6616
Langmuir 2008, 24, 6616-6623
The Local Microenvironment Surrounding Dansyl Molecules Attached to Controlled Pore Glass in Pure and Alcohol-Modified Supercritical Carbon Dioxide Phillip M. Page, Taylor A. McCarty, Chase A. Munson, and Frank V. Bright* Department of Chemistry, Natural Sciences Complex, UniVersity at Buffalo, The State UniVersity of New York, Buffalo, New York 14260-3000 ReceiVed February 17, 2008. ReVised Manuscript ReceiVed April 29, 2008 We report on the local microenvironment surrounding a free dansyl probe, dansyl attached to controlled pore glass (D-CPG), and dansyl molecules attached to trimethylsilyl-capped CPG (capped D-CPG) in pure and alcohol-modified supercritical CO2. These systems were selected to provide insights into the local microenvironment surrounding a reactive agent immobilized at a silica surface in contact with pure and cosolvent-modified supercritical CO2. Local surface-bound dansyl molecule solvation on the CPG surface depends on the dansyl molecule surface loading, the surface chemistry (uncapped versus capped), the bulk fluid density, and the alcohol gas phase absolute acidity. At high dansyl loadings, the surface-bound dansyl molecules are largely “solvated” by other dansyl molecules and these molecules are not affected significantly by the fluid phase. When the dansyl surface loading decreases, dansyl molecules can be accessed/solvated/wetted by the fluid phase. However, at the lowest dansyl loadings studied, the dansyl molecules are in a fluid inaccessible/restrictive environment and do not sense the fluid phase to any significant degree. In uncapped D-CPG, one can poise the system such that the local concentration of an environmentally less responsible cosolvent (alcohol) in the immediate vicinity of surface-immobilized dansyl molecules can approach 100% even though the bulk solution contains orders of magnitude less of this less environmentally responsible cosolvent. In capped C-CPG, the surface excess is attenuated in comparison to that of uncapped D-CPG. The extent of this cosolvent surface excess is discussed in terms of the dansyl surface loading, the local density fluctuations, the cosolvent and surface silanol gas phase acidities, and the silica surface chemistry. These results also have implications for cleanings, extractions, heterogeneous reactions, separations, and nanomaterial fabrication using supercritical fluids.
Introduction With the growing emphasis on the environment, researchers in academia, the government, and industry have been exploring several environmentally responsible solvent systems. Agreements (e.g., Montreal and Kyoto Protocols) and major awards (e.g., the 2007 Nobel Prize) have served to heighten awareness of our environment and point to the economic advantages of more environmentally benign strategies. In this context, many of the more common liquid organic solvents are not particularly environment friendly, they can be expensive to purchase, and proper recycling can necessitate complicated/expensive protocols. Thus, considerable effort has been directed toward developing more environmentally responsible strategies that might eliminate or limit certain liquid organic solvent use. Supercritical CO2 (scCO2) represents an attractive alternative to conventional liquid organic solvents because it is nontoxic, nonflammable, and relatively inexpensive, it can be easily separated from a system by a decompression step, and it can be recycled. The physicochemical properties of scCO2 (density, dielectric constant, viscosity, diffusivity, and refractive index) are intermediate between a gas and liquid, and a supercritical fluid’s properties can be continuously tuned by adjusting the system pressure and/or temperature.1 In addition, the critical pressure and temperature (Pc ) 73.8 bar and Tc ) 304.3 K) for CO2 are readily created in the laboratory or plant. * To whom correspondence should be addressed. Voice: (716) 645-6800 ext. 2162. Fax: (716) 645-6963. E-mail:
[email protected]. (1) McHugh, M. A.; Krukonis, V. J. Supercritical Fluids Extraction: Principles and Practice, 2nd ed.; Butterworth-Heinemann: Boston, 1994.
ScCO2 has been used extensively for extractions, heterogeneous catalysis, and separations.2 ScCO2 has also been used for polymer film swelling,3 nanosynthesis,4 microelectronic processing,5 deposition,6 engineered powder formulations for drug delivery,7 device fabrication,8 polymer impregnation,9 magnetic media production,10 metallic film deposition strategies,11 surfacepromoted reactions,12 and surface condensation and grafting13 among other applications and uses. An interface and the adsorption (2) (a) Taylor, L. T. Supercritical Fluid Extraction; John Wiley & Sons: New York, 1996. (b) Smith, R. M.; Hawthorne, S. B. Supercritical Fluids in Chromatography and Extraction; Elsevier, Oxford, 1997. (c) Jessop, P. G.; Leitner, W. Chemical Synthesis Using Supercritical Fluids; Wiley-VCH: Germany, 1999. (d) Clifford, T. Fundamentals of Supercritical Fluids; Oxford University Press: New York, 1999. (e) Zougagh, M.; Valcarcel, M.; Rios, TrAC, Trends Anal. Chem. 2004, 23, 399–405. (f) Henry, M. C.; Yonker, C. R. Anal. Chem. 2006, 78, 3909–3915. (3) (a) Kazarian, S. G.; Vincent, M. F.; Bright, F. V.; Liotta, C. L.; Eckert, C. A. J. Am. Chem. Soc. 1996, 118, 1729–1736. (b) Li, Y.; Park, E. J.; Lim, K. T.; Johnston, K. P.; Green, P. F. J. Polym. Sci., Part B: Polym. Phys. 2007, 45, 1313–1324. (4) Dahl, J. A.; Maddux, B. L. S.; Hutchison, J. E. Chem. ReV. 2007, 107, 2228–2269. (5) XiaoGang, Z.; Johnston, K. P. Chin. Sci. Bull. 2007, 52, 27–33. (6) Zhang, Y.; Erkey, C. J. Supercrit. Fluids 2006, 38, 252–267. (7) Chan, H.-K. Colloids Surf., A 2006, 284-285, 50–55. (8) O’Neil, A.; Watkins, J. J. MRS Bull. 2005, 30, 967–975. (9) (a) Kikic, I.; Vecchione, F. Curr. Opin. Solid State Mater. Sci. 2003, 7, 399–405. (b) Fleming, O. S.; Kazarian, S. G. In Supercritical Carbon Dioxide; Kennere, M. F., Meyer, T., Eds.; Wiley-VCH: Weinheim, Germany, 2005; pp 205-238. (10) Johns, K. Tribol. Int. 1998, 31, 485–90. (11) Cabanas, A.; Shan, X.; Watkins, J. J. Chem. Mater. 2003, 15, 2910–2916. (12) Weinstein, R. D.; Renslo, A. R.; Danheiser, R. L.; Tester, J. L. J. Phys. Chem. B 1999, 103, 2878–2887. (13) (a) O’Neil, A. S.; Mokaya, R.; Poliakoff, M. J. Am. Chem. Soc. 2002, 124, 10636–10637. (b) Xie, B.; Muscat, A. J. IEEE Trans. Semicond. Manuf. 2004, 17, 544–553. (c) Wang, Z.-W.; Wang, T.-J.; Wang, Z.-W.; Jin, Y. Powder Technol. 2004, 139, 148–155. (d) Domingo, C.; Loste, E.; Fraile, J. J. Supercrit. Fluid 2006, 37, 72–86.
10.1021/la8005184 CCC: $40.75 2008 American Chemical Society Published on Web 06/07/2008
MicroenVironment Surrounding Dansyl Molecules
Langmuir, Vol. 24, No. 13, 2008 6617
Figure 1. Dansyl-based subsystems that were used in this research. See text for abbreviations.
and desorption of major/minor components from the fluid phase onto an interface play an important role in each of the aforementioned areas. Researchers have previously reported on the adsorption equilibria of scCO2 and solutes dissolved in scCO2 onto solid supports. For example, Strubinger and Parcher determined the surface excess of pure CO2 on C18-silica,14 Morbidelli and coworkers measured pure CO2 adsorption isotherms on silica,15 and Humayun and Tomasko determined the adsorption of pure CO2 on activated carbon.16 In all cases, in proximity to the fluid critical density, there is an excess of CO2 at the interface in comparison to the bulk fluid density. This behavior has been rationalized by the long-range CO2 density fluctuations affecting the Gibbs CO2 adsorption. There have also been related studies on the equilibrium adsorption of cosolvents dissolved in scCO2 (e.g., acetone,17 benzene,17 ethylacetate,18 p-dichlorobenzene,19 and toluene19) onto silica surfaces. These experiments showed that there is an increase in the concentration of adsorbed cosolvent in proximity to the mixture critical density. These results have been explained in terms of density fluctuations, a competition between CO2 and cosolvent adsorption, and the cosolvent ability to interact with the silica surface via, for example, hydrogen bonds and/or van der Waals interactions. Controlled pore glass (CPG) is a Si-based solid support that is commonly used as a chromatographic stationary phase20 and as a platform for immobilizing a wide variety of active agents.21 The CPG surface consists of siloxanes and geminal, isolated, and vicinal or hydrogen-bonded silanols.22 The silanol groups provide convenient attachment sites for the introduction of functional groups onto the CPG surface (e.g., amine groups) and (14) Strubinger, J. R.; Parcher, J. F. Anal. Chem. 1989, 61, 951–955. (15) Di Giovanni, O.; Do¨rfler, W.; Mazzotti, M.; Morbidelli, M. Langmuir 2001, 17, 4316–4321. (16) Humayun, R.; Tomasko, D. L. AIChE J. 2000, 46, 2065–2075. (17) Jin, D. W.; Nitta, T.; Park, D. W. Bull. Chem. Soc. Jpn. 1997, 70, 2987– 2993. (18) Lochmuller, C. H.; Mink, L. P. J. Chromatogr. 1987, 409, 55–62. (19) Yang, X.; Matthews, M. A. Chem. Eng. J. 2003, 93, 163–172. (20) Schnabel, R.; Langer, P. J. Chromatogr. 1991, 544, 137–146. (21) (a) Pluym, B.; Slegers, G.; Claeys, A. Enzyme Microb. Technol. 1988, 10, 656–659. (b) Go¨bo¨lo¨s, S.; Ta´las, E.; Hegedu¨s, M.; Berto´ti, I.; Margitfalvi, J. L. Appl. Catal., A 1997, 152, 63–68. (c) Heckel, A.; Seebach, D. Chem.sEur. J. 2002, 8, 559–572. (22) Unger, K. K. Porous Silica; Elsevier Scientific Publishing Co.: New York, 1979.
Figure 2. Schematic representation and abbreviation notation illustrating the solvation of D-CPG (panels A-C) and capped D-CPG (panel D) model interfaces at various dansyl surface loadings (D/A).25 The colorization around the dansyl molecules represents the solvent responsivity of the CPG-bound dansyl moieties. Red depicts residues/ domains that are the least solvent sensitive. Green denotes residues/ domains that are the most solvent sensitive. The behaviors exhibited by capped D-CPG at D/A ) 0.3, 0.6, and 10-5 are comparable to those of the uncapped D-CPG cases provided to the left and thus are not repeated.
are mostly responsible for the interactions between the CPG surface and potential adsorbates.20–23 Fluorescence spectroscopy has been used to investigate silica interfaces by using covalently attached or physiosorbed probe molecules.24 Previously, we reported on the effects of surface loading, solvent type (ionic, molecular liquids) and polarity, and surface residue end capping on the local microenvironment surrounding a solvent sensitive fluorescent probe molecule when it is covalently attached to the CPG surface.25 These results revealed the important role surface loading and encapping plays on the ability of surface-immobilized species to interact with molecular and ionic liquids. In the current research, we report on the behavior of dansylpropylsulfonamide (DPSA) (the free probe), dansylated aminopropyl controlled pore glass (D-CPG), and D-CPG that has been end capped with trimethylchlorosilane (capped D-CPG) (Figure 1) when they are in contact with (i) pure scCO2 at a reduced temperature (Tr ) Texp/Tc) of 1.02 between a reduced density (Fr ) Fexp/Fc) of 0 and 1.90 and (ii) binary mixtures of (23) (a) Tripp, C. P.; Combes, J. R. Langmuir 1998, 14, 7348–7352. (b) Granqvist, B.; Sandberg, T.; Hotokka, M. J. Colloid Interface Sci. 2007, 310, 369–376. (24) (a) Lochmu¨ller, C. H.; Marshall, D. B.; Wilder, D. R. Anal. Chim. Acta 1981, 130, 31–43. (b) Wong, A. L.; Hunnicutt, M. L.; Harris, J. M. J. Phys. Chem. 1991, 95, 4489–4495. (c) Nigam, S.; Rutan, S. Appl. Spectrosc. 2001, 55, 362A370A.
6618 Langmuir, Vol. 24, No. 13, 2008
CO2 that contain 3 mol % methanol, ethanol, isopropyl alcohol, 2,2,2-trifluoroethanol, or phenol at Tr ) 1.04 between Fr ) 0 and Fr ) 1.40. These particular alcohols were selected initially because they span a range of pKa (aqueous) values (ca. 10-17).26 The DPSA experiments serve as a benchmark to estimate how the probe molecule itself is solvated intrinsically in the absence of a CPG surface (uncapped or capped). The D-CPG experiments were carried out to determine how pure CO2 and alcohol-modified CO2 modulate the local microenvironment that surrounds the surface-bound dansyl molecules (i.e., the cybotactic region). Together, these experiments provide insights into the molecularlevel solvation of these at silica surfaces in contact with supercritical fluids. Dansyl was selected as our fluorescent probe molecule because its emission spectrum is solvatochromic and it is straightforward to label the aminopropyl-CPG surface amines with dansyl chloride. Thus, changes in the physicochemical properties surrounding the CPG-bound dansyl probe are reflected in the dansyl molecule’s emission spectrum. Figure 2 summarizes how the dansyl-to-surface amine molar ratio (D/A) influences the solvation of the dansyl molecule on uncapped and capped D-CPG.25 The colorization around the dansyl molecules represents the solvent responsivity of the CPG-bound dansyl molecules. Red depicts residues/domains that are the least solvent sensitive. Green denotes residues/domains that are the most solvent sensitive.
