Deuterium Nuclear Magnetic Resonance Spectroscopy as a Probe For

Department of Chemistry, University of Vermont, Burlington, Vermont 05405-0125. Deuterium nuclear ... of slurry samples, consisting of both mobile- an...
0 downloads 0 Views 74KB Size
Anal. Chem. 1998, 70, 602-607

Deuterium Nuclear Magnetic Resonance Spectroscopy as a Probe For Reversed-Phase Liquid Chromatographic Bonded-Phase Solvation. 3. Tetrahydrofuran and Water Binary Systems Jessica L. Wysocki† and Karen B. Sentell*

Department of Chemistry, University of Vermont, Burlington, Vermont 05405-0125

Deuterium nuclear magnetic resonance (2H NMR) spectroscopy is a useful tool in probing molecular association between components in a binary hydro-organic mixture, such as tetrahydrofuran (THF) and water. Measurements of 2H longitudinal relaxation times (T1) for deuterated analogs of THF and water provide this type of information, since the magnitude of T1 is indicative of molecular association and mobility. This technique is also useful in determining the interactions between the mobile phase and reversed-phase stationary phase. T1 measurements of slurry samples, consisting of both mobile- and stationary-phase components, allow for changes in the longitudinal relaxation time of mobile-phase components upon contact with the stationary phase (∆T1) to be monitored. These ∆T1 values are indicative of the degree of association of the THF-water mobile-phase components with the stationary phase. A reversed-phase liquid chromatographic (RPLC) separation is a thermodynamic process involving the transfer of a solute between the stationary and mobile phases. Even though RPLC is a widely used separation technique, improved knowledge about solvent association with the stationary phase in these chromatographic systems and the role that this solvation plays in RPLC retention mechanisms at the molecular level is important in better understanding and optimizing real-world chromatographic applications. Katz et al.1 have shown that what might be described as a binary solvent mixture of tetrahydrofuran (THF) and water on a macroscopic scale is, in fact, at least a ternary mixture microscopically. This mixture includes self-associated organic modifier molecules, water molecules associated with other water molecules, and “mixed” species, wherein organic modifier molecules are associated with water molecules. Due to differing degrees of association between water and THF molecules as a function of composition, the extent of interaction between these two mobilephase components and between the mobile and stationary phases

will be a function of the bulk composition of the mobile phase. This has a marked effect on solute retention in chromatographic systems. Previous studies2-22 have demonstrated the usefulness of solution-state nuclear magnetic resonance (NMR) spectroscopy in the investigation of molecular interactions in RPLC systems. However, many of these studies were performed with uncommon solvents or only encompassed a narrow range of binary compositions.1,5,8-12,16 Previous work in our laboratory2-4 has indicated that, with an appropriate wetting procedure for the stationary phase, the entire binary composition range for hydro-organic mixtures of acetonitrile and methanol becomes accessible for study via solution-state deuterium (2H) NMR. Therefore, this type of study has been extended to THF-water solutions in order to investigate the association behavior between these cosolvents. Using deuterated analogs of these mobile-phase components, their degrees of interaction as a function of bulk solution composition can be probed by measurement of the solution-state deuterium longitudinal relaxation time (2H T1) of the mobile-phase component of interest.

* Author to whom correspondence should be addressed. Present address: Ciba Vision Corp., 11460 Johns Creek Pkwy., Duluth, GA 30097-1556. E-mail: [email protected]. Fax: (770) 418-2681. † Present address: Department of Chemistry, Florida State University, Tallahassee, FL 32306-4390. E-mail: [email protected]. Fax: (850) 6448281. (1) Katz, E. D.; Ogan, K.; Scott, R. P. W. J. Chromatogr. 1986, 352, 67.

(2) Bliesner, D. M. Ph. D. Dissertation, University of Vermont, 1992. (3) Bliesner, D. M.; Sentell, K. B. J. Chromatogr. 1993, 631, 23. (4) Bliesner, D. M.; Sentell, K. B. Anal. Chem. 1993, 65, 1819. (5) Marshall, D. B.; McKenna, W. P. Anal. Chem. 1984, 56, 2090. (6) Sentell, K. B. J. Chromatogr. A 1993, 656, 231. (7) Gilpin, R. K.; Gangoda, M. E. J. Chromatogr. Sci. 1983, 21, 352. (8) Gilpin, R. K.; Gangoda, M. E. Anal. Chem. 1984, 56, 1470. (9) Gilpin, R. K.; Gangoda, M. E. J. Magn. Reson. 1985, 64, 2090. (10) Gilpin, R. K.; Gangoda, M. E. J. Magn. Reson. 1987, 74, 134. (11) Gilpin, R. K.; Gangoda, M. E. Langmuir 1990, 6, 941. (12) Ellison, E. H.; Marshall, D. B. J. Phys. Chem. 1991, 95, 808. (13) Maciel, G. E.; Zeigler, R. C.; Taft, R. K. In Silanes, Surfaces and Interfaces (Chemically Modified Surfaces Series, Vol. 1); Leyden, D. E., Ed.; Gordon and Breach: New York, 1986; p 413. (14) Maciel, G. E.; Zeigler, R. C. In Chemically Modified Surfaces in Science and Industry (Chemically Modified Surfaces Series, Vol. 2); Leyden, D. E., Ed.; Gordon and Breach: New York, 1988; p 319. (15) Maciel, G. E.; Zeigler, R. C. J. Am. Chem. Soc. 1991, 113, 6349. (16) Albert, K.; Evers, B.; Bayers, E. J. Magn. Reson. 1985, 62, 428. (17) Bayer, E.; Paulus, A.; Peters, B.; Laupp, G.; Reiners, J.; Klaus, A. J. Chromatogr. 1986, 364, 25. (18) Albert, K.; Pfleiderer, B.; Bayer, E. In Chemically Modified Surfaces in Science and Industry (Chemically Modified Surfaces Series, Vol. 2); Leyden, D. E., Ed.; Gordon and Breach: New York, 1988; p 287. (19) McNally, M. E.; Rogers, L. B. J. Chromatogr. 1985, 331, 23. (20) Shah, P.; Rogers, L. B.; Fetzer, J. C. J. Chromatogr. 1987, 388, 411. (21) Claessens, H. A.; van de Ven, L.; de Haan, J.; Cramers, C. A.; Vonk, N. J. High Resolut. Chromatogr. Chromatogr. Commun. 1983, 6, 433. (22) Kelusky, E. C.; Fyfe, C. A. J. Am. Chem. Soc. 1986, 108, 1746.

602 Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

S0003-2700(97)00507-6 CCC: $15.00

© 1998 American Chemical Society Published on Web 02/01/1998

2H

NMR is a popular technique for studying changes in the local environment of the deuterated species23 because longitudinal relaxation for quadrupolar nuclei is dominated by intermolecular interactions (i.e., hydrogen-bonding, dipole-dipole, or van der Waals interactions) rather than by intramolecular effects. Intramolecular dipole-dipole interactions, as well as quenching by paramagnetic impurities, compete poorly, if at all, with the quadrupolar relaxation process. For these reasons, 2H T1 NMR studies are an effective means for studying not only binary solution structure but also associative effects of the stationary phase while in contact with the binary mobile phase. Because deuterium is a quadrupolar nucleus with a spin quantum number I ) 1, it has a solution-state longitudinal relaxation rate (1/T1) described by eq 1:

1/T1 ) 3π2

2I + 3 2 χ τc I(2I - 1)

(1)

