Determination of Enantiomeric Excess by Solid-Phase Extraction Using a Chiral Metal-Organic Framework as Sorbent

Metal-organic frameworks (MOFs) have recently attracted considerable attention because of their fascinating structures and intriguing potential applications in diverse areas. In this study, we developed a novel method for determination of enantiomeric excess (ee) of (±)-1,1′-bi-2-naphthol by solid-phase extraction (SPE) using a chiral MOF, [Co(l-tyr)]n(l-tyrCo), as sorbent. After optimization of the experimental conditions, a good linear relationship between the ee and the absorbance of the eluate (R2 = 0.9984) was obtained and the standard curve was established at the concentration of 3 mmol L−1. The ee values of (±)-1,1′-bi-2-naphthol samples can be rapidly calculated using the standard curve after determination of the absorbance of the eluate. The method showed good accuracy, with an average error of 2.26%, and is promising for ee analysis.


Introduction
Optical purity is important in many industries, particularly in the development and production of pharmaceuticals. In general, one enantiomer of chiral drugs is safe and effective, while the other is may be ineffective or even toxic [1]. Consequently, the determination of enantiomeric excess (ee) of chiral compounds is very important. At present, chiral chromatographic methods, such as high performance liquid chromatography (HPLC) and gas chromatography (GC) using chiral stationary phases (CSPs), are most commonly used to determine ee values. However, these methods have some drawbacks, such as the high cost of chiral columns and long analysis times. Therefore, the development of simple and fast methods for ee analysis has great value. A number of techniques for rapid ee analysis have been developed, including circular dichroism [2,3], UV and colorimetric methods [4,5], fluorescence spectroscopy [6,7], mass spectrometry [8], and molecularly imprinted polymer based assays [9].
Solid-phase extraction (SPE) is a valuable sample pretreatment technique that enables concentration and purification of analytes from complex matrices. It has been widely used for its simplicity, rapidity, low cost, and compatibility with different detection techniques in both on-line and off-line modes [10,11]. The adsorbent is the core in SPE. High enrichment efficiency and selectivity of SPE are particularly dependent on the adsorbent. Thus, development of novel and efficient adsorbents for SPE is an ongoing research interest. selectivity of SPE are particularly dependent on the adsorbent. Thus, development of novel and efficient adsorbents for SPE is an ongoing research interest.
The chiral 3D MOF, [Co(L-tyr)]n(L-tyrCo), composed of metal dimers (Co II ) linked via L-tyrosine ligand (Figure 1a) [40]. In this paper, we report the determination of ee values of (±)-1,1′-bi-2-naphthol ( Figure 1b) by SPE using [Co(L-tyr)]n(L-tyrCo) as the adsorbent. The experimental results showed that the ee values of (±)-1,1′-bi-2-naphthol were linearly correlated with the absorbance of the eluates, and a standard curve could be established for ee analysis. The ee values of (±)-1,1′-bi-2-naphthol samples can be rapidly measured using the standard curve by determining the absorbance of the eluates after SPE. The method has advantages that include of good accuracy, low cost, simple operation, and suitability for batch analysis.

Characterization of the Synthesized [Co(L-tyr)]n(L-tyrCo)
The synthesized [Co(L-tyr)]n(L-tyrCo) was characterized by powder X-ray diffraction (PXRD). As shown in Figure 2a, the experimental PXRD pattern of the synthesized [Co(L-tyr)]n(L-tyrCo) crystals was in good agreement with that simulated from single crystal data, indicating successful synthesis. After several extraction experiments and soaking in a large volume of methanol and n-hexane/isopropanol (70/30, v/v) for 48 h, the PXRD pattern was not significantly changed, suggesting good stability of the MOF, and that it would not be dissolved during the experiments

Characterization of the Synthesized [Co(L-tyr)] n (L-tyrCo)
The synthesized [Co(L-tyr)] n (L-tyrCo) was characterized by powder X-ray diffraction (PXRD). As shown in Figure 2a, the experimental PXRD pattern of the synthesized [Co(L-tyr)] n (L-tyrCo) crystals was in good agreement with that simulated from single crystal data, indicating successful synthesis. After several extraction experiments and soaking in a large volume of methanol and n-hexane/isopropanol (70/30, v/v) for 48 h, the PXRD pattern was not significantly changed, suggesting good stability of the MOF, and that it would not be dissolved during the experiments

Selection of UV Wavelength for Analysis of Eluate
To obtain the highest sensitivity, the optimum UV absorption wavelength of the analyte should be determined. The UV-visible absorption spectrum of (±)-1,1′-bi-2-naphthol in methanol was measured over 200-400 nm. As can be seen in Figure 3, the solution of (±)-1,1′-bi-2-naphthol showed strong absorptions at 278 and 335 nm. In principle, both of these wavelengths could be used to determine the absorbance of the eluate. However, some impurities in the eluent absorbed at around 271 nm, which would affect the accuracy of the experimental result. In contrast, impurities in the eluent had relatively low absorption at 335 nm, so this wavelength was used in subsequent experiments.

Optimization of SPE Conditions
The sorption of (±)-1,1′-bi-2-naphthol on a SPE cartridge packed with [Co(L-tyr)]n(L-tyrCo) can be influenced by many experimental conditions, such as solvent composition, sample solution volume, analyte concentration, and the type and volume of eluent. To obtain good extraction efficiency, these parameters were optimized.

Selection of UV Wavelength for Analysis of Eluate
To obtain the highest sensitivity, the optimum UV absorption wavelength of the analyte should be determined. The UV-visible absorption spectrum of (±)-1,1 -bi-2-naphthol in methanol was measured over 200-400 nm. As can be seen in Figure 3, the solution of (±)-1,1 -bi-2-naphthol showed strong absorptions at 278 and 335 nm. In principle, both of these wavelengths could be used to determine the absorbance of the eluate. However, some impurities in the eluent absorbed at around 271 nm, which would affect the accuracy of the experimental result. In contrast, impurities in the eluent had relatively low absorption at 335 nm, so this wavelength was used in subsequent experiments.

Selection of UV Wavelength for Analysis of Eluate
To obtain the highest sensitivity, the optimum UV absorption wavelength of the analyte should be determined. The UV-visible absorption spectrum of (±)-1,1′-bi-2-naphthol in methanol was measured over 200-400 nm. As can be seen in Figure 3, the solution of (±)-1,1′-bi-2-naphthol showed strong absorptions at 278 and 335 nm. In principle, both of these wavelengths could be used to determine the absorbance of the eluate. However, some impurities in the eluent absorbed at around 271 nm, which would affect the accuracy of the experimental result. In contrast, impurities in the eluent had relatively low absorption at 335 nm, so this wavelength was used in subsequent experiments.

Optimization of SPE Conditions
The sorption of (±)-1,1′-bi-2-naphthol on a SPE cartridge packed with [Co(L-tyr)]n(L-tyrCo) can be influenced by many experimental conditions, such as solvent composition, sample solution volume, analyte concentration, and the type and volume of eluent. To obtain good extraction efficiency, these parameters were optimized.

Optimization of SPE Conditions
The sorption of (±)-1,1 -bi-2-naphthol on a SPE cartridge packed with [Co(L-tyr)] n (L-tyrCo) can be influenced by many experimental conditions, such as solvent composition, sample solution volume, analyte concentration, and the type and volume of eluent. To obtain good extraction efficiency, these parameters were optimized.

Effect of Sample Solution Volume
The adsorption efficiency can also be influenced by the volume of the sample solution. Volumes of sample solution (3 mmol L −1 ) in the range of 1.0-3.0 mL were passed through the SPE cartridge, and the adsorbed analytes were eluted with methanol (4 mL). As shown in Figure 5, the absorbance of the eluate increased as the sample solution volume was increased from 1.0 mL to 2.0 mL, but remained stable as the volume was increased to 3.0 mL. Thus, the optimum volume was 2.0 mL.

