Enantioselective HPLC of Liquid Crystals Bearing a Small and Flexible 2‐Methylbutyl Chiral Center

ABSTRACT

The conformationally flexible 2‐methylbutyl structural motif represents a small chiral center, which cannot engage in polar interactions with chiral stationary phases (e.g., hydrogen bonds), which often play a key role in enantioselective recognition. In this contribution, racemic liquid crystalline materials sharing the 2‐methylbutyl group were analyzed by enantioselective high‐performance liquid chromatography, and the studied analyte set also included diastereomeric compounds with additional chiral centers based on the lactate unit. As the non‐chiral parts of the analytes are rigid and bulky, this disbalance in size and flexibility further complicates enantioseparation of such materials. The separations were done under either polar organic mode or reversed‐phase liquid chromatography conditions to enhance the less polar interactions with the chiral stationary phases. Two polysaccharide chiral stationary phases were employed—coated tris(5‐chloro‐2‐methylcarbamate) amylose and immobilized tris(3‐chloro‐5‐methylphenylcarbamate) amylose. Baseline enantioseparation of all racemic samples was accomplished in less than 9 minutes. Effects of analytes’ structural variations on enantioselectivity are discussed, and the differences between results on the two chiral stationary phases are assessed.

Enantioseparation of liquid crystalline compounds with a 2‐methylbutyl chiral center is studied on two amylose chiral stationary phases in polar organic and reversed‐phase mode. The effects of mobile phase composition and of the solutes‘ structure on their chromatographic behavior is discussed. All compounds were baseline separated.

1. Introduction

Thermotropic liquid crystalline materials (LCMs) are promising materials capable of self‐assembling into various structural forms (mesophases) within a defined temperature range between the isotropic and crystalline phases [1]. Depending on their chemical structure and mesophase type, LCMs can respond to external stimuli, such as electromagnetic radiation or electric fields, enabling their application as functional materials in fields extending far beyond liquid crystal displays, including medicine [2], polymer science [3], soft robotics [4], photonics [5], and security technologies [6]. Chiral LCMs constitute a major area of LCM research because molecular chirality promotes the formation of additional mesophase types and mesomorphic properties that are absent in achiral systems. Molecular chirality in tilted smectic phases can induce ferroelectricity or helicity [7]. The chiral purity of these materials plays a crucial role in determining their properties [8, 9, 10], thereby driving the need for enantioseparation methods for chiral purity control and assessment.

Liquid chromatography is a prominent separation method for determining the content of both achiral and chiral impurities in studied materials [11]. Enantioselective chromatography with a chiral stationary phase (CSP) enables direct separation of enantiomers, with the derivatives of amylose or cellulose polysaccharide being the most widely used CSPs, which are commercially available [12]. Chlorinated CSP Lux Amylose‐2 (coated tris(5‐chloro‐2‐methylphenylcarbamate) amylose) and Lux i‐Amylose 3 (immobilized tris(3‐chloro‐5‐methylphenylcarbamate) amylose) were selected for this work based on previous research showing that chlorinated polysaccharide derivatives often offer higher enantioselective potential for separation of chiral LCMs [13, 14] or similar compounds [15]. Unfortunately, neither coated tris(3‐chloro‐5‐methylphenylcarbamate) amylose nor immobilized tris(5‐chloro‐2‐methylphenylcarbamate) amylose are currently commercially available, preventing direct assessment of chiral selector anchoring technology on the enantioseparation of our analytes.

This study focuses on LCMs sharing the same chiral fragment—a flexible 2‐methylbutyl moiety derived from 2‐methylbutanol. This chiral alcohol (in S configuration) is an important structural fragment used in materials science. Its primary importance stems from its status as a relatively cheap, widely commercially available, and naturally sourced “chiral pool” building block due to being a product of fermentation. The location of the chiral center at the C2 position adjacent to the point of attachment does not require enantioselective procedures of synthesis [16]. A racemic mixture (or R form) is usually synthesized in much smaller amounts as a reference material for evaluating the influence of chiral purity. For structures of materials used in this work, see Figure 1. The first group consists of photo sensitive azo materials, the second of materials derived from lactic acid, and the third of photo sensitive cinnamoyl derivatives also containing an internal lactic acid unit. The chiral fragment (2‐methylbutyl) has no heteroatom vicinal to the asymmetric carbon, which makes the chiral moiety low polar and restricts the potential for polar non‐covalent interactions (e.g., hydrogen bonds) with CSPs. Thus, the enantioseparation of these materials is problematic in the normal phase (NP) liquid chromatography environment, which favors polar interactions [17] and is expected to be more plausible in aqueous/more polar mobile phases, where (given the hydrophobic nature of the analytes) the solvophobic effect should promote adsorption of the analytes onto the CSP [18].

FIGURE 1.

Molecular structures of studied racemates: azo compounds G₁ and G₂, compounds Z12/*, ZL12/* and ZLL12/*, and two racemates of the compound 8SHL* differing in configuration of the lactate. The (2‐methylbutyl) chiral centers are highlighted in color.

In the past, partial enantioseparation of compounds G₁ and G₂ was only achieved under NP conditions after the samples were irradiated and converted into their Z isomers that are more polar than E isomers [19]. Compounds G₁, G₂, and Z12/* have been used in a work focused on the enantioselective potential of various CSPs in the environment of supercritical fluid chromatography (SFC), and none of these compounds could be separated under any of the tested conditions [20]. This result is not unexpected because the SFC environment behaves similarly to NP chromatography by favoring polar interactions, even though the retention mechanisms can differ substantially [21]. In contrast, G₁ and G₂ materials could be separated under reversed‐phase (RP) conditions, but with excessively long retention times [22]. Meanwhile, Z12/* was successfully resolved using polar organic mode (POM), both with conventional solvents [14] and with environmentally friendly (“green”) solvents [23].

The aim of this work is to survey enantioseparation of racemic materials with a 2‐methylbutyl chiral center under either POM or RP conditions and to achieve baseline enantioseparation of all analytes. The structural similarity of the compounds will be used to discuss the relationship between analyte structure and its chromatographic behavior.

2. Materials and Methods

2.1. Instrumentation

Chromatography was performed on an Agilent 1260 Infinity II HPLC system consisting of a quaternary gradient solvent pump, an autosampler, a column thermostat, and a diode array detector with continual spectrum acquisition. OpenLab CDS software was used for system control and data analysis. Chromatograms were recorded at various wavelengths corresponding to the absorption maximum of each compound. Columns Lux i‐Amylose‐3 (immobilized amylose tris(3‐chloro‐5‐methylphenylcarbamate)) and Lux 2‐Amylose (coated amylose tris(5‐chloro‐2‐methylphenylcarbamate)), both with 3 µm silicagel particles and column dimensions 150 × 4.6 mm i.d. were obtained from Phenomenex (USA). All experiments were performed at 40°C with a mobile phase flow rate of 1 mL∙min⁻¹ and 3 µL sample injection volume.

2.2. Materials

Azo materials G₁ and G₂ were synthesized at the Faculty of Chemistry, University of Wrocław (Poland) [19]. The compounds Z12/*, ZL12/*, ZLL12/* and 8SHL*, were synthesized and studied at the Institute of Physics of the Czech Academy of Sciences, Czechia [22, 24, 25, 26]. All samples were prepared at a concentration of 0.2 mg·mL⁻¹ by dissolving both the racemic mixture and the S enantiomer in pure acetonitrile, with samples of pure S enantiomers used to assign enantiomer elution order (EEO). The only exception is 8SHL* which is only available in the form of (S,S), (S,rac), (rac,S), and (rac,rac) mixtures, and the results of (R,rac) and EEO had to be extracted from the data post‐analysis. Solvents used as MP components were acetonitrile Chromasolv (ACN; gradient grade, ≥99.9%) purchased from Altium International (Czech Republic); methanol HiPerSolv (MeOH; gradient grade, ≥99.8%) and 1‐propanol (PrOH; for HPLC, ≥99.8%) purchased from VWR International (Czech Republic); 2‐propanol Rotisolv (IPA; for HPLC, ≥99.9%) purchased from P‐LAB (Czech Republic); and ethanol LiChrosolv (EtOH; gradient grade, ≥99.9%) and 1‐butanol (BuOH; for HPLC, ≥99.7%) purchased from Merck Life Science (Czech Republic). Deionized water was prepared on the ULTRAPUR water purification system obtained from Watrex Praha (Czech Republic).

3. Results and Discussion

The studied samples were initially analyzed in mobile phases: acetonitrile (ACN), methanol (MeOH), ethanol (EtOH), 1‐propanol (PrOH), 1‐butanol (BuOH), and 2‐propanol (IPA) and mixtures of acetonitrile with each of the alcohols in 90/10, 80/20, 20/80, and 10/90 (v/v) ratio to cover the POM conditions; and in ACN/water mixtures in 97.5/2.5, 95/5, and 90/10 (v/v) ratio for the RP experiments. Additional mobile phases (or non‐default flow‐rate or column temperatures) were used in case they served to achieve baseline resolution or shorten excessive retention times. For complete results, see Tables S1 and S2 in Supplementary Material. The compounds 8SHL*, ZL12/*, and ZLL12/* have two or three chiral centers, but for ZL12/* and ZLL12/* racemic mixtures are available only with the S‐ configuration of the incorporated lactic acid moieties. However, for 8SHL* with two chiral centers, all following forms are available: (rac,rac), (S,rac), (rac,S), and (S,S); enabling studying 8SHL*(S,rac) and 8SHL*(R,rac) as two separate analytes.

3.1. Mobile Phase Flow‐Rate and Column Temperature

The number of theoretical plates was found to increase with decreasing flow‐rate as is usual on polysaccharide CSPs with bulky aromatic analytes [23, 27, 28], with the trend only reversing at very little flow‐rates, see Figure 2a. As a result, the resolution can be notably improved by lowering the flow‐rate down to 0.2 mL∙min⁻¹, but the analysis is prolonged accordingly, and thus it is better to start with a higher flow‐rate for the screening runs. Regarding the column temperature, the efficiency decreased with decreasing the temperature, see Figure 2b. Changing the column temperature might (apart from its effect on efficiency) lead to dramatic changes in enantioselectivity, including its (dis)appearance [14], and the temperature optimization is thus often much less straightforward than flow‐rate optimization. In Figure 2b the increase in selectivity at lower temperatures leads to increased resolution despite the loss of efficiency.

FIGURE 2.

A Van Deemter curve of S (●) and R (°) enantiomer of G₁(rac) sample in neat ACN displayed as the height equivalent of a theoretical plate (HETP). B Effect of column temperature on the plate number (N), enantioresolution (R) and enantioselectivity (α) of ZLL12/* in 95/5 ACN/IPA. Measured on Lux Amylose‐2 column.

3.2. Mobile Phase Composition

For all materials, the POM revealed a U‐shaped retention trend where the retention in mixtures of acetonitrile and alcohol is weaker than in either of the neat solvents. This is a result of a gradual transition from one governing retention mechanism to another (from more polar interactions in neat acetonitrile to less polar interactions and solvophobic effects in neat alcohols. All these interactions are weakened in mobile phases containing both solvents, resulting in shorter retention times and the U shape. This trend is associated with analysis of aromatic analytes on multi‐modal aromatic stationary phases such as polysaccharide derivatives [14, 29, 30, 31] and may or may not include change of EEO, see Figure 3.

FIGURE 3.

U‐shaped retention trend of A ZL12/* without EEO reversal and B Z12/* with EEO reversal on the Lux i‐Amylose‐3 column. Results plotted as the retention factor (log k, purple) of the S enantiomer (●) and the R enantiomer (°) and resulting enantioselectivity (α, green).

From the alcohols used in POM separations, BuOH was the strongest eluent and MeOH the weakest as the retention decreased with increasing alkyl chain length. This is probably caused by the longer alcohols promoting a weaker solvophobic effect and hindering π‐π and/or dipole‐dipole interactions between the solute and the CSP [32]. In MeOH, the LCMs have a notably worse peak shape that can be improved by the addition of ACN, see Figure S1 in Supplementary material. This can be explained by the solvophobic effect since the solubility of LCMs in MeOH is very low. Neat ACN was always a stronger eluent than MeOH, but (depending on the solute) could be either stronger or weaker eluent than other alcohols, for illustration, see Figure S2 in Supplementary Material.

