Alkyl Chain Conformation of Octadecylsilane Stationary Phases by

Department of Chemistry, University of Arizona, Tucson, Arizona 85721 ... liquidlike disordered state characteristic of these station- ary phases at r...
0 downloads 0 Views 85KB Size
Anal. Chem. 1998, 70, 4915-4920

Alkyl Chain Conformation of Octadecylsilane Stationary Phases by Raman Spectroscopy. 1. Temperature Dependence Mankit Ho† and Jeanne E. Pemberton*

Department of Chemistry, University of Arizona, Tucson, Arizona 85721

Raman spectroscopy is used for the first time to probe the effect of temperature on the conformational order of polymeric and monomeric octadecylsilane stationary phases. Spectral data in the ν(C-C) and ν(C-H) regions are interpreted in terms of alkyl chain conformational state and its dependence on temperature. In contrast to the liquidlike disordered state characteristic of these stationary phases at room temperature, at liquid N2 temperatures, the alkyl chains exist in a more ordered state with a residual level of gauche conformational defects. Systematic studies between -15 and 95 °C reveal more subtle changes in conformational order as ascertained from empirical spectral indicators including the intensity ratios I[νa(CH2)]/I[νs(CH2)] and I[ν(C-C)T]/I[ν(C-C)G]. Plots of these ratios as a function of temperature reveal two distinct regimes of behavior. By extrapolating the linear regions of these plots, a surface “phase transition” temperature of ∼20 °C for both surface-confined octadecylsilane stationary phases is estimated that represents subtle changes in alkyl chain conformational order from a more ordered phase to a slightly more disordered phase. The similarity in behavior between the polymeric and monomeric octadecylsilane stationary phases is interpreted as evidence for similar interchain spacing of the alkylsilanes on these silica surfaces. Chemically modified silicas are popular stationary phases for reversed-phase liquid chromatography (RPLC) and solid-phase extraction (SPE).1-3 These stationary phases are typically formed by covalent attachment of alkylchlorosilanes to a silica gel support using mono-, di-, or trichlorosilanes.3 The properties of these stationary phases play an important role in governing the selectivity and efficiency of chromatographic separations.2-5 To elucidate the retention mechanisms for such separations, an understanding of the stationary-phase behavior under various separations conditions is essential. A variety of experimental techniques including Raman spectroscopy,6,7 fluorescence spectroscopy,8-14 NMR,15-20 † Present address: National Institute for Occupational Safety and Health, 4676 Columbia Parkway, MS-R7 Cincinnati, OH 45226. (1) Berendsen, G. E.; DeGalan, L. J. Liq. Chromatogr. 1978, 1, 561-586. (2) Dorsey, J. G.; Cooper, W. T. Anal. Chem. 1994, 66, 857A-867A. (3) Dorsey, J. G.; Dill, K. A. Chem. Rev. 1989, 89, 331-346. (4) Tchapla, A.; Heron, S.; Lesellier, E. J. Chromatogr., A 1993, 656, 81-112. (5) Schunk, T. C.; Burke, M. F. J. Chromatogr., A 1993, 656, 289-316. (6) Ho, M.; Cai, M.; Pemberton, J. E. Anal. Chem. 1997, 69, 2613-2616.