where I and the quadrupolar coupling constant, χ, are constants for each type of nucleus in a particular environment.24 The molecular correlation time, τc, is a measure of the rate at which the molecule of interest is able to rotate through one radian and is, therefore, a measure of molecular mobility. Since the longitudinal relaxation time (T1) is inversely related to τc, the magnitude of T1 is a convenient means of expressing extent of molecular motion. The amount of intermolecular association experienced by a molecular species is, therefore, directly related to the longitudinal relaxation time. The greater the degree of molecular association present, the lower the mobility of the molecular species of interest. This is demonstrated by a subsequent decrease in the longitudinal relaxation time, T1, of the species. It has been shown that the composition and thickness of the solvation layer of a chromatographic system depend upon the composition of the bulk mobile phase, type of bonded phase ligand, degree of silica surface modification, temperature and number and type of residual silanols present.25-33 In this study, the solvation layer for two well-characterized monomeric C18 stationary phases in contact with THF-water solutions was studied by 2H T1 experiments. The T1 values for solvent in contact with stationary phase are a weighted average of the free solution and stationary-phase-associated solvent species, due to rapid exchange of solvent species occurring between the bulk solution and species associated with the solid surface.34 This two-site rapid-exchange model is an appropriate model for solvent interactions with microporous silica gel5 and solid materials.5,34,35 The observed relaxation rate for this two-site rapid exchange model is, therefore,4,5,34,35 (23) Mantsch, H. H.; Saito, H.; Smith, I. C. D. Prog. Nucl. Magn. Reson. Spectrosc. 1977, 11, 211. (24) Harris, R. K. Nuclear Magnetic Resonance Spectroscopy: A Physicochemical View; Pitman Books: London, 1983. (25) Martire, D. E.; Boehm, R. E. J. Phys. Chem. 1983, 87, 1045. (26) McCormick, R. M.; Karger, B. L. Anal. Chem. 1980, 52, 2249. (27) Yonker, C. R.; Zwier, T. A.; Burke, M. F. J. Chromatogr. 1982, 241, 257. (28) Yonker, C. R.; Zwier, T. A.; Burke, M. F. J. Chromatogr. 1982, 241, 269. (29) Scott, R. P. W.; Simpson, C. F. Faraday Symp. Chem. Soc. 1980, 15, 69. (30) Marqusee, J. A.; Dill, K. A. J. Chem. Phys. 1986, 85, 434. (31) Dill, K. A. J. Phys. Chem. 1987, 91, 1980. (32) Dorsey, J. G.; Dill, K. A. Chem. Rev. 1989, 89, 331. (33) Gilpin, R. K.; Squires, J. A. J. Chromatogr. Sci. 1981, 19, 195. (34) Glasel, J. A.; Lee, K. H. J. Am. Chem. Soc. 1974, 96, 970. (35) Woessner, D. E.; Snowden, B. S. J. Colloid Interface Sci. 1970, 34, 290.

1/T1(obs) ) (Fb)/T1b + (1 - Fb)/T1s

(2)

where T1(obs) is the observed T1 value of the solution component in contact with the stationary phase, Fb is the fraction of solvent molecules in the associated state, T1b is the relaxation time for associated species, and T1s is the T1 value for the component in the bulk solution. This equation states that the T1 value observed is a weighted average of free-solution solvent species as well as solvent molecules associated with stationary phase5 and is valid only when the relaxation rate of associated solvent is greater than that of bulk solvent and the fraction of solvent molecules in the associated state is small,4,5,35 such as in these studies, where only 3% deuterated solvent components are used. This change in the observed T1 (∆T1 ) T1{bulk solution} - T1{solvent associated with stationary phase}) is indicative of the degree of association of the solvent in contact with the stationary phase. EXPERIMENTAL SECTION Chemicals. [2H8]Tetrahydrofuran (Isotec, Miamisburg, OH), HPLC-grade tetrahydrofuran (Fisher, Fair Lawn, NJ), and deuterium oxide (Aldrich, Milwaukee, WI) were used without further purification. HPLC-grade water was obtained in-house using a Barnstead Nanopure (Sybron, Boston, MA) water purification system. Two different monomeric C18 stationary phases were studied. The first, designated LT1, was a 10-µm non-endcapped C18 phase of high alkyl chain bonding density (4.4 µmol/m2), which was synthesized in our laboratory using previously described techniques.36 The second, designated ODS1, was a commercially available Spherisorb (Phase Separations, Norwalk, CT) 5-µm partially endcapped C18 phase with low bonding density (1.4 µmol/m2). Sample Preparation. Approximately 3% (by volume) [2H8]THF or 2H2O was added to its nonlabeled analog. The solution samples were hand-mixed at each binary composition. The stationary phase was dried under vacuum for 12 h at 110 °C and then hand-packed into a 10-cm × 4.6-mm stainless steel column, and the desired mobile phase composition ratio (v/v% organic cosolvent/water) was pumped through the column by a Shimadzu LC6 pump (Kyoto, Japan) at a flow rate sufficient to generate at least 1500 psi back pressure. A sufficient volume of mobile phase (at least 50 column volumes) was pumped through the column in both directions to allow complete wetting of the stationary phase. Samples were prepared in 10% or smaller volume increments in the range from 100:0 (v/v) to 0:100 (v/v) [2H8]THF/ 2H O. T values were then measured for the deuterons of the 2 1 labeled component of the binary mobile-phase mixture. Initial experiments were carried out with both mobile-phase components deuterated, but this led to a smaller relaxation time than that measured for individually labeled components in the same mixtures. For this reason, only one labeled mobile-phase component was used for any individual experiment. NMR Measurements. T1 values for the deuterons of labeled components of THF-water mobile-phase mixtures were measured on a Bruker WM-250 NMR spectrometer (Billerica, MA), operating at a field strength of 5.875 and a frequency of 38.4 MHz for 2H. The standard inversion recovery pulse sequence was used: (36) Sentell, K. B.; Barnes, K. W.; Dorsey, J. G. J. Chromatogr. 1988, 455, 95.

Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

603

(180°-τ-90°-acquire-delay)n where the pulse delay time, τ, ranged from 16 to 1.0 × 10-5 s. Each sequence employed 9-10 τ values. Solution mixtures and stationary-phase slurry acquisitions required at least 4 or 32 scans, respectively, for a sufficient signal-to-noise ratio. The relaxation delay between successive pulse sequences was greater than 5T1 in order to allow the system to return to its equilibrium magnetization state between measurements. The total acquisition time for all τ values of a single binary mixture composition ranged from 10 to 90 min, as opposed to 1.5-4.5 h for a single mobile phase-stationary phase slurry experiment. Reflected power was minimized to reduce errors in the 180° and 90° pulse angles, which were 25° and 10°, respectively. Automatic field frequency lock was used to acquire spectra. The temperature was set at 303 K (30 °C). Once the inversion recovery sequence is performed on each sample, the longitudinal relaxation time is calculated using the inversion recovery equation:

I ) I1(1 - e-τ/R1) - I2

(3)

where I is the experimental peak intensity for each τ value, I1 is the peak intensity after infinite recovery time, τ is the experimental pulse delay time after which each peak intensity is measured, R1 is the time constant of recovery, which is 1/T1, and I2 is the peak intensity immediately after the experiment is begun (time ) 0). Since a plot of peak intensity versus τ follows the expected exponential behavior, an iterative least-squares fit of experimental peak intensity (I) versus pulse delay time (τ) was used to obtain the value of R1 (1/T1) for each experimental condition. All T1 values are reported at a confidence level of (1 standard deviation. RESULTS AND DISCUSSION Hyperfine splitting between two chemically different deuteron environments causes two different chemical shifts to be observed in the proton-decoupled [2H]8THF NMR spectrum. The two deuteron resonances for THF differ in their chemical environment due to their position relative to the oxygen in the heterocyclic ring. Those deuterons closer to the oxygen are deshielded as a result of their proximity to the electron-withdrawing oxygen and, therefore, are shifted farther downfield in the spectrum. Although these two environments are chemically different, Figure 1A demonstrates that the relaxation mechanisms of both deuterated resonances as functions of binary solution composition are similar. Due to the similar trends seen for both deuteron environments, only one resonance (i.e., downfield-shifted resonance) will be shown for the remaining figures. Neat Mobile Phase T1 Measurements. Molecular associations such as hydrogen-bonding, dipole-dipole interactions, or van der Waals interactions between the components of a binary mixture are the main contributors to quadrupolar relaxation of the deuterated mobile-phase components. THF molecules in neat solution will undergo weaker THF-THF hydrogen-bonding and van der Waals associations, which allow more mobility within the solution structure than when more strongly hydrogen-bonding species are present. THF-THF interactions are weaker than water-water or water-THF interactions because THF can serve only as a proton acceptor. For these reasons, the mobility of THF molecules and associated THF-water molecules is dominated by 604

Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

Figure 1. (A) Deuterium longitudinal relaxation time for downfield (less shielded) [9] and upfield (more shielded) [0] resonances of [2H]8THF in THF-water mobile phase versus volume percent THF in bulk mobile phase. (B) Deuterium longitudinal relaxation time for 2H2O [9] in a THF-water mobile phase versus volume percent THF in bulk mobile phase.

the strong hydrogen bond lattice formed between water molecules.1 Hydrogen bonding between THF and water molecules will also decrease the mobility of THF molecules. The magnitude of the T1 for deuterium oxide (2H2O) as a function of solution composition indicates strong hydrogenbonding interactions and an efficient relaxation process (Figure 1B). At compositions with