A White Paper: Why Everyone Should Have Silica Hydride Based HPLC Columns in Their Lab

Professor Joseph J. Pesek Department of Chemistry San Jose State University San Jose, CA 95112

Dr. Maria T. Matyska MicroSolv Technology Corporation 1 Industrial Way West, Bldg. E Eatontown, NJ 07724

Abstract The usefulness of silica hydride-based HPLC stationary phases, commercially available today, is described and a number of unique features of these materials are illustrated. These stationary phases can be used for a broad range of applications including traditional reversed-phase (with unique selectivity) as well as both aqueous and organic normal phase analyses. This suite of columns is becoming a very popular tool for orthogonal data during method development. The ability to analyze hydrophobic and hydrophilic compounds in the same run, to separate basic compounds at low pH and for methods with very high organic mobile phases is especially interesting for use with mass spectroscopy as a detector. Several types of phases are available so the correct choice of separation material can be made depending on the degree of hydrophobic/hydrophilic selectivity required.

Introduction Silica-hydride based HPLC stationary phases have been investigated over the last ten years and possess both high stability and a broad range of chromatographic properties that are unmatched by any other single separation medium (1,2). A number of these features are especially attractive when coupling HPLC to mass spectroscopy. These separation materials span the range of mobile phase compositions from 100% aqueous to pure non polar organic solvents. Thus they can be used at high water (reversed-phase), high organic with some water present (aqueous normal phase) and pure organic (organic normal phase). The reversed-phase and aqueous normal phase modes are highly

complementary and it is possible to rapidly move between them with the silica hydridebased stationary phases due to the rapid equilibration of these separation materials and in some cases both mechanisms operate simultaneously thus retaining both hydrophobic and hydrophilic compounds in a single isocratic run. Because there are several types of silica hydride-based stationary phases currently available, it is possible to adjust the relative amount of reversed-phase and aqueous normal phase capabilities by varying the surface composition. The surface composition is a combination of the base silica-hydride and the organic moiety attached to it. There are currently five different types of silica hydride stationary phases to consider when assessing particular separation goals. The materials are in order of increasing hydrophobicity: Silica-C™ (silica hydride only surface); Diamond Hydride™ (a small amount of carbon on the silica hydride surface); UDCCholesterol™; Bidentate C8™(octyl); and Bidentate C18 (octadecyl) .

Experimental Columns: All of the columns used in this study were obtained from MicroSolv Technology (Eatontown, NJ) www.mtc-usa.com. Column sizes ranged from 75 to 150 mm all with a standard 4.6 mm id. for UV detection and with 2.1 mm i.d. for mass spectroscopy detection. Instrumentation: All LC experiments were conducted on either an Agilent 1050 HPLC system equipped with a diode arrary detector or an Agilent 1050 HPLC system interfaced to a Waters Micromass Platform LC mass spectrometer having ESI and APCI sources with 0

Masslynx 3.4 for data analysis. Temperature was maintained at 25 C for all experiments.

Results and Discussion The Silica-C™ column has the greatest retention capabilities for hydrophilic compounds since there are no organic groups attached to the hydride surface and thus is most suitable for aqueous normal phase (ANP) operation. It also possesses good selectivity for certain functional groups like hydroxyl, carbonyl, carboxylic acid, and halides so it is also a suitable separation material for organic normal phase separations. Amino acids are difficult to retain on typical reversed phase columns and generally elute at or near the dead volume. The only choice is to derivatize these compounds and make them more hydrophobic. This adds an additional step to the analysis and is a source of error that can lead to non-reproducible results. With underivatized amino acids, HILIC columns have been used for retention. However, HILIC depends on a uniform and consistent water layer on the surface for reproducible results (3). This is often difficult to achieve and is probably the main reason some laboratories report poor reproducibility on HILIC columns for the analysis of many hydrophilic compounds from run to run. Figure 1 shows the separation of phenylglycine and phenylalanine on the Silica-C™ column. Repeated injections of these and other amino acids give RSD values less than 0.3% in both isocractic and gradient methods. An even more versatile material for the analysis of hydrophilic compounds is the new Cogent Diamond Hydride™. With a small amount of carbon engineered onto the silica-hydride surface, the polar retention properties are diminished slightly with respect to Silica-C™. An example of the retention capabilities of this stationary phase is shown in Figure 2 for the separation of adenosine and guanine in the ANP mode. When this mixture is injected ten consecutive times, the RSD of the retention for these two

compounds is less 0.1%. Another important feature of the Diamond Hydride™ is the relatively low amount of TFA in the mobile phase needed to maintain excellent peak shape; in this case only 0.001%. This extremely low concentration of TFA is particularly important when using mass spectroscopy for detection since the amount typically used on other phases is 0.1% which can significantly depress the MS signal and lower sensitivity. The Diamond Hydride™ can be used for a wide variety of metabolites including amino acids, small organic acids and carbohydrates. Both the Silica-C™ and the Diamond Hydride™ do not have a strong affinity for water, which leads to the high reproducibility and fast equilibration after gradients in the ANP mode in comparison to HILIC. The lack of water affinity also accounts for the fact that both of these materials are excellent choices for organic normal phase (ONP) methods. Water does not have to be added to the mobile phase in trace amounts as is common practice with typical stationary phases used in ONP methods (silica, amino, cyano and diol) and this feature also makes gradients more feasible in this mode of operation. The attachment of bonded organic groups to the hydride surface results in more hydrophobic retention and a diminishing, but not elimination, of the ANP retention. An example of ANP retention with a bonded organic group is shown in Figure 3 for the retention of the drug Tobramycin™ on a silica hydride-based UDC-Cholesterol™ phase. The drug is highly polar with five amine groups and five hydroxyls. The peak shape is very symmetric and the retention, as on the Silica-C™ and Diamond Hydride™ phases is quite reproducible. Attempts to run this compound on several types of typical reversedphase columns resulted in very low retention and significant tailing, even for end capped materials. The interesting feature of silica hydride-based materials is that the addition of

organic groups on the surface result in the attainment of reversed-phase (RP) properties. The extent of RP retention depends on the type of moiety bonded to the surface. In this case the reversed-phase properties of the UDC-Cholesterol stationary phases are shown in Figure 4 where a mixture of steroids is separated in a mobile phase consisting of 50:50 methanol/water. This is a typical RP mobile phase composition. Detection is by mass spectroscopy using the APCI+ mode. In addition to its ANP and RP retention capabilities, the UDC-Cholesterol™ bonded material also has good shape selectivity for such compounds as polycyclic aromatic hydrocarbons and steroids. This feature is a result of the liquid crystal properties of cholesterol, some of which are preserved even when bonded to a silica surface. Of the other commercially available silica hydride-based phases, the Bidentate C8™ is slightly more hydrophobic than the UDC-Cholesterol™ phase and the Bidentate C18™ is the most hydrophobic. In general, a longer alkyl chain with no functional groups produces greater hydrophobic retention. However, the presence of the silica hydride surface imparts different properties to the material and hence selectivity is often different than most other reversed-phase materials. This can be advantage in at least two circumstances. First, different selectivity may allow for the resolution of analytes that cannot be separated on other stationary phases with the same bonded group. Second, it can provide a second (orthogonal) means for identifying one or more components in a mixture. This capability is especially useful for impurity profiling, assays, drug metabolism and stability indicating methods. A totally different column can provide the confirmation for an analysis that is often required in quality control and regulatory situations. The properties of more than four hundred stationary phases have been

evaluated including a large number of octadecyl materials (4). Their properties have been cataloged and an examination of the silica hydride based phase shows that the parameters used for determining column equivalency are much different for the hydride materials than many others thus making it possible for their use in the two situations described above. The versatility of the silica hydride-based C18 is shown in the next two figures. In Figure 5 a mixture containing both hydrophilic and hydrophobic compounds is separated on the Bidentate C18™ column. Under the mobile phase conditions used (60:40 acetonitrile/water) the reversed-phase mechanism is dominant since the first three compounds are hydrophilic and the last four are hydrophobic. When the amount of acetonitrile is increased, the two groups switch position as the ANP retention becomes stronger and the RP effects are diminished. The chromatogram shown represents the best resolution of this mixture. However, in other samples using a mobile phase where ANP retention dominates will produce the best resolution. The excellent RP capabilities of the Bidentate C18™ silica hydride-based phase are shown in Figure 6 where a group of carbohydrate isomers are separated in a 100% aqueous mobile phase. The excellent reproducibility of this phase is illustrated by comparing the 1st and 10th injections, which are essentially identical. The identity of the structural isomers is accomplished through mass spectroscopy since each compound has unique fragment ions. Another significant advantage of using mass spectroscopy for detection in combination with the ANP mode is the expansion of solvent choices. For example, it has been demonstrated that acetone is a very useful solvent for ANP applications. It has different solvent properties than acetonitrile and hence different selectivity. With UV

