Anal. Chem. 2006, 78, 4155-4160
Articles
Ice Chromatography. Characterization of Water-Ice as a Chromatographic Stationary Phase Yuiko Tasaki and Tetsuo Okada*
Department of Chemistry, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8551, Japan
Water-ice has been characterized as a stationary phase for liquid chromatography. Solutes having two or more polar groups are retained on this stationary phase with THF/hexane as the mobile phase, suggesting that multipoint interactions are required for measurable solute retention. Chromatographic separation of phenols or crown ethers on water-ice is possible. The ice surface is expected to provide two different adsorption sites coming from the OH and O dangling bonds. Although the solute partition into the quasiliquid layer is also considered, the dependence of the retention times on the THF concentration implies that the interaction of solutes with the waterice surface rather than the partition into the quasiliquid layer is responsible for solute retention. A retention model suggests that the number of adsorption sites for a crown ether depends on its ring size, whereas two sites are involved for the retention of phenols having two hydroxyl groups. Although hydroxyl groups can act as both a hydrogen bond donor and an acceptor, the interaction with the ice OH sites, which are exposed to the surroundings in comparison with the ice O sites, is more important. However, when an acyclic polyether is added to the mobile phase, its adsorption onto the water ice surface allows the creation of the O sites that phenols can approach without steric hindrance. In the presence of the polyethers adsorbed on the ice surface, the retention of phenols is enhanced, whereas crown ethers become less retained due to the competitive adsorption of the polyethers. The development of stationary phases that provide desired separation selectivity and high separation performance is one of the most important tasks to enhance the usefulness and applicability of chromatographic methodology. There are two major approaches for exploitation of novel chromatographic stationary phases; that is, chemical modifications of known materials1-7 and * To whom correspondence should be addressed. Phone and Fax: +81-35734-2612. E-mail:
[email protected]. (1) Ruderisch, A.; Iwanek, W.; Pfeiffer, J.; Fischer, G.; Albert, K.; Schurig, V. J. Chromatogr., A 2005, 1095, 40. (2) Huang, X.; Wang, J.; Wang, Q.; Huang, B. Anal. Sci. 2005, 21, 253. (3) Lai, X.-H.; Ng, S.-C. J. Chromatogr., A 2004, 1031, 135. (4) Kurata, K.; Ono, J.; Dobashi, A. J. Chromatogr., A 2005, 1080, 140. (5) Hasegawa, T.; Umemura, T.; Koide, A.; Chiba, K.; Ueki, Y.; Tsunoda, K.; Haraguchi, H. Anal. Sci. 2005, 21, 913. 10.1021/ac0602470 CCC: $33.50 Published on Web 05/04/2006
© 2006 American Chemical Society
developments of new base materials.8-13 The surface modifications of silica gel and polymer beads have been extensively studied in various branches of chemistry. The introduction of ligand molecules onto the base packing materials through an appropriate covalent bond is typically used for the preparation of these stationary phases.1-3 In addition, the physisorption of functional molecules, such as surfactants, onto a base stationary phase is also a versatile and convenient means.4-7 One recent innovation in base materials is the development of monolithic stationary phases.7-13 Such stationary phases have been widely used because of their potentially high separation performance and short analysis times, similar to an open tubular capillary. However, although monolithic stationary phases promise high efficiency at high speed, their surface is basically the same as usual packed stationary phases, and they provide no new avenues to selectivity. Green chemistry is a key ingredient for a sustainable development. Ecologically friendly materials need to be used in all aspects of science and technology; separation science is no exception. Instead of organic solvents, water at high-temperatures, even in the supercritical state, has been used as the chromatographic mobile phase.14-17 Under such conditions, water becomes capable of dissolving normally water-insoluble organic molecules; thus, the consumption of organic solvents can be reduced. Ionic liquids have also received much attention as ecologically friendly substitutes for organic solvents,18-25 albeit their applications in (6) Okada, T. Anal. Chim. Acta 2005, 540, 139. (7) Hu, Z.-X.; Zhang, W.-N.; He, H.-B.; Feng, Y.-Q.; Da, S.