Effect of Sample Solution Volume
The adsorption efficiency can also be influenced by the volume of the sample solution. Volumes of sample solution (3 mmol L −1 ) in the range of 1.0-3.0 mL were passed through the SPE cartridge, and the adsorbed analytes were eluted with methanol (4 mL). As shown in Figure 5, the absorbance of the eluate increased as the sample solution volume was increased from 1.0 mL to 2.0 mL, but remained stable as the volume was increased to 3.0 mL. Thus, the optimum volume was 2.0 mL.

Effect of Sample Solution Volume
The adsorption efficiency can also be influenced by the volume of the sample solution. Volumes of sample solution (3 mmol L −1 ) in the range of 1.0-3.0 mL were passed through the SPE cartridge, and the adsorbed analytes were eluted with methanol (4 mL). As shown in Figure 5, the absorbance of the eluate increased as the sample solution volume was increased from 1.0 mL to 2.0 mL, but remained stable as the volume was increased to 3.0 mL. Thus, the optimum volume was 2.0 mL.

Effect of Analyte Concentration
Five different concentrations (1.0, 2.0, 3.0, 4.0, and 5.0 mmol L −1 ) of (±)-1,1 -bi-2-naphthol were investigated. As shown in Figure 6, the absorbance of the eluate increased as the analyte concentration was increased from 1.0 to 3.0 mmol L −1 , with no further increase as the concentration was raised to 5.0 mmol L −1 . The optimal concentration was therefore 3.0 mmol L −1 .
Molecules 2018, 23, x FOR PEER REVIEW 5 of 10 concentration was increased from 1.0 to 3.0 mmol L −1 , with no further increase as the concentration was raised to 5.0 mmol L −1 . The optimal concentration was therefore 3.0 mmol L −1 .

Effect of Type and Volume of Elution Solvent
The nature of the elution solvent plays an important role in the effective desorption of analyte from adsorbent. The influence of elution solvent on desorption was evaluated under the optimum conditions using six common organic solvents: Methanol, Ethanol, Isopropanol, Acetonitrile, Ethyl acetate, and n-Hexane. As shown in Figure 7a, the maximum absorbance of the eluate was obtained when using methanol as the desorption solvent. Therefore, methanol was adopted as the optimal desorption solvent. The volume of methanol was also optimized because the sorption analyte cannot be completely eluted by a small amount of methanol, and long volatilization time will be taken when excessive methanol is used. As presented in Figure 7b, the absorbance of the eluate increased as the volume of methanol was increased from 1 to 4 mL, and no significant changes were observed as the volume was increased further. Consequently, 4 mL of methanol was sufficient for the analytes to be completely eluted and was chosen as the optimized volume.

Reproducibility and Reusability of the SPE Column
To evaluate the reproducibility of the SPE column, the sample solution (3 mmol L −1 , 2 mL) was added to each of 30 SPE cartridges filled with different batches of [Co(L-tyr)]n(L-tyrCo) particles. As shown in Figure 8a, the differences in the absorbance values of the eluates among the 30 columns were very small, indicating the good reproducibility of the SPE column. The reusability of the SPE column was also investigated. After each SPE experiment, the column was washed sequentially with methanol (10 mL) and n-hexane/isopropanol (70/30, v/v, 10 mL), and then reused for the next extraction cycle. As presented in Figure 8b, no obvious change in absorbance of the eluate was observed after five SPE cycles, indicating good reusability of the SPE column.

Effect of Type and Volume of Elution Solvent
The nature of the elution solvent plays an important role in the effective desorption of analyte from adsorbent. The influence of elution solvent on desorption was evaluated under the optimum conditions using six common organic solvents: Methanol, Ethanol, Isopropanol, Acetonitrile, Ethyl acetate, and n-Hexane. As shown in Figure 7a, the maximum absorbance of the eluate was obtained when using methanol as the desorption solvent. Therefore, methanol was adopted as the optimal desorption solvent. The volume of methanol was also optimized because the sorption analyte cannot be completely eluted by a small amount of methanol, and long volatilization time will be taken when excessive methanol is used. As presented in Figure 7b, the absorbance of the eluate increased as the volume of methanol was increased from 1 to 4 mL, and no significant changes were observed as the volume was increased further. Consequently, 4 mL of methanol was sufficient for the analytes to be completely eluted and was chosen as the optimized volume.
Molecules 2018, 23, x FOR PEER REVIEW 5 of 10 concentration was increased from 1.0 to 3.0 mmol L −1 , with no further increase as the concentration was raised to 5.0 mmol L −1 . The optimal concentration was therefore 3.0 mmol L −1 .