10.1021/ac980471s CCC: $15.00 Published on Web 10/28/1998

© 1998 American Chemical Society

IR spectroscopy,21-24 and chromatographic methods25-28 have been used to study these stationary phases. Despite the valuable information provided by these studies, a complete understanding of retention has not yet been achieved. This deficiency is due in part to a lack of understanding of the intermolecular interactions between the alkyl moieties of the stationary phase, the solute, and the mobile phase in the separation process. This information can best be provided by investigating the stationary phase using techniques that are sensitive to alkyl chain conformation. Raman spectroscopy is a powerful tool for the characterization of conformational changes in alkyl chains.29-31 In contrast to most previous methods used to study these alkylsilane stationary phases (in which conformational state of the alkane is indirectly inferred from the data), Raman spectroscopy is particularly useful due to the direct measurement of gauche and trans conformers in the alkane. Furthermore, these measurements are relatively free from (7) Doyle C. A.; Vickers, T. J.; Mann, C. K.; Dorsey, J. G. J. Chromatogr., A 1997, 779, 91-112. (8) Lochmuller, C. H.; Colborn, A. S.; Hunnicutt, M. L.; Harris, J. M. J. Am. Chem. Soc. 1984, 106, 4077-4082. (9) Lochmuller, C. H.; Marshall, D. B.; Wilder, D. R. Anal. Chim. Acta 1981, 130, 31-34. (10) Carr, J. M.; Harris, J. M. Anal. Chem. 1987, 59, 2546-2550. (11) Montgomery, M. E.; Green, M. A.; Wirth, M. J. Anal. Chem. 1992, 64, 1170-1175. (12) Burbage, J. D.; Wirth, M. J. J. Phys. Chem. 1992, 96, 5943-5948. (13) Montgomery, M. E.; Wirth, M. J. Anal. Chem. 1994, 66, 680-684. (14) Zull, S. L.; Kovaleski, J. M.; Zhu, X. R.; Harris, J. M.; Wirth, M. J. Anal. Chem. 1994, 66, 1708-1712. (15) Sindorf, D. W.; Maciel, G. E. J. Am. Chem. Soc. 1983, 105, 1848-1851. (16) Kelusky, E. C.; Fyfe, C. A. J. Am. Chem. Soc. 1986, 108, 1746-1749. (17) Shah, P.; Rogers, L. B.; Fetaer, J. C. J. Chromatogr. 1987, 388, 411-419. (18) Bayer, E.; Paulus, A.; Peters, B.; Laupp, G.; Reiners, J.; Albert, K. J. Chromatogr. 1986, 364, 25-37. (19) Albert, K.; Bayer, E. J. Chromatogr. 1991, 544, 345-370. (20) Albert, K.; Brindle, R.; Martin, P.; Wilson, I. D. J. Chromatogr., A 1994, 665, 253-258. (21) Sander, L. C.; Callis, J. B.; Field, L. R. Anal. Chem. 1983, 55, 1068-1075. (22) Ohtake, T.; Mino, N.; Ogawa, K. Langmuir 1992, 8, 2081-2083. (23) Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1120-1126. (24) Jinno, K.; Wu, J.; Ichikawa, M.; Takata, I. Chromatographia 1993, 37, 627634. (25) Gilpin, R. K.; Gangoda, M. E.; Krishen, A. E. J. Chromatogr. Sci. 1982, 20, 345-348. (26) Schunk, T. C.; Burke, M. F. Int. J. Environ. Anal. Chem. 1986, 25, 81103. (27) Cole, L. A.; Dorsey, J. G. Anal. Chem. 1992, 64, 1317-1323. (28) Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292. (29) Bryant, M. A.; Pemberton, J. E. J. Am. Chem. Soc. 1991, 113, 3629-3637. (30) Bryant, M. A.; Pemberton, J. E. J. Am. Chem. Soc. 1991, 113, 8282-8293. (31) Pemberton, J. E.; Bryant, M. A.; Sobocinski, R. L. J. Phys. Chem. 1992, 96, 3776-3782.

Analytical Chemistry, Vol. 70, No. 23, December 1, 1998 4915

spectral interference from the silica or any adsorbed water which plague FT-IR studies of these systems. In a previous report,6 we demonstrated the feasibility of using Raman spectroscopy for the elucidation of conformational information about alkysilane-modified silica-based chromatographic stationary phases. As part of an ongoing project to elucidate retention mechanisms in liquid chromatography, we present here a Raman spectroscopic investigation of the effect of temperature on the conformational state of polymeric and monomeric octadecylsilane stationary-phase alkyl chains. Liquid chromatographic separations are commonly carried out at ambient temperature; temperature control is normally performed to gain greater reproducibility.32 However, several reports33-35 describe attempts to enhance chromatographic selectivity by adjusting column temperature. In the few cases in which this approach was used, improvements in separation selectivity were usually attributed to a change in interfacial structure of the stationary phase. The investigation of temperature effects on the conformational state of stationary phases, particularly the existence of melting-like phase transitions of the alkyl moieties of stationary phases, has been the focus of several reports.32-46 One popular approach is the van’t Hoff analysis. In the early 1980s, Morel and Serpinet36-38 reported the observation of deviations from linearity in van’t Hoff plots for monomeric octadecylsilane stationary phases in the absence of solvent using gas chromatographic techniques. They attributed these deviations to phase transitions of the stationary phases’ alkyl moieties. Van’t Hoff analysis in liquid chromatography is more complicated due to solvation of the stationary phases. Nonlinear van’t Hoff plots in liquid chromatography are usually attributed to an irreversible reorganization of alkyl chain structure due to solvation.39 However, although nonlinear van’t Hoff behavior is indicative of a change in retention mechanism, this change is not necessarily due to structural changes of the bonded phase.32 Differential scanning calorimetry (DSC)41-43 is another technique that has been used to study the effect of temperature on stationary-phase structure. The presence of a distinct endothermic peak in a DSC scan has been suggested as evidence for a phase transition. However, octadecylsilane stationary phases with bonding densities less than 2.5 µmol/m2 failed to show distinct phase transitions in DSC.42 Although DSC provides important thermodynamic information about stationary phases, it does not provide direct information about conformational state of the alkyl chains. Therefore, techniques sensitive to alkyl chain conformation are (32) Wheeler, J. F.; Beck, T. L.; Klatte, S. J.; Cole, L. A.; Dorsey, J. G. J. Chromatogr., A 1993, 656, 317-333. (33) Bell, C. M.; Sander, L. C.; Wise S. A. J. Chromatogr., A 1997, 757, 29-39. (34) Jinno, K.; Lin, Y. Chromtographia 1995, 41, 311. (35) Sander, L. C.; Wise, S. A. Anal. Chem. 1989, 61, 1749-1754. (36) Morel, D.; Serpinet, J. J. Chromatogr. 1980, 200, 95-104. (37) Morel, D.; Serpinet, J. J. Chromatogr. 1981, 214, 202-208. (38) Morel, D.; Serpinet, J. J. Chromatogr. 1982, 248, 231-240. (39) Gilpin, R. K.; Squires, J. A. J. Chromatogr. Sci. 1981, 19, 195-199. (40) Hansen, S. J.; Callis, J. B. J. Chromatogr. Sci. 1983, 21, 560-563. (41) Claudy, P.; Letoffe, J. M.; Gaget, C.; Morel, D.; Serpinet, J. J. Chromatogr. 1985, 329, 331-349. (42) Morel, D.; Tabar, K.; Serpinet, J.; Claudy, P.; Letoffe, J. M. J. Chromatogr. 1987, 395, 73-84. (43) Pursch, M.; Strohschein, S.; Handel, H.; Albert, K. Anal. Chem. 1996, 68, 386-393. (44) Kelusky, E. C.; Fyfe, C. A. J. Am. Chem. Soc. 1986, 108, 1746-1749. (45) Gangoda, M. E.; Gilpin, R. K. J. Magn. Reson. 1983, 53, 140-143. (46) Thompson, W. R.; Pemberton J. E. Anal. Chem. 1994, 66, 3362-3370.

4916 Analytical Chemistry, Vol. 70, No. 23, December 1, 1998

needed for a more detailed investigation of the effect of temperature on stationary-phase structure. Recently, FT-IR spectroscopy21 and cross-polarization magic angle solid-state NMR44-46 were employed to provide a more molecular view of stationary-phase structure. These studies provided important information about mobility and conformational order of alkyl chains at different temperatures and supporting evidence for phase transitions. However, systematic studies on phase transitions or chain reorganization transitions are still lacking. Furthermore, the impact of such a phase transition on retention has not been fully elucidated. In this paper, we begin to address these issues through a systematic Raman spectroscopic study of the dependence of alkyl chain conformational order on temperature for two SPE octadecylsilane stationary phases on silica, Isolute C18 MF (referred to throughout as C18MF) and Isolute C18 TF (referred to throughout as C18TF). C18MF is a monomeric stationary phase formed from dimethyloctadecylchlorosilane, and C18TF is a polymeric stationary phase formed from octadecyltrichlorosilane. C18MF can only form one bond with the surface per molecule, and hence, these alkylsilane molecules have the opportunity to be distributed more uniformly across the surface. In contrast, the trifunctional C18TF system can form multiple bonds with the surface or with other octadecylsilane molecules leading to surfaces that are expected to possess a more heterogeneous distribution of alkylsilanes, possibly as islands. Indeed, differences in bonding of these alkylsilanes on silica surfaces have resulted in differences in chromatographic selectivity for these stationary phase systems.2 Thus, in the work reported here, C18MF and C18TF were both studied in an attempt to elucidate any disparities in the response of conformational order to temperature changes that might be attributable to differences in surface distribution of the alkylsilanes in these systems. EXPERIMENTAL SECTION Instrumentation. Raman spectra were collected using 5070 mW of 514.5-nm radiation from a Coherent Innova 90-5 Ar+ laser on a Spex 1877 Triplemate spectrograph; this system has been described in detail previously.29-31 Slit settings of the Triplemate were 0.5/7.0/0.15 mm for these studies. Laser polarization was parallel to the plane of incidence. The detector in these experiments was a Princeton Instruments charge-coupled device (CCD) system based on a thinned, back-illuminated, antireflection coated RTE-1100-PB CCD of pixel format 1100 × 330 which was cooled with liquid N2 to -90 °C. Samples in air were sealed in NMR tubes to prevent moisture condensation at the colder temperatures. These NMR tubes were held in a copper sample mount through which a temperature control medium (50: 50 water/ethylene glycol) was circulated using a Neslab NTE110 temperature controller. Samples were equilibrated at the appropriate temperature for at least 30 min prior to spectral acquisition based on the observation that the Raman spectral response is relatively unchanged at longer equilibration times. Integration times for each spectrum are provided in the figure captions. Low laser power and short acquisition times were used to avoid unnecessary sample heating by the laser beam. Materials. Octadecyltrichlorosilane (OTS, 95%) was purchased from Aldrich and used as received. The two commercial SPE stationary phases, Isolute C18 MF and Isolute C18 TF, were a gift from International Sorbent Technology. The manufacturer’s