detection acetone is not a suitable mobile phase component due to its own strong absorption a lower wavelengths. Thus an expansion of separation capabilities is possible when the detector is not sensitive to the mobile phase composition. Another example of the same insensitivity to the solvent is the evaporative light scattering detector (ELSD). From a practical point, the other common solvent in aqueous based mobile phases is methanol. However, methanol being more polar than acetonitrile or acetone generally produces less ANP retention than these two solvents. ANP behavior can be observed for strongly polar (usually basic) compounds like tobramycin with methanol. The ANP mode offers another significant advantage with respect to mass spectroscopy detection. At higher amounts of organic in the mobile phase, sensitivity increases in MS due to lower noise in the spectrum. This can sometimes result in up an order of magnitude improvement in the lower limit of detection. Finally, ANP is a more instrument friendly method for the analysis of basic compounds than reversed phase. In RP methods the pH must be increased to the range of 8-10, which can lead to not only shorter column lifetimes but more importantly an increase in the amount of instrument down time due to replacement of valves, seals and other parts degraded by exposure to basic mobile phases.

Conclusions Silica hydride-based stationary phases are emerging as a unique type of separation material that provides analytical capabilities not available in a single type of ordinary bonded silica phase, and in some instances not in any other commercial column. The type and degree of selectivity can be controlled by starting with the base material (SilicaC) and increasing the amount of hydrophobicity on the surface by increasing the amount of bonded material (either via higher surface coverage or using larger molecular weight

hydrocarbon chains). Thus the selectivity changes from strong retention for hydrophilic compounds to increased retention for hydrophobic compounds when the surface has a high amount of C18 bonded. However, the ANP mode does not disappear even for the octadecyl phase so this stationary phase can provide significant retention for both hydrophobic and hydrophilic compounds as the mobile phase composition spans the range from 100% aqueous to essentially pure organic. For these reasons, and others, many labs worldwide have adopted these columns to challenge every method they develop, for complex mixtures, for compounds not typically retained in RP and as a way to extend the range and reach of each and every HPLC method.

References (1) J.J. Pesek and M.T. Matyska, J. Sep. Sci. 28, 1845–1854 (2005). (2) J.J..Pesek and M.T. Matyska, J. Liq. Chromatogr & Rel. Technol. 29, 1105–1124 (2006). (3) D.L. Roush, L.Y. Hwang and F.D. Anita, J. Chromatogr A 1098, 55–65 (2005). (4) http://www.usp.org/USPNF/columns.html

Figure 1

Figure 1. Separation of phenylglycine (1) and phenylalanine (2) using a Silica-C hydride based stationary phase. Mobile phase: 80:20 acetonitrile/DI water + 0.5% formic acid. Detection at 254 nm.

Figure 2

Figure 2. Separation of adenosine (1) and guanine (2) using a Diamond Hydride™ silica column. Mobile phase: 80:20 acetonitrile/DI water + 0.001% TFA. Detection at 232 nm.

Figure 3

Figure 3. Separation of uracil and tobramycin on a silica hydride-based cholesterol column. Mobile phase: 70:30 acetonitrile/DI water + 0.5% formic acid. Detection by mass spectrometry using the APCI+ ionization mode with single ion monitoring.

Figure 4

5.00

10.00 15.00

20.00 25.00 30.00 35.00 40.00 45.00 50.00 Time (min)

Figure 4. Separation of a six steroid component mixture on a silica hydride-based cholesterol column. Mobile phase 50:50 methanol/DI water. Detection by mass spectroscopy in the APCI+ ionization mode with single ion monitoring. Solutes: 13.29 min. = andrenosterone; 16.99 min = corticosterone; 27.46 min = 4 –androstene-3,17dione; 32.12 min = 11α-acetoxyprogesterone; 38.07 min. = estrone; and 48.89 min. = estradiol

Figure 5

Figure 5. Separation of a seven-component mixture containing both polar and nonpolar compounds on the Bidentate C18 silica hydride stationary phase. Mobile phase: 60:40 acetonitrile/water. Detection: 254 nm. Compounds 1-3 are polar and 4-7 are nonpolar.

Figure 6

Figure 6. Separation of carbohydrate isomers with MW = 505 on the Bidentate C18 silica hydride stationary phase using a 100% aqueous mobile phase. Detection by mass spectroscopy in the APCI+ mode.