Fluids Processing Centre, a special research centre of the Australian Research Council (ARC).
References [1] A. Daehler, G.W. Stevens, A.J. O’Connor, Bioadsorption and separation with nanoporous materials, in: G.Q.M. Lu, X.S. Zhao (Eds.), Nanoporous Materials—Science and Engineering, Imperial College Press, London, 2004, pp. 812–848. [2] J. Deere, E. Magner, J.G. Wall, B.K. Hodnett, Adsorption and activity of cytochrome c on mesoporous silicates, Chem. Commun. 5 (2001) 465–466. [3] J.F. Diaz, K.J. Balkus, Enzyme immobilization in MCM-41 molecular sieve, J. Mol. Catal. B 2 (2–3) (1996) 115–126. [4] J. Kisler, G.W. Stevens, A.J. O’Connor, Adsorption of proteins on mesoporous molecular sieves, Mater. Phys. Mech. 4 (2) (2001) 89–93. [5] J. Kisler, A. Daehler, G.W. Stevens, A.J. O’Connor, Separation of biological molecules using mesoporous molecular sieves, Microporous Mesoporous Mater. 44–45 (2001) 769–774. [6] J. Kisler, G.W. Stevens, A.J. O’Connor, Protein adsorption on mesoporous silicates, in: S. Iveson (Ed.), Chemeca’99, The Institution of Engineers Australia, Canberra, 1999, p. 134. [7] J. Kisler, G.W. Stevens, A.J. O’Connor, Mesoporous molecular sieves as adsorbents for bioseparations, in: 13th International Zeolite Conference Recent Research Reports, Groupe Franc¸ais des Z´eolithes, Paris, 2001, p. 18-R-08. [8] H. Takahashi, B. Li, T. Sasaki, C. Miyazaki, T. Kajino, S. Inagaki, Immobilized enzymes in ordered mesoporous silica materials and improvement of their stability and catalytic activity in an organic solvent, Microporous Mesoporous Mater. 44 (2001) 755– 762.
201
[9] J. Yang, A. Daehler, G.W. Stevens, A.J. O’Connor, Adsorption of lysozyme and trypsin onto mesoporous silica materials, Stud. Surf. Sci. Catal. 146 (2002) 775–778. [10] H.H.P. Yiu, C.H. Botting, N.P. Botting, P.A. Wright, Size selective protein adsorption on thiol-functionalised SBA-15 mesoporous molecular sieve, Phys. Chem. Chem. Phys. 3 (15) (2001) 2983–2985. [11] G.M.S. El Shafei, Effect of pore structure on surface characteristics of zirconium phosphate-modified silica, J. Coll. Int. Sci. 250 (2) (2002) 394–399. [12] S. Ernst, M. Hartmann, S. Munsch, Adsorption of selected amino acids from aqueous solutions on mesoporous molecular sieve, Stud. Surf. Sci. Catal. 135 (2001), 28-P-10. [13] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, A new family of mesoporous molecular sieves prepared with liquid crystal templates, J. Am. Chem. Soc. 114 (1992) 10834–10843. [14] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Ordered mesoporous molecular sieves synthesized by a liquid-crystal template mechanism, Nature 359 (1992) 710–712. [15] J.A. Herbst, C.T. Kresge, D.H. Olson, K.D. Schmitt, J.C. Vartuli, D.I.C. Wang, Method for Separation of Substances, Mobil Oil Corporation, USA, 1995, US 5378440. [16] H. Nagai, G. Carta, Lysine adsorption on cation exchange resin. I. Ion exchange equilibrium and kinetics, Sep. Sci. Tech. 39 (16) (2004) 3691–3710. [17] K.J. Edler, J. Dougherty, R. Durand, L. Iton, G. Kirton, G. Lockhart, Z. Wang, R. Withers, J.W. White, Small angle X-ray scattering from MCM-41 and its synthesis gels: optimisation of the synthesis parameters, Coll. Surf. A 102 (1995) 213–230. [18] X.S. Zhao, F. Audsley, G.Q. Lu, Irreversible change of pore structure of MCM-41 upon hydration at room temperature, J. Phys. Chem. B 102 (1998) 4143–4146. [19] J. Kisler, M.L. Gee, G.W. Stevens, A.J. O’Connor, Comparative study of silylation methods to improve the stability of silicate MCM41 in aqueous solutions, Chem. Mater. 15 (3) (2003) 619–624. [20] X.S. Zhao, G.Q. Lu, G.J. Millar, Advances in mesoporous molecular sieve mcm-41, Ind. Eng. Chem. Res. 35 (1996) 2075–2090. [21] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319. [22] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, second ed., Academic Press, London, 1982. [23] D. Dollimore, G.R. Heal, J. Appl. Chem. 14 (1964) 109–114. [24] D. Dollimore, G.R. Heal, J. Colloid Interface Sci. 33 (1970) 508–519. [25] R.K. Iler, The Chemistry of Silica, John Wiley & Sons, New York, 1979. [26] N.N. Vlasova, L.P. Golovkova, The adsorption of amino acids on the surface of highly dispersed silica, Colloid J. 66 (6) (2004) 657–662. [27] S. Munsch, M. Hartmann, S. Ernst, Adsorption and separation of amino acids from aqueous solutions on zeolites, Chem. Commun. 19 (2001) 1978–1979. [28] X.S. Zhao, G.Q. Lu, A.K. Whittaker, G.J. Millar, H.Y. Zhu, Comprehensive study of surface chemistry of MCM-41 using 29Si CP/MAS NMR, FTIR, pyridine-TPD, and TGA, J. Phys. Chem. B 101 (1997) 6525–6531. [29] A. Stein, B.J. Melde, R.C. Schroden, Hybrid inorganic–organic mesoporous silicates—nanoscopic reactors coming of age, Adv. Mater. 12 (19) (2000) 1403–1419.