Table 1. Manufacturer’s Specifications for C18MF and C18TF

surface area (m2/g) average pore size (Å) carbon loading (C%) surface coverage (µmol/m2) average particle diam (µm)

Isolute C18 MF (monomeric)

Isolute C18 TF (polymeric)

300 60 13.9 1.9 63

561 54 19.3 1.6 62

Table 2. Raman Peak Frequencies and Assignments for Octadecylsilane Environments Raman spectral bands (cm-1) C18MFa (liq N2 temp)

C18TFb (liq N2 temp)

C18MF (room temp)

C18TF (room temp)

assignment

2933 shd 2886 sh 2878 2847 1462 1441 1297

2962 2930 2900 2887 sh 2854 1454 1440 1301

2962 2931 2888 sh 2852 1454 1441 1303

νa(CH3)c νs(CH2)FRh νa(CH3) νa(CH2) νs(CH2) CH2 scissoring CH2 scissoring CH2 twistT,Ge,f

1124

1125

1078 1065

1081 1065

ν(C-C)T CH2 twistT ν(C-C)G ν(C-C)T CH3 rkTg

2961 2901 2880 2848 1456 1438 1295 1170 1130 1097 1080 1062

1130 1098 1063 890

a C18MF: Isolute C18 MF, monomeric octadecylsilane stationary phase. b C18TF: Isolute C18 TF, polymeric octadecylsilane stationary phase. c ν, stretch. d sh, shoulder. e T, trans. fG, gauche. g rk, rock.h FR, Fermi resonance.

RESULTS AND DISCUSSION Initial attempts to assess the effect of temperature on the conformational order of Isolute octadecylsilane alkyl chains began with a simple comparison of these stationary phases at room and liquid N2 temperatures. Figure 1 presents Raman spectra in the ν(C-C) region between 800 and 1600 cm-1 and the ν(C-H) region between 2700 and 3100 cm-1 for C18MF and C18TF at these two temperatures. Peak frequencies and band assignments for these octadecylsilane stationary-phase materials are given in Table 2. Assignments of the Raman spectra of these systems are based on those previously reported for similar alkyl chain systems.46,47 As discussed in earlier reports,28-30 Raman spectra can provide detailed information about alkyl chain conformation. In the ν(C-H) region, although the observed spectral envelope encompasses as many as 11 vibrational modes, 4 prominent bands can be clearly identified through their peak maximums at ∼2852, 2888, 2931, and 2962 cm-1. These are assigned to the νs(CH2), νa(CH2),

νs(CH2)FR, and νa(CH3) modes, respectively. Previous studies47,48 have shown that the peak intensity ratio of the antisymmetric (at ∼2886 cm-1) to symmetric (at ∼2852 cm-1) ν(CH2) bands is an excellent empirical indicator of alkyl chain conformational order. In the liquid state, for example, the intensity ratio of these methylene bands is ∼0.6-0.9, whereas in the crystalline solid, this ratio increases to ∼1.6-2.0. The peak frequencies of these ν(CH2) modes are also sensitive to the degree of alkyl chain order.48,49 The νs(CH2) and νa(CH2) modes are observed at ∼2856 and ∼2888 cm-1, respectively, for alkyl chains in the liquid state; these bands both shift to lower frequency by about 6 and 8 cm-1, respectively, for alkyl chains in a crystalline solid. The bands of interest in the ν(C-C) region are the CH3 rock of trans conformers at 890 cm-1, the ν(C-C) of trans conformers at 1063 cm-1, the ν(C-C) of gauche conformers at 1080 cm-1, and the CH2 twist modes for gauche and trans conformers which appear as overlapped bands at ∼1300 cm-1. Several spectral indicators are available with which to assess the degree of conformational order of the alkyl chains in this region including the relative intensity of the ν(C-C) bands associated with gauche and trans conformers and the peak shape and peak position of the combined gauche-trans CH2 twist mode. This latter indicator is based on the observed shift of the combined gauche-trans CH2 twist mode to lower frequency and its concomitant decrease in width and increase in symmetry as the degree of conformational order of the alkyl chain increases. As reported in our earlier paper,6 Raman spectra of both the monomeric (C18MF) and polymeric (C18TF) octadecylsilane stationary phases indicate that the alkyl chains of both systems, despite the potential for substantially different organization at the silica substrate, adopt a relatively disordered, liquidlike conformational state at room temperature. The disordered nature of these alkyl chains is evident by the relatively low value of the

(47) Lin-Vien, J. G.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. In The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press: New York, 1991.

(48) Wallach, D. F. H.; Varma, S. P.; Fookson, J. Biochim. Biophys. Acta 1979, 559, 153-208. (49) Gaber, B.; Peticolas, W. L. Biochim. Biophys. Acta 1977, 465, 260-274.

Figure 1. Raman spectra in the ν(C-C) (left panel) and ν(C-H) (right panel) regions for (a) C18MF at room temperature, (b) C18MF at liquid N2 temperature, (c) C18TF at room temperature, and (d) C18TF at liquid N2 temperature. Integration times: (a) 20, (b) 20, (c) 5, and (d) 5 min in the ν(C-C) region; (a) 5 min, (b) 1 min, (c) 30 s, and (d) 30 s in the ν(C-H) region.

specifications of these materials are given in Table 1. Both packing materials have similar average particle sizes (∼60 µm), average pore sizes (∼60 Å), and alkyl chain surface coverages (1.9 µmol/m2 for C18MF and 1.6 µmol/m2 for C18TF). These materials were used as received.

Analytical Chemistry, Vol. 70, No. 23, December 1, 1998

4917

I[νa(CH2)]/I[νs(CH2)] ratio (∼1.0) in the ν(C-H) region, the relatively large ν(C-C)G intensity at ∼1080 cm-1, and the breadth and asymmetry of the CH2 twistT,G at 1302 cm-1 (gauche and trans denoted by subscripts G and T, respectively). The similarity in spectral response of the monofunctional and trifunctional alkylsilane systems is noteworthy and suggests that the alkane chains of each system reside in conformationally similar chemical environments. The significance of this observation is discussed further below. Upon cooling to liquid N2 temperatures, the alkyl chains of both C18MF and C18TF adopt a more ordered conformation. This transformation is evidenced by an increase in intensity of the two ν(C-C)T bands at ∼1062 and ∼1130 cm-1 and a concomitant decrease in bands associated with gauche conformations, such as the ν(C-C)G at 1080 cm-1. In addition, the CH2 twist at ∼1300 cm-1 increases in intensity, shifts 6 cm-1 to lower frequency, and becomes narrower and more symmetric reflecting a predominantly trans conformational environment. Finally, the I[νa(CH2)]/I[νs(CH2)] value increases to ∼1.4. Although this value is not as high as observed in pure crystalline solids, it is considerably higher than observed at room temperature as would be expected in an alkane environment of greater conformational order. Despite the high degree of conformational order expected at liquid N2 temperatures, a small number of gauche conformers still exist within the alkyl chains as indicated by the presence of a small ν(C-C)G band at ∼1080 cm-1. Given the small surface coverage of the alkyl chains in both systems and the fact that one end of each molecule is tethered to the surface, the presence of gauche conformers, even at these temperatures, is not surprising. Nonetheless, these spectral data collectively indicate that the silica surface-bound alkyl chains of both the polymeric and monomeric octadecylsilane stationary phases are proximate enough to engage in significant chain-chain van der Waals interactions leading to a relatively, although not perfectly, ordered state at liquid N2 temperature. Such intermolecular interactions are not surprising for the trifunctional polymeric system given its propensity for cross-linking. However, the partial surface crystallization of the monomeric dimethyloctadecylsilane indicates that these chains, despite their low surface coverage equivalent to an average of ∼1 octadecyl chain/100 Å2, bind to the surface close enough to each other to engage in collective chain phenomena. This observation clearly supports the notion of island formation of these dimethyloctadecyl species even in the absence of crosslinking. Furthermore, these observations imply flexibility of the alkyl chains in their surface-bound environment resulting in a substantial effect of temperature on the overall interfacial structure. A more detailed investigation of the temperature-dependent behavior of C18TF is accomplished by systematically varying the temperature from -15 to +95 °C. These studies were begun by investigating the effect of temperature on neat OTS, a precursor for the C18TF stationary-phase material. Figure 2 shows Raman spectra for OTS over this temperature range. Two distinct alkyl chain conformational states are indicated by the spectra. At low temperatures (e.g. -15 and +5 °C), ordered crystalline alkyl chains exist as suggested by the large value (∼1.8) of I[νa(CH2)]/ I[νs(CH2)], the relatively large intensity of the two ν(C-C)T bands at ∼1062 and ∼1130 cm-1, the complete absence of a ν(C-C)G band at 1080 cm-1, and a narrow, symmetric CH2 twistT at ∼1298 4918 Analytical Chemistry, Vol. 70, No. 23, December 1, 1998

Figure 2. Raman spectra in the ν(C-C) (left panel) and ν(C-H) (right panel) regions for OTS at temperatures of (a) -15, (b) +5, (c) +35, (d) +65, and (e) 95 °C. Integration times: 1 min for spectra in the ν(C-C) region and 5 s for spectra in the ν(C-H) region.

Figure 3. Plots of (a) I[νa(CH2)]/I[νs(CH2)] and (b) I[ν(C-C)T]/I[ν(C-C)G] as a function of temperature for OTS.

cm-1. At higher temperatures, the alkyl chains of OTS exist in an increasingly disordered state as shown by a decrease in the I[νa(CH2)]/I[νs(CH2)] value and an increase in relative intensity of the ν(C-C)G band. A plot of I[νa(CH2)]/I[νs(CH2)] as a function of temperature is shown in Figure 3a and reveals a sharp and clear transition from the crystalline to the liquid state at ∼10 °C. Similarly, the relative intensity changes of the ν(C-C)G and ν(C-C)T bands in Figure 3b also show the phase transition of neat OTS. Similar controlled-temperature experiments were performed on C18TF in air. Figure 4 presents representative Raman spectra of C18TF over the temperature range between -15 and +95 °C. In contrast to the behavior of pure OTS, these spectra indicate that the alkyl chains exist in a largely disordered state throughout this entire temperature range based on the indicators described above. However, a more careful examination of the spectra reveals quantitative temperature-dependent differences. For example, at -15 °C, the larger I[νa(CH2)]/I[νs(CH2)] value and the larger

Figure 4. Raman spectra in the ν(C-C) (left panel) and ν(C-C) (right panel) regions for C18TF at temperatures of (a) -15, (b) +5, (c) +35, (d) +65, and (e) +85 °C. Integration times: 5 min for spectra in the ν(C-C) region and 30 s for spectra in the ν(C-H) region.

Figure 5. Plots of (a) I[νa(CH2)]/I[νs(CH2)] and (b) I[ν(C-C)T]/I[ν(C-C)G] as a function of temperature for C18TF.

intensity of the ν(C-C)T band at 1064 cm-1 relative to the ν(CC)G band at 1080 cm-1 are consistent with a greater degree of alkyl chain order compared with these features at 95 °C. A plot of I[νa(CH2)]/I[νs(CH2)] as a function of temperature for C18TF is shown in Figure 5a. Although the range of values over which this parameter varies is relatively small for this temperature regime (total change of ∼0.07), this plot clearly shows in more detail that C18TF conformational changes do occur. Although the changes observed in these values are much smaller than those of pure OTS, two regions of distinct temperature response can be identified in the plot. The first region is observed for temperatures greater than ∼20 °C where relatively small changes in alkyl chain conformation with temperature occur. Below 15 °C, the alkyl chains appear to be more sensitive to temperature as indicated by the steeper slope of the values. This two-region behavior is attributed to a surface phase transition of C18TF on the stationary-phase support similar to that observed for pure OTS at low temperatures. By extrapolating the two