Theory Consider a solvatochromic fluorescent probe molecule dissolved at low concentration in pure nonpolar and polar miscible solvents, and a binary mixture of the two solvents. To a first approximation, the Lippert-Mataga expression24,25,27,28 predicts that the Stokes shift in the pure nonpolar solvent will be smaller in comparison to the Stokes shift (SS ) υabs - υem, in cm-1) observed in the pure polar solvent and the mixture Stokes shift will be between the two pure Stokes shift values. Assuming a linear relationship between the observed Stokes shift and the mole fraction of polar and nonpolar solvent molecules that surround the solvatochromic fluorophore,28 one can use the experimentally measured Stokes shifts to estimate the local mol % of the polar cosolvent (LMPC) surrounding the solvatochromic fluorescent probe molecule in any binary mixture:
LMPC ) [(SSpolar - SSmixture) ⁄ ((SSpolar - SSmixture) + (SSmixture - SSnonpolar))] (1) In this expression, the subscripts “nonpolar”, “polar”, and “mixture” represent the Stokes shift (in cm-1) reported by the fluorescent probe molecule in the pure nonpolar, pure polar, and binary mixture, respectively. If we extend this scenario to the situation of density-dependent Stokes shifts like one would anticipate in a supercritical fluid system,29 one can estimate the percentage of the polar component (alcohol in the current experiments) that surrounds the fluorescent (25) (a) Page, P. M.; Munson, C. A.; Bright, F. V. Langmuir 2004, 20, 10507– 16. (b) Page, P. M.; McCarty, T. M.; Baker, G. A.; Baker, S. N.; Bright, F. V. Langmuir 2007, 23, 843–849. (26) CRC Handbook of Chemistry and Physics, 74th Ed.; Chemical Rubber Publishing Co.: Boca Raton, FL, 1993-1994. (27) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.; Plenum Press: New York, 1999; Chapters 6 and 7. (28) (a) Suppan, P. J. J. Chem. Soc., Faraday Trans. 1 1987, 83, 495–509. (b) Suppan, P.; Ghoneim, N. SolVatochromism; Royal Society of Chemistry: Cambridge, UK, 1997. The assumption of linearity has less of an impact in the current research because we ultimately compare the free probe and CPG-bound probe to one another when arriving at an estimate of the effects of the interface on the local composition of the cosolvent that surrounds the surface immobilized probe molecules.
Page et al.
Figure 3. Hypothetical density-dependent Stoke’s shifts for solvatochromic surface-immobilized probe molecules in pure scCO2, pure ROH, and cosolvent-modified CO2 (CO2 + ROH). The estimated ROH % surrounding the surface-immobilized solute is also given.
probe molecule (Figure 3). Finally, by extending this concept to a surface-attached fluorescent probe molecule like in the current research (Figure 1) and by using an appropriate control experiment (i.e., compare the free probe molecule fluorescence to the fluorescence from the surface-attached probe), one can estimate the density-dependent local composition of polar solvent component that surrounds the surface-immobilized probe molecule and explore the role surface and polar component chemistry play in determining the local microenvironment surrounding the surface-attached probe molecule.
Experimental Section Chemicals and Reagents. Aminopropyl-CPG was purchased from CPG, Inc. (now Millipore). The vendor reports the following characteristics: particle size, 34-74 µm; mean pore diameter, 182 Å; surface area, 113 m2/g; and surface amine concentration, 701.9 µmol/g. Dansyl chloride (98%) was purchased from Aldrich and used without further purification. All alcohol cosolvents were from Sigma-Aldrich (HPLC grade, 99.8+%). CO2 (SFC grade) was obtained from Scott Specialty Gases. Binary mixtures of CO2 and a particular alcohol were prepared within a home-built stainless steel high-pressure bomb and then transferred into the high-pressure syringe pump. Dansylpropylsulfonamide (DPSA), Dansyl-Labeled Controlled Pore Glass (D-CPG), and Capped D-CPG Synthesis. The syntheses are described in ref 25a. Dansyl Surface Loading. The protocol for determining the immobilized dansyl-to-surface amine concentration ratio (D/A) on these D-CPG samples is described in ref 25a. Measurement System. Figure 4 presents a simplified schematic of the high-pressure spectroscopic system used in this research.30 The CPG samples are loaded into 1.5 mm quartz capillaries (Technical Glass, Inc.). For the DPSA experiments, the capillary is removed and DPSA is added directly into the high-pressure optical cell. All fluorescence spectra were measured by using an SLM-AMINCO model 8100 instrument. The excitation wavelength was 342 nm. The excitation and emission spectral bandpasses were maintained at 2 and 4 nm, respectively, for all experiments. All emission spectra were background corrected by using appropriate blanks. The blank contribution was always