-L. J. Chromatogr., B 2005, 827, 173. (8) Li, Z.; Jaroniec, M. Anal. Chem. 2004, 76, 5479. (9) Tanaka, N.; Kobayashi, H.; Nakanishi, K.; Minakuchi, H.; Ishizuka, N. Anal. Chem. 2001, 73, 420A. (10) Eeltink, S.; Desmet, G.; Vivo-Truyols, G. l.; Rozing, G. P.; Schoenmakers, P. J.; Kok, W. T. J. Chromatogr., A 2006, 1104, 256. (11) Fan, Y.; Zhang, M.; Feng, Y.-Q. J. Chromatogr., A 2005, 1099, 84. (12) Jia, L.; Tanaka, N.; Terabe, S. Electrophoresis 2005, 26, 3468. (13) Liu, Z.-S.; Xu, Y.-L.; Wang, H.; Yan, C.; Gao, R.-Y. Anal. Sci. 2004, 20, 673. (14) Sanagi, M. M.; See, H. H. J. Liq. Chromatogr. Relat. Technol. 2005, 28, 3065. (15) Yarita, T.; Nakajima, R.; Shimada, K.; Kinugasa, S.; Shibukawa, M. Anal. Sci. 2005, 21, 1001. (16) Yang, Y. LC-GC Europe 2003, 16, 37. (17) Pyo, D.; Lim, C. Bull. Korean Chem. Soc. 2005, 26, 312. (18) Ding, J.; Welton, T.; Armstrong, D. W. Anal. Chem. 2004, 76, 6819. (19) Anderson, J. L.; Armstrong, D. W. Anal. Chem. 2005, 77, 6453. (20) Planeta, J.; Roth, M. J. Phys. Chem., B 2005, 109, 15165. (21) Stalcup, A. M.; Cabovska, B. J. Liq. Chromatogr. Relat. Technol. 2004, 27, 1443. (22) Li, S.; He, C.; Liu, H.; Li, K.; Liu, F. J. Chromatogr., B 2005, 826, 58.
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4155
separation have been limited to gas chromatography18,19 and solvent extraction.22-25 Ecologically adaptable liquids which are potentially applicable to various liquid-phase separations have, thus, been extensively investigated.21 In contrast, no efforts have been made to develop ecologically friendly solid phases. As far as chromatography is concerned, not only the mobile phases but also the stationary phase should be investigated from this viewpoint. Dasgupta and Mo26 studied water-ice as a chromatographic stationary phase. In their study, the feasibility of chromatographic separation on ice was presented with two dye compounds (methyl red and methyl violet 2B) as samples. These solutes have entirely different structures, and therefore, systematic discussions on the retention mechanism were not possible. To the best of our knowledge, this remains the only attempt to use water-ice as a liquid chromatographic stationary phase. Aside from complete environmental compatibility, water-ice has an unusual and important feature that material separation occurs on well-defined crystalline surfaces. Molecular interactions taking place on the crystalline surfaces have mostly been characterized with spectroscopy, electrochemistry, or microscopy;27-29 studies based on differing sorption properties are rare. Although separation on a single crystal ice phase is difficult at the present stage, such chromatography obviously offers an opportunity to probe and systematically discuss molecular interactions on water-ice surfaces. We present here fundamental aspects and potential usefulness of such an approach, which is hereinafter called “ice chromatography”. Throughout the rest of the manuscript, ice connotes water-ice. EXPERIMENTAL SECTION The chromatographic system was composed of a Shimadzu HPLC pump, model LC-10A; a Rheodyne injector equipped with a 100-µL sample loop; a Tosoh UV detector, model UV-8020 set to 275 nm; and an Advantec low-temperature bath, model TBT220DA. Ice particles were prepared by introducing nebulized water droplets directly into liquid nitrogen. The microscopically measured diameters of ice particles ranged from 20 to 100 µm and were typically 50 µm. They were put into a PEEK column (7.5 mm i.d × ∼150 mm length) immersed in liquid nitrogen and were tapped down with a plastic rod. After packing, the column was transferred to the low-temperature bath. The volume ratio of the ice stationary phase to the entire column was ∼70% (water volume was measured after melting the column packing). The temperature of the separation column, the sample loop, and the solvent reservoir was maintained at approximately -12 to -15 °C. The mobile phase was precooled by passing through a coiled tube connected between the pump and the sample injector. Ultrasonically degassed THF/hexane mixtures were used as the mobile phase. Small amounts of ice were put in the mobile (23) Luo, H.; Dai, S.; Bonnesen, P.; Haverlock, T.; Moyer, B.; Buchanan, A. Solvent Extr. Ion Exch. 2006, 24, 19. (24) Dietz, M. L.; Stepinski, D. C. Green Chem. 2005, 7, 747. (25) Abbott, A. P.; Capper, G.; Davies, D. L.; Rasheed, R. K.; Shikotra, P. Inorg. Chem. 2005, 44, 6497. (26) Dasgupta, P. K.; Mo, Y. Anal. Chem. 1997, 69, 4079. (27) Conway, B. E. J. Electroanal. Chem. 2002, 524-525, 4. (28) Hafner, J. J. Mol. Struct. 2003, 651-653, 3. (29) Zhang, J.; Chi, Q.; Albrecht, T.; Kuznetsov, A. M.; Grubb, M.; Hansen, A. G.; Wackerbarth, H.; Welinder, A. C.; Ulstrup, J. Electrochim. Acta 2005, 50, 3143.
4156 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006
Figure 1. Strucutures of solute molecules and schematic illustration of the water-ice (0001) surface.
phase to saturate it with water at the operating temperature. The water concentration in the mobile phase measured by a Karl Fischer method was lower than the detection limit of the instrument used, a Metrohm, model KF701. Water coming from additives should also be removed by keeping the mobile phase at the operating temperature. Crown ethers were synthesized according to the literature.30 The concentrations of solutes were 10-80 µM, depending on their absorptivity at the detection wavelength. Reagents of the highest available grade were used as received. RESULTS AND DISCUSSION Possible Interactions on an Ice Surface. Thermodynamic considerations and experimental evidence indicate that a quasiliquid layer is present on the surface of ice;31-34 this has been of particular interest in meteorological research, because it is relevant to the phenomena taking place in the troposphere. The available literature pertains to ice exposed to water-saturated air. Ice is here exposed to a water-saturated organic solvent; a similar liquid layer should, nevertheless, be present. The nature of this quasiliquid layer is not well-understood, despite substantial efforts to characterize it. It is known that the quasiliquid layer becomes thicker as the temperature approaches the melting point of ice. Although the only previous work26 presumed that this liquid layer is responsible for the chromatographic retention, the contribution of this layer to the solute retention in ice chromatography may not be high because the operating temperature (-12 °C) is substantially below freezing. Its thickness also depends on the pressure. The operating pressure was low enough (0.1-0.2 MPa) to neglect the pressure effect on the thickness of the quasiliquid layer. Nevertheless, we took the effect of the quasiliquid layer into account in the following discussion on the retention characteristics of solutes. The other mechanism of chromatographic retention on ice is the adsorption of solutes on the ice surface through hydrogen bonds. There are two types of dangling bonds on the water-ice surface; that is, OH sites and O sites, as schematically illustrated in Figure 1.35,36 Although the ice particles used in this work must be polycrystalline without a strictly defined surface structure, it is useful to discuss solute retention on the basis of the structural match of the solute with the ice surface. The molecular structures (30) Pedersen, C. J. J. Am. Chem. Soc. 1967, 89, 7017. (31) Dash, J.G.; Haiying, F.; Wettlaufer, J. S. Rep. Prog. Phys. 1995, 58, 115. (32) Knight, C. A. J. Geophys. Res. 1996, 101, 12921. (33) Baker, M. B.; Dash, J. G. J. Geophys. Res. 1996, 101, 12929. (34) Do ¨ppenschmidt, A.; Butt, H.-J. Langmuir 2000, 16, 6709. (35) Nada, H.; Furukawa, Y. J. Crystal Growth 2005, 283, 242. (36) Eisenberg, D.; Kauzmann, W. The Structure and Properties of Water; Oxford University Press: Oxford, 1969, Chapter 3.
Chart 1
of sample compounds with ether or hydroxyl groups used in this work are also depicted in Figure 1. Polyethers likely interact with the OH sites on the ice surface, whereas phenolic compounds can form the hydrogen bonds with both the OH sites and the O sites. The number of potential interacting groups contained in the solute is also an important factor to govern its retention. Monoand dihydroxyl compounds were examined as solutes because solutes with more hydroxyl groups have limited solubility in the mobile phase used. In contrast, for polyethers, varying the number of polar groups over the wider range is possible. Crown ethers having up to 10 ethereal oxygen atoms as well as Triton X-100, a mixture of acyclic polyethers, were studied (see Chart 1). These structural variations of the solute compounds were expected to provide an insight into the retention mechanisms involved in ice chromatography. Retention and Separation of Selected Molecules on an Ice Stationary Phase. Figures 2 and 3 show typical ice chromatograms of phenols and crown ethers. The relatively large particle sizes of the ice stationary phase result in peak broadenings and, in turn, the low theoretical plate number (∼250 plates for DB24C8 in Figure 3). A column pressure was typically 0.2 MPa, which is ∼1/20 as low as that measured for a usual HPLC separation column, suggesting that the packing is loose for the ice column prepared in the way described above. It is known that the ice particles fuse into larger clumps even below the melting temperature, at least in air. A similar phenomenon was observed in THF/ hexane solutions saturated with water; ice particles packed in a separation column fused into a porous rod after use as a stationary
Figure 2. Ice chromatograms for phenols. Mobile phase, 2% (v/v) THF in hexane.