For compilation of the highest resolutions and enantioselectivities obtained on both stationary phases, see Table 1 and Table 2. Longer alcohols did not yield unique enantioselectivity (compared to MeOH) on either of the CSPs and no baseline resolution at all on Lux Amylose‐2. On the other hand, mobile phases with IPA brought the only baseline enantioseparation of ZLL12/* (S,S,rac) and its combination with MeOH lead to higher resolution of ZL12/* (S,rac) and 8SHL* (S,rac) than with either of the alcohols alone. Polarity of linear alcohols decreases with increasing alkyl size, and the longer alcohols are thus better at weakening low polar interactions and worse at enhancing solvophobic interactions. At the same time, the linear alkyl moieties are flexible and less bulky than the branched isopropyl group of IPA and are therefore likely to affect the conformation of the CSP to a lesser extent than isopropanol [33]. These combined effects may explain the underperformance of linear alcohols and discourage their further use for enantioseparation of LCMs, but it should be noted that these outcomes are solute‐specific and not universal [34].

TABLE 1.

Best obtained enantioresolutions of racemic samples obtained at retention times under 30 min. The mobile phases are denoted as 100 ACN for neat acetonitrile, MeOH/EtOH/IPA and water for either neat solvent or its mixture with acetonitrile, and special for other mobile phases specified in table footnotes. (x) means the enantioresolution was too low to be calculated.

Chiral stationary phase	Mobile phase	G1 (rac)	G2 (rac)	Z12/ a (rac)	ZL12/ a (S,rac)	ZLL12/ a (S,S,rac)	8SHL a (S,rac)	8SHL a (R,rac)
Lux i‐amylose‐3	100 ACN	0	0	2.80	0	0	0	1.81
	MeOH	0.80	0.38	1.65	2.30	0.40	1.29	1.61
	EtOH	0	0	1.89	1.60	x	0.21	1.65
	PrOH	0	0	x	0.61	0	0	x
	BuOH	0	0	x	0	0	0	0.35
	IPA	0	0	0	1.46	0	0.82	1.46
	water	0	0	1.30	0	0.66	0	1.19
	special				3.65 a	0.74 b	2.22 a
Lux amylose‐2	100 ACN	1.43	1.28	1.18	0	0	0	0
	MeOH	0.62	0.73	2.23	0	0	0	x
	EtOH	x	0	x	0	0	0	0
	PrOH	0	0	x	0	0.38	0	0
	BuOH	x	0	1.02	0	0	0	0
	IPA	0	0	0	0	2.46	0	0
	water	2.97	2.85	2.16	0	0	0	0

^a 50/50 MeOH/IPA.

^b 90/5/5 ACN/MeOH/water.

TABLE 2.

Highest obtained enantioselectivity of racemic samples obtained in screening mobile phases. The mobile phases are denoted as 100 ACN for neat acetonitrile and alcohol or water for either neat solvent or its mixture with acetonitrile.

Chiral stationary phase	Mobile phase	G1 (rac)	G2 (rac)	Z12/* (rac)	ZL12/* (S,rac)	ZLL12/* (S,S,rac)	8SHL* (S,rac)	8SHL* (R,rac)
Lux i‐amylose‐3	100 ACN	1	1	1.140	1	1	1	1.132
	MeOH	1.038	1.028	1.085	1.124	1.042	1.065	1.060
	EtOH	1	1	1.058	1.118	1.026	1.035	1.076
	PrOH	1	1	1.033	1.079	1	1	1.056
	BuOH	1	1	1.028	1	1	1	1.055
	IPA	1	1	1	1.096	1	1.058	1.037
	water	1	1	1.057	1	1.060	1	1.066
Lux amylose‐2	100 ACN	1.133	1.131	1.208	1	1	1	1
	MeOH	1.106	1.164	1.405	1	1	1	1.086
	EtOH	1.066	1	1.060	1	1	1	1
	PrOH	1	1	1.051	1	1.077	1	1
	BuOH	1.070	1	1.098	1	1	1	1
	IPA	1	1	1	1	1.281	1	1
	water	1.163	1.166	1.167	1	1	1	1

