Hydrophilic interaction chromatography (HILIC) provides stronger retention for small polar compounds as well as different selectivity than reversed-phase liquid chromatography (RPLC) [1], [2], [3], [4]. Polar compounds gain retention through hydrophilic partitioning between the mobile phase and an immobilized water-rich liquid layer on the packing surface, surface adsorption through polar interactions with polar functional groups on stationary phases, and/or electrostatic interactions if both the analytes and stationary phases contain charged groups [5], [6], [7], [8]. The observed retention (k’obs) is the sum of contributions of all the retention processes:kobs′=kpar′+kads′+kelec′where k’par, k’ads and k’elec represent the retention factors contributed by hydrophilic partitioning, surface adsorption and electrostatic interactions, respectively. In the hydrophilic partitioning process, the retention is governed by distribution coefficient (K) and phase ratio (Ф) [9]:kpar′=K∅

In chromatographic separations, selectivity is characterized by selectivity factor as defined by the ratio of retention factors:α=k2′k1′

For non-ionized compounds without electrostatic interactions in HILIC, the selectivity factor can be expressed by combining Eqs. 1-3:α=k2′k1′=K2∅+k2,ads′K1∅+k1,ads′where K1 and K2 are the partitioning coefficients, k’1,ads and k’2,ads represent the retention factors contributed by surface adsorption. Eq. 4 indicates that selectivity is potentially dependent on both hydrophilic partitioning and surface adsorption in HILIC. Selectivity tests such as the Tanaka's selectivity test are commonly used for column comparison in RPLC [10,11]. In HILIC, selectivity has been investigated by comparing the elution patterns or the selectivity factors of selected probe compounds on various polar stationary phases [12], [13], [14]. These studies provide a direct comparison of selectivity among different stationary phases; however, it is difficult to generalize the findings from different studies since observed selectivity may vary depending on the selected probe compounds. To standardize selectivity comparison in HILIC, Ikegami and co-workers proposed a series of selectivity tests using a group of test compounds [15]. Selectivity factors are measured by the retention factor ratio of a pair of selected compounds (Eq.3). For example, hydroxy selectivity is measured by uridine and 2-dexoyuridine, and methylene selectivity by uridine and 5-methylurudine. Configurational selectivity is probed by two pairs of isomers, vidarabine/adenosine, and 4-nitrophenyl-α/β-D-glucopyranoside. Regioselectivity is evaluated by regio-isomer pairs, 2’-deoxyguanosine and 3’-deoxyguanosine, and theophylline and theobromine. These selectivity tests have been applied to differentiate monomeric and polymeric stationary phases commonly used in HILIC [16]. The Ikegami selectivity tests provide a tool to directly compare different stationary phases in HILIC; however, the selectivity factors based on observed retention factors do not differentiate different mechanisms (i.e., partition, adsorption, and/or electrostatic interactions) involved. Therefore, the selectivity data does not provide any insights into the driving force for selectivity and is not directly linked to the characteristics of the stationary phase.

One approach to investigate how selectivity is related to the retention mechanisms is to understand how much each potential mechanism contributes to the observed retention. Eq. 4 indicates that if the relative contributions of hydrophilic partitioning (k’par) and surface adsorption (k’ads) are known for a pair of non-ionized compounds, it would be feasible to determine whether the selectivity is controlled by hydrophilic partitioning or surface adsorption. Furthermore, the selectivity factors can also be linked to the phase ratio, which is related to the stationary phase. Therefore, the selectivity data can be interpreted based on stationary phase properties and better understood in terms of the differences among various columns.

We have developed a methodology to quantitate relative contributions of hydrophilic partitioning and surface adsorption to the retention of non-ionized polar compounds in HILIC [17]. This methodology is based on establishing linear correlation between the observed retention factor and phase ratio. The slope of the linear regression line provides the partitioning coefficients of polar compounds in the hydrophilic partitioning process, and the intercept represents the retention contribution of surface adsorption. Previous research shows that the phase ratio can be varied by changing the salt concentration in the mobile phase. Here the phase ratio (Ф) is defined as the volume of the adsorbed water layer and the mobile phase, and can be experimentally determined using the elution volume of toluene in pure acetonitrile (VACN) and mobile phases (VM) [18]:∅=VACNVM−1

In this study, three uridine test compounds (uridine, 2-deoxyuridine, and 5-methyluridine) for hydrophilic selectivity in the Ikegami selectivity tests were selected. Two pairs of positional isomers (theophylline/theobromine, 2’-deoxyguanosine/3’-deoxyguanosine) and one pair of configurational isomers (adenosine/vidarabine) were also included in this study. The epimer pair in the Ikegami selectivity test (4-nitrophenyl-α/β-D-gluocopyranoside) was not evaluated due to sample availability issues. The main objective of this study was to identify the major mechanism that controls selectivity in the Ikegami selectivity tests. This would help explain the selectivity test results and better understand the difference among HILIC columns.