phase for several hours. As the particle size becomes smaller, the interparticle fusion of water-ice should be enhanced by higher column pressures. In addition, the melting point of the waterice becomes low with increasing pressure; this is probably not very serious because the rate of a melting point decrease with increasing pressure is only ∼7 × 10-3 K atm-1. From these viewpoints, Figures 2 and 3 may show the best performance that we can attain with usual chromatographic instruments. The preparation of the ice columns is a key technique, and improvements are still necessary for successful separation, as pointed out by Dasgupta and Mo.26 The nonporous nature of ice particles obviously results in a small stationary phase area, which possibly leads to a low loading capacity. In the present study, the concentration of samples was kept as low as possible, typically 10-30 µM. Effects of the sample concentration on peak shapes were not found; no overloading
Figure 3. Ice chromatographic separation of crown ethers. Mobile phase: (a) 3% (v/v) THF in hexane, (b) 4.9% (v/v) THF in hexane. Other conditions are given in the text.
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006
4157
Table 1. Partition Coefficients for Selected Solute between THF/Hexane and Water THF concn,% solute
4
5
6
7
2
3
4
Partition Coefficient DB18C6 4.1 4.1 4.3 DB24C8 2.2 2.3 2.3 2.4 resorcinol 0.46 0.57 0.69 hydroquinone 2.8 × 10-3 3.0 × 10-3 3.4 × 10-3
occurs. The large interparticle distances also require long times for solute diffusion and possibly cause peak broadening. However, in the present case, this does not induce peak tailing because flow rates did not influence peak shapes. Interestingly, DB24C8 gives a sharper peak than DB18C6, whereas the former is eluted later. Thus, the peak shape strongly depends on the nature of individual molecules. The diffusion and adsorption kinetics should be investigated to specify the source of peak tailing; this is a future task. It should be noted that the retention of phenols having one hydroxyl group, such as phenol and cresols, was very weak. Although their retention was confirmed with hexane as the mobile phase, they were not retained when THF was added to hexane. In contrast, compounds having two hydroxyl groups gave measurable retention, even with THF/hexane mobile phases. Figure 3 shows that the retention of polyethers is also enhanced by an increasing number of ethereal oxygen atoms. Although results are not shown here, we confirmed by HPLC analyses of effluents obtained with Triton X-100 injection that longer acyclic polyethers are more strongly adsorbed on water-ice than shorter counterparts. These results strongly suggest that multipoint hydrogen bond formations are important for retention on the ice surface if the retention is caused by hydrogen bond formation between the ice surface and a solute molecule. However, the water solubility of a compound is, in general, enhanced with an increasing number of polar groups contained in the molecules. We cannot exclude the possibility that the retention is caused by the partition of solutes into the quasiliquid layer. To obtain further information on the ice chromatographic retention mechanism, the partition coefficients of some solute molecules were determined with water/THF/hexane systems. Although the structure and properties of the quasiliquid layer may be different from those of bulk water to some extent, these two different states of water should have a similar partition nature. Effects of the THF concentration on the partition coefficients of selected solutes are summarized in Table 1. The partition coefficients are almost constant for crown ethers and slightly increase for phenols with increasing THF concentration. It must be significant to compare these data with the dependence of ice chromatographic retention times of these solutes on the THF concentration in the mobile phase. Figure 4 shows the effects of the THF concentration in the mobile phase on the retention times. Ice chromatographic retention is much more sensitive to the THF concentration in the mobile phase than the partition coefficients. Although the addition of a polar solvent into the mobile phase usually causes a reduction of retention in normal phase chromatography, the effects of a polar component are generally much more modest. This result should suggest that the mechanism 4158 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006
Figure 4. Effects of THF concentration in the mobile phase on solute retention. Mobile phase was prepared in hexane. Curves are fitted to eq 1 or 2 assuming KTHF ) 5. Parameters used are listed in Table 2.