In RP mode, for the majority of solutes, the retention increased markedly even with small amounts of water, and no HILIC‐like behavior was observed, as is usual for LCMs [13]. Since the LCMs are hydrophobic in nature, the presence of water in the mobile phase often leads to a solvophobic effect, favoring the solute partitioning onto the CSPs rather than staying solvated in the mobile phase. The only exception was 8SHL*(R,rac) on Lux i‐Amylose‐3, whose retention slightly decreased in 97.5/2.5 ACN/water compared to neat acetonitrile before increasing in 95/5 ACN/water. Interestingly, this did not happen for the 8SHL*(S,rac), which rules out non‐enantiospecific interactions (e.g., with the silica support) and solvating effects of the mobile phase as the possible cause for this behavior since those would be identical for any 8SHL* isomer. Thus, the possible cause may be competition for hydrogen‐bonding sites inside the chiral groove of the amylose helix, where the (R,rac) isomers have better fit than the (S,rac) isomers.

3.3. Analyte Structure

Molecules G₁ and G₂ differ only in the presence of one oxygen atom at the opposite end of the molecule form the chiral center. These two racemates could not be separated under SFC and NP conditions, probably due to the absence of hydrogen‐bond donoring sites and only being weak hydrogen‐bond acceptors. Correspondingly, partial enantioseparation of irradiated Z isomers under NP conditions (for which the azo group is more polar and exposed to interactions) was achieved [19]. In acetonitrile and acetonitrile‐rich MPs, G₁ is retained stronger than G₂ on both CSPs, and the difference is highest with aqueous MPs, which corresponds to the simple C8 driving stronger hydrophobic and π‐alkyl interactions with the CSP, as opposed to the ether O‐C8 terminal chain of G₂, see Figure S3 in Supplementary Material for illustration. This affects only the non‐enantiospecific retention, as the enantioselectivity was found to be nearly identical for both LCMs under RP conditions.

The retention of Z12/* is notably higher than for G₁ and G₂, mostly due to the additional benzoyl group and partly also due to the longer terminal alkyl. The enantioselectivity of both CSPs towards Z12/* is much higher, and Z12/* is the only material for which a reversal of the EEO was found, induced by MP composition, see Figure 3b. Based on our previous results [14] and the fact that the results of no other solute from our set show EEO reversal on the 2‐methylbutyl chiral center, we assume that the presence of both a carbonyl group and a rigid (aromatic) ring in the vicinity of the asymmetric carbon is crucial for the occurrence of EEO reversal.

ZL12/*(S,rac) is structurally similar to 8SHL*(S,rac). ZL12/* is longer retained due to the additional benzoyl group and partly also due to the longer terminal alkyl. However, these structural differences are remote from the chiral center and the enantioselectivity trends are almost identical for the two analytes on both CSPs, for illustration see Figure 4. On the contrary, ZLL12/*(S,S,rac) contains one more S‐lactate unit than ZL12/*(S,rac) next to the chiral center. This difference caused only slightly longer retention times; however, it yields opposite EEO (under all conditions where it could be determined) together with disparate trends of enantioselectivity and with the two compounds successfully separated on different CSPs, see Figure 4. This figure shows the baseline enantioseparation of 8SHL*(S,rac), ZL12/*, and ZLL12/*.

FIGURE 4.

Analysis of 8SHL*(S,rac), ZL12/*(S,rac), and ZLL12/*(S,S,rac) on Lux i‐Amylose in 45/45/10 MeOH/IPA/ACN and on Lux Amylose‐2 in 90/10 IPA/ACN. Temperature: 40°C.

In the case of the 8SHL* compound, the (rac,rac) mixture was available, which enabled to the survey of the effect of configuration on the lactate‐derived chiral center on the enantioselectivity of the small chiral center. On CSP1 the (S,rac) and (R,rac) mixtures were separated nearly exclusively, see Figure 5a. The (S,rac) mixture was separated in alcohol‐rich MPs, under RP conditions, and in a mixture of ACN/IPA. On the other hand, the (R,rac) mixture was separated in acetonitrile‐rich MPs. Simultaneous baseline separation of all four isomers constituting the (rac,rac) sample could not be achieved. Regarding the lactate chiral center (see *(1) in Figure 5a), the enantioseparation is much easier—high resolution and enantioselectivity values are obtained even with retention factors k∼1. Additionally, reversals of EEO were found on both CSPs, see Figure 5b. The S, S and S, R enantiomers elute first in mobile phases of higher relative permittivity (e.g., aqueous mobile phases) and last in mobile phases of lower relative permittivity (e.g., neat propanol), similarly to previous results of lactic acid‐based chiral LCMs [14].

FIGURE 5.

A Enantioseparation results of 8SHL*(rac,rac) in ACN/MeOH mobile phases on Lux i‐Amylose‐3, demonstrating different selectivity for the second chiral center based on the configuration of the first chiral center. B Enantiomer elution orders on the chiral center (1) depending on the type and amount of cosolvent (B) in the acetonitrile. The order on the chiral center (2) is not specified.

3.4. Stationary Phase

The two CSPs, Lux Amylose‐2 and Lux i‐Amylose‐3, are both chlorinated amylose derivatives that differ in the position of methyl (5‐chloro‐2‐methylphenylcarbamate vs. 3‐chloro‐5‐methylphenylcarbamate) and in the anchoring of the chiral selector onto the silica gel particles (coated vs. immobilized). For the LCMs in this study, the two columns showed notable complementarity, with 4 out of 7 racemates baseline separated on Lux Amylose‐2 and 4 out of 7 racemates baseline separated on Lux i‐Amylose‐3, with only Z12/* separated on both (see Table 1).

Interestingly, the G₁ and G₂ racemates could only be baseline separated on the coated Lux Amylose‐2. Despite similar or stronger retention, the enantioselectivity is completely absent on the immobilized Lux i‐Amylose‐3, see Figure 6. In the past, separation of G₁ and G₂ was tried on immobilized and coated tris(3,5‐dimethylphenylcarbamate) amylose CSPs and brought no quantifiable resolution (unpublished results, see Table S3 in Supplementary Material). It might appear that the covalent immobilization hinders the enantioselectivity of polysaccharide CSPs towards these analytes, however, this cannot be confirmed, as a coated analogue of Lux i‐Amylose‐3 is not commercially available. Moreover, the successful enantioseparation of the smallest analytes on the tris(5‐chloro‐2‐methylphenylcarbamate) amylose might also be explained by smaller chiral cavities of the polysaccharide due to the 2‐methyl substituent [35], where the hindered cavity might be beneficial for chiral recognition of small analytes.

FIGURE 6.

Analysis of racemic G₂ on Lux i‐Amylose‐3 (left) and Amylose‐2 (right) in neat acetonitrile and in acetonitrile/water mobile phases.

4. Conclusion

The series of chiral LCMs were studied by enantioselective HPLC, and all the studied racemic materials were successfully resolved. Despite containing the identical chiral moiety (2‐methylbutyl), the materials required various separation conditions to acquire baseline resolution, including different stationary phases and/or different chromatographic modes (some separated in POM, some in RP). This shows the influence of the overall analyte structure on its enantioseparation. For the 8SHL* compound with two chiral centers, the chiral configuration on the first asymmetrical carbon strongly affected the enantiorecognition on the second, with the (S,rac) and (R,rac) racemates exhibiting disparate chromatographic behavior. For the smallest molecules (G₁, G₂, and Z12/*), the coated Lux Amylose‐2 brought the best (or the only) successful separation, while the immobilized Lux i‐Amylose‐3 was overall better for the larger analytes. Methanol and 2‐propanol yielded partially complementary results, while longer linear alcohols did not provide any results that would encourage their use in polar organic mode. This work provides a concise overview of the enantioseparation of these analytes and may serve as a useful background for the future development of enantioselective methods for structurally similar analytes.

Author Contributions

Terézia Jurkovičová: writing – original draft, investigation, conceptualization, writing – review and editing. Martin Cigl: writing – review and editing, resources, funding acquisition, formal analysis. Věra Hamplová: supervision, funding acquisition, writing – review and editing, resources. Vladimíra Novotná: funding acquisition, writing – review and editing, resources, project administration. Petra Vaňkátová: writing – original draft, investigation, visualization, conceptualization, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: jssc70552‐sup‐0001‐SuppMat.docx

Data Availability Statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.