Figure 6. Plot of I[ν(C-C)T]/I[ν(C-C)G] as a function of temperature for C18MF.

approximately linear regions (Figure 5a, dashed lines), a surface phase transition temperature of ∼20 °C is estimated. As expected, the plot of I[ν(C-C)T]/I[ν(C-C)G] as a function of temperature in Figure 5b also reveals the occurrence of this phase transition. The data suggest that, in contrast to the sharp transition observed for bulk OTS, the transition for C18TF occurs over a much broader temperature range. This behavior is rationalized by the fact that these alkyl moieties are covalently bonded to the silica surface and, thus, much more difficult to crystallize with neighboring alkyl groups. Furthermore, this broad phase transition is consistent with the heterogeneous nature of this chemically modified silica surface. The chains eventually approach a crystalline state at much lower temperatures (liquid N2) with far fewer gauche defects than exist in the temperature range encompassed by the data in Figure 5a. The surface phase transition temperature of ∼20 °C observed here is in good agreement with the gas chromatographic van’t Hoff analysis results of Morel and Serpinet36 and DSC results on monomeric octadecylsilane stationary phases.41,42 As discussed above, neither van’t Hoff analysis nor DSC is a direct indicator of conformational changes in alkyl chains. However, the Raman spectra reported here provide very direct evidence for temperature-dependent conformational changes of these stationary-phase alkyl chains. Significantly, phase transitions for stationary phases with such low coverages (1.6 µm/m2) have not been previously observed; the results reported here demonstrate the extreme sensitivity of Raman spectroscopy to very small changes in conformational state of such stationary-phase materials that are likely to be those that dictate changes in chromatographic efficiency. A similar set of experiments was performed for C18MF. For this system, spectral interference from the νs(CH3) mode at ∼2900 cm-1, attributed to the two methyl groups bound to Si in C18MF, precludes conformational analysis using the value of I[νa(CH2)]/ I[νs(CH2)], especially for relatively disordered alkyl chains. Therefore, the relative intensity of the ν(C-C)G and ν(C-C)T bands was used to assess the effect of temperature on conformational order of these systems. Figure 6 shows a plot of I[ν(CC)T]/I[ν(C-C)G] as a function of temperature for C18MF. C18MF exhibits an almost identical response to temperature as C18TF with a broad surface phase transition of the chains evident at ∼20 Analytical Chemistry, Vol. 70, No. 23, December 1, 1998

4919

°C. The observation of a surface phase transition for C18MF, especially one that is essentially identical to that of C18TF, is significant in that it signals a degree of chain-chain cooperativity and confirms the proximate nature of the alkyl chains postulated above for this system. The conclusion one draws from these observations is that the octadecyl chains of C18MF, despite the presence of the two methyl groups that prohibit intermolecular cross-linking, bind to the underlying silica surface close enough to each other for their alkyl chains to interact to essentially the same degree as in the cross-linked C18TF system. Thus, van der Waals interactions between the chains must play a substantial role in controlling the intermolecular spacing during the reaction chemistry which results in covalent attachment of the alkylsilanes in both C18MF and C18TF. In summary, the results presented here demonstrate the ability of the alkyl moieties of silica-based stationary-phase materials to attain different degrees of conformational order in response to temperature perturbations. However, in contrast to the behavior of the bulk alkylsilanes, the phase transition of the surface-confined

4920 Analytical Chemistry, Vol. 70, No. 23, December 1, 1998

alkylsilanes occurs over a much broader temperature range due to the substantially greater heterogeneity of the environments of these molecules on the silica surface. It should be noted that factors such as solvent composition and alkyl chain surface coverage are expected to additionally affect this phase transition phenomenon.32 Investigations of temperature-dependent conformational changes in different solvents are currently underway in this laboratory and will be reported at a later date. ACKNOWLEDGMENT The authors are grateful for support of this research by the Department of Energy (Grant DE-FG03-95ER14546). The authors also express appreciation to Professor Michael F. Burke and International Sorbent Technology for providing the SPE stationaryphase materials. Received for review April 28, 1998. Accepted September 17, 1998. AC980471S