Effect of Type and Volume of Elution Solvent
The nature of the elution solvent plays an important role in the effective desorption of analyte from adsorbent. The influence of elution solvent on desorption was evaluated under the optimum conditions using six common organic solvents: Methanol, Ethanol, Isopropanol, Acetonitrile, Ethyl acetate, and n-Hexane. As shown in Figure 7a, the maximum absorbance of the eluate was obtained when using methanol as the desorption solvent. Therefore, methanol was adopted as the optimal desorption solvent. The volume of methanol was also optimized because the sorption analyte cannot be completely eluted by a small amount of methanol, and long volatilization time will be taken when excessive methanol is used. As presented in Figure 7b, the absorbance of the eluate increased as the volume of methanol was increased from 1 to 4 mL, and no significant changes were observed as the volume was increased further. Consequently, 4 mL of methanol was sufficient for the analytes to be completely eluted and was chosen as the optimized volume.

Reproducibility and Reusability of the SPE Column
To evaluate the reproducibility of the SPE column, the sample solution (3 mmol L −1 , 2 mL) was added to each of 30 SPE cartridges filled with different batches of [Co(L-tyr)]n(L-tyrCo) particles. As shown in Figure 8a, the differences in the absorbance values of the eluates among the 30 columns were very small, indicating the good reproducibility of the SPE column. The reusability of the SPE column was also investigated. After each SPE experiment, the column was washed sequentially with methanol (10 mL) and n-hexane/isopropanol (70/30, v/v, 10 mL), and then reused for the next extraction cycle. As presented in Figure 8b, no obvious change in absorbance of the eluate was observed after five SPE cycles, indicating good reusability of the SPE column.

Reproducibility and Reusability of the SPE Column
To evaluate the reproducibility of the SPE column, the sample solution (3 mmol L −1 , 2 mL) was added to each of 30 SPE cartridges filled with different batches of [Co(L-tyr)] n (L-tyrCo) particles. As shown in Figure 8a, the differences in the absorbance values of the eluates among the 30 columns were very small, indicating the good reproducibility of the SPE column. The reusability of the SPE column was also investigated. After each SPE experiment, the column was washed sequentially with methanol (10 mL) and n-hexane/isopropanol (70/30, v/v, 10 mL), and then reused for the next extraction cycle. As presented in Figure 8b, no obvious change in absorbance of the eluate was observed after five SPE cycles, indicating good reusability of the SPE column.
Given that an enantiomerically pure of L-tyrosine was used as ligand in the synthesis, the synthesized [Co(L-tyr)]n(L-tyrCo) is chiral. The 2D structure of [Co(L-tyr)]n(L-tyrCo) showing cobalt oxide layers pillared with L-tyrosine linkers and produced large chiral pores (Figure 1a). The influence of the chiral microenvironment on the enantioselective adsorption is complicated. It is difficult to completely understand the mechanism of interaction between the MOF and the two enantiomers. The reason for the selective adsorption of (S)-1,1′-bi-2-naphthol may be the high degree of space matching between (S)-1,1′-bi-2-naphthol and the chiral structures of the crystal resulting in different forces.
Given that an enantiomerically pure of L-tyrosine was used as ligand in the synthesis, the synthesized [Co(L-tyr)]n(L-tyrCo) is chiral. The 2D structure of [Co(L-tyr)]n(L-tyrCo) showing cobalt oxide layers pillared with L-tyrosine linkers and produced large chiral pores (Figure 1a). The influence of the chiral microenvironment on the enantioselective adsorption is complicated. It is difficult to completely understand the mechanism of interaction between the MOF and the two enantiomers. The reason for the selective adsorption of (S)-1,1′-bi-2-naphthol may be the high degree of space matching between (S)-1,1′-bi-2-naphthol and the chiral structures of the crystal resulting in different forces.  Given that an enantiomerically pure of L-tyrosine was used as ligand in the synthesis, the synthesized [Co(L-tyr)] n (L-tyrCo) is chiral. The 2D structure of [Co(L-tyr)] n (L-tyrCo) showing cobalt oxide layers pillared with L-tyrosine linkers and produced large chiral pores (Figure 1a). The influence of the chiral microenvironment on the enantioselective adsorption is complicated. It is difficult to completely understand the mechanism of interaction between the MOF and the two enantiomers. The reason for the selective adsorption of (S)-1,1 -bi-2-naphthol may be the high degree of space matching between (S)-1,1 -bi-2-naphthol and the chiral structures of the crystal resulting in different forces.
To verify the accuracy of the methodology, solutions of (±)-1,1 -bi-2-naphthol at a variety of ee values were determined by this method. The absorbances of the eluates were determined and the ee values were calculated using the standard curve (Table 1). Compared with the actual ee values measured by chiral HPLC, this method had an average error of 2.26%. This method can therefore be used for evaluation of the ee of (±)-1,1 -bi-2-naphthol with good accuracy.

Instrumentations
Powder X-ray diffraction (PXRD) patterns were recorded on a D/max-3B diffractometer (Tokyo, Japan) using Cu K α radiation. Scanning electron microscopy (SEM) images were recorded on a FEI Quanta FEG 650 scanning electron microscope (Hillsboro, OR, USA). The absorbance of eluate was determined on a TU-1901 UV/Vis spectrophotometer (Beijing Purkinje General Instrument Co., Ltd., Beijing, China). Commercial ChiralPak OD-H (250 mm × 4.6 mm, Daicel, Shanghai, China) and a Shimadzu HPLC system (Tokyo, Japan) consisting of an LC-15C HPLC pump and SPD-15C UV/Vis detector were used for the actual ee analyses.

Synthesis of [Co(L-tyr)] n (L-tyrCo)
[Co(L-tyr)] n (L-tyrCo) was synthesized according to the method of Rocha et al. [40]. Typically, Co(CH 3 COO) 2 ·4H 2 O and L-tyrosine in a molar ratio of 1:2 were mixed with ultrapure water (20 mL) in a Teflon-lined bomb and stirred with a magnetic bar. The pH of the mixture was adjusted to 9~10 by adding aqueous solution of NaOH (1 mol L −1 ). The bomb was sealed and heated at 130 • C for 3 days. After cooling to room temperature, the mixture was filtered and the purple crystals were washed thoroughly with distilled water and ethanol. Finally, the solid was dried at 100 • C.

Preparation of SPE Columns
The synthesized crystals of [Co(L-tyr)] n (L-tyrCo) were large and heterogeneous, which made them unsuitable for direct use as adsorbent in the preparation of SPE columns. The crystals were therefore milled and a suitable particle size was selected by preparing an ethanol suspension. Subsequently, each SPE cartridge was packed with the selected crystal particles (0.2000 g) using polypropylene frits at each end of the cartridge to keep the packing in place. The outlet tip of the cartridge was connected to a vacuum pump, and methanol was continuously added at the inlet end.

SPE Procedures
Before the extraction experiment, the crystals of the [Co(L-tyr)] n (L-tyrCo) packed cartridge were preconditioned by washing with methanol (5 mL) and activated with sample solvent (5 mL). A known volume of the sample was passed through the activated cartridge by gravity action. To remove residual sample on the surface of the crystals, additional sample solvent (5 mL) was added after the sample solution had completely passed through. Subsequently, the cartridge was eluted with methanol (4 mL). The eluate was volatilized, diluted with methanol to a final volume of 4 Ml, and the absorbance was determined by UV-visible spectrophotometry.

Conclusions
In this work, we reported a novel and simple method for the determination of ee value by SPE using a chiral MOF as sorbent. After adsorption and desorption, a standard curve was established that demonstrated a good linear relationship between ee and absorbance of the eluate. The ee value of (±)-1,1 -bi-2-naphthol can be rapidly calculated after SPE using the standard curve.

Conflicts of Interest:
The authors declare no conflict of interest.