other than the partition into the surface liquid layer is mainly responsible for the determination of ice chromatographic retention. As mentioned above, the quasiliquid layer becomes thinner with decreasing temperature. Such a thin liquid layer loses its bulk properties and cannot accommodate a solute molecule therein. Even though the quasiliquid layer actually exists on the water-ice surface, it should be strongly affected by the surface structures of water-ice and may behave as a part of solids rather than as a liquid as far as molecular interactions are concerned. The molecular processes taking place on the water ice surface can, therefore, be reasonably discussed on the basis of the hydrogen bond formation between a solute and the water-ice surface. As shown in Figure 2, the retention of phenols decreases in the order of p-hydroxybenzyl alcohol (HBA) > resorcinol > p-hydroquinone. A simple consideration may support an idea that the match of the length between hydroxyl groups with that between active sites on the ice surface is an important factor to determine the retentivity of a solute. However, the situation is very complex because of the amphoteric nature of the hydroxyl groups; i.e., they act as both hydrogen bond donor and acceptor and possibly interact with both the OH and O sites. In addition, the benzyl hydroxyl group is structurally more flexible than phenolic hydroxyl groups, and this flexibility possibly results in high retentivity of HBA. In contrast, the interaction of polyethers with the ice surface is simpler than that for phenols, because polyethers should interact only with the OH sites. As noted already, the number of hydrogen bond formation is an important factor to govern the overall interaction energy: the more interacting points, the stronger the interaction. However, it should be
Table 2. Parameters for Describing Retention Changes with THF Concentration n solute
KTHF ) 2
KTHF ) 5
KTHF ) 50
DB18C6 DB24C8 HBA resorcinol hydroquinone
1.41 3.05 2.28 2.44 2.54
1.26 2.60 2.07 2.20 2.28
1.16 2.45 1.97 2.05 2.14
noted that the interaction is entropically enhanced with increasing number of polar groups involved in a molecule, even if the number of interacting points is not varied. It must, thus, be reasonable that for steric reasons, a part of the ethereal oxygen atoms, rather than all of them, must be involved in the simultaneous hydrogen bond formation. These factors enhance the retentivity of crown ethers with larger ring sizes. Further Considerations of Retention Mechanisms. As described above, the retention on the ice surface is principally caused by the hydrogen bond formation. As shown in Figure 4, THF acts as a competitor both for phenols and for polyethers. Changes in the retention of crown ethers with THF concentration should simply be interpreted by the competitive adsorption with THF, because both crown ethers and THF are adsorbed onto the OH sites. The adsorption of THF on an open OH site is given by
ΓOH-THF [THF]ΓOH
KOH-s
s + n-OH {\} -OHn-s
k)
)
ΓOHn - s [s]ΓOHn
AΓOHn - s V[s]
AΓtOHnKOH-s V(1 + KTHF[THF])n
KO-s
s + mO {\} Om-s
k)
where Γ denotes the surface concentration. If n open sites are necessary for solute retention, its retention factor can be described by
KOH-s )
KOH-s
s + nOH {\} OHn-s
THF added in the mobile phase affects the adsorption on the OH sites, but it does not directly influence the adsorption on the O sites. Thus, the retention factor of phenols is given by
KTHF
THF + -OH {\} -OH-THF KTHF )
values calculated by assuming KTHF to be in the range 2-50 M-1 (∼38-94% active sites are occupied by THF in 2%(v/v) ∼0.3 M solution). Since both KTHF and n similarly affect the results of curve-fitting, likely values were first assumed for KTHF, and then n was optimized; otherwise, estimated values should have no physical meanings. The n values for crown ethers are varied with ring sizes; n ) 1.2-1.4 for DB18C6 and 2.5-3.0 for DB24C8. The n value for DB24C8 suggests that two to three ethereal oxygen atoms act for its adsorption on water-ice, and in turn, most of the oxygen atoms contained in this molecule do not take part in the interaction. Even in such cases, an entropic effect should result in the preferable interaction of the compounds having more polar atoms. This effect may also lead to measurable retention of DB18C6, despite a smaller n value. However, the peak width of DB18C6 is obviously larger than that for other solutes (see Figure 3), and thus, a different mechanism may be involved in its retention process. In contrast, the situation is more complex for phenolic solutes. The adsorption of phenols on the ice surface must be described by the following two equilibriums:
(1)
where A and V are the surface area of the stationary phase and the void volume of the column, ΓtOH is the total surface concentration of the OH sites, and KOH-s is the adsorption constant of a solute to the adsorption sites. The adsorption of THF is expected to be rather weak, because no retention has been confirmed for other monoethers. However, the large dependence of the retention of solutes on the THF concentration implies that a substantial part of the surface active OH sites are occupied by the THF molecules over the concentration studied ranges in Figure 4. Table 2 lists n
{
}
ΓtOHnKOH-s A + KO-sΓOm V (1 + K [THF])n THF
(2)
The k-[THF] plots for phenols imply that the contribution from the second term is not very important because the dependence of the retention on the THF concentration is similar to that obtained for the crown ether. The OH sites are exposed to the surroundings on the basal plane of the ice crystal (the (0001) plane, see Figure 1), and as a result, a solute more easily approaches the OH sites than the O sites. Thus, the adsorption of phenols on the O sites appears to be less important than that on the OH sites in the determination of their retention on the ice stationary phase. The results of curve-fitting for phenols are also listed in Table 2. Interestingly, n is almost 2 for phenolic solutes, irrespective of assumed KTHF values, suggesting that two OH sites are necessary for their retention on the ice stationary phase. This is consistent with the fact that the retention of monohydroxylic compounds was much weaker than that of dihydroxylic solutes. Polyethers were also studied as mobile phase additives; triethyleneglycol dimethyl ether (triEGDE), tetraethyleneglycol dimethyl ether (tetraEGDE) were examined. These polyether additives are expected to be adsorbed onto the ice surface and more strongly affect solute retention than THF. The results obtained with tetraEGDE are shown in Figures 5. Both triEGDE (results are not shown) and tetraEGDE should interact with the OH sites on the ice surface and, thus, interfere with the adsorption of polyether solutes. As expected, the addition of these polyethers Analytical Chemistry, Vol. 78, No. 12, June 15, 2006
4159
Figure 5. Effect of addition of tetraEGDE on solute retention. Mobile phase, 3% (v/v) THF in hexane. Curves are for a guide for the eye.
in the mobile phase reduced the retention for crown ethers. Although basic trends were identical, more marked effects were seen for tetraEGDE than for triEGDE, being consistent with the stronger adsorption of longer polyether chains. In contrast, it is very interesting that the retention of phenolic solutes increases with increasing concentration of the polyether additives in the mobile phase. Although, as stated above, phenolic compounds can interact either with the O sites or with the OH sites on the waterice surface, the interaction with the O sites on the water-ice surface is less important. However, the O sites, which are newly created by the adsorption of polyether additives on the ice surface, act as the adsorption sites for phenols. Again, not all of the ethereal oxygen atoms necessarily participate in the surface adsorption, and therefore, the adsorption of polyethers allows the production of the free O sites, which are exposed to the surroundings, in comparison with the O sites originally aligned on the water-ice surface. Long-chain alcohols, that is, octanol and dodecanol, were also tested as mobile-phase additives. Both of them caused decreases in solute retention. Figure 6 illustrates the effects of dodecanol on the retention of selected solutes. Obviously, the added alcohol competes with phenolic and polyether solutes for the same adsorption sites on the ice surface. This result, thus, supports the above inference that the interaction with the ice OH site is more important, even for phenolic solutes. In conclusion, we have shown some examples of ice chromatographic separation and discussed the retention mechanisms.
4160 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006
Figure 6. Effect of addition of dedecanol on the retention of hydroquinone, DB24C8, and HBA (p-hydroxybenzyl alcohol). Mobile phase, 3% (v/v) THF in hexane. Curves are for a guide for the eye.
The importance of ice chromatography is summarized as follows: (1) its high ecological adaptation, (2) potential usefulness as a new stationary phase, (3) elucidation of the molecular interactions on the water-ice surface, and (4) capability of discussion on the physicochemical nature of water-ice surface itself. At the present stage, the identical ice stationary phase cannot be prepared. The surface area; particle sizes; and, in turn, the retention ability are different for every preparation. This generates difficult rigorous discussions on the physicochemical nature of the ice surface and possibly results in poor separation performance. More effort should be made to overcome this problem. Apart from this point, we believe that ice chromatography will find more applications because inexpensive stationary phase preparation must be attractive for preparative use of chromatography. ACKNOWLEDGMENT This work has been supported in part by a Grant-in-Aid for an Exploratory Research from the Ministry of Education, Culture, Sports, Science and Technology and by a research fund from the Sekisui Chemical Foundation. Received for review February 7, 2006. Accepted April 3, 2006. AC0602470
