A countercurrent chromatography solvent system based on deep eutectic solvents for separation of cis‐ and trans‐crocetin from Gardenia jasminoides Ellis

Abstract

Due to the structural similarity and large difference in concentration, the separation of trans‐ and cis‐crocetin has been challenging, and the cis‐crocetin is usually neglected. In this work, a countercurrent chromatography method was developed for the quick separation of trans‐crocetin and cis‐crocetin from the hydrolytic extract of Gardenia jasminoides Ellis. High purity of trans‐crocetin (>95%) and cis‐crocetin (>91%) were prepared simultaneously for the first time through a novel biphasic system based on deep eutectic solvents, n‐heptane/n‐butyl alcohol/13 mmol/L Na₂CO₃ in water/acetamide‐benzyltrimethylammonium chloride (4:1, mol/mol) (4:7:9:1, v/v). The addition of deep eutectic solvent significantly improved the separation efficiency. The two targets can be easily recovered from the separation system through simple acidification and precipitation. It has potential for preparative separations on a large scale.

Abbreviations

Ac

acetamide

CCC

countercurrent chromatography

DAD

diode array detector

DES

deep eutectic solvent

DSC

differential scanning calorimeter

FTIR

Fourier transform infrared spectroscopy

GA

glycolic acid

HBA

hydrogen bond acceptor

HBD

hydrogen bond donor

HSCCC

high‐speed countercurrent chromatography

LA

levulinic acid

TEBAC

benzyltriethylammonium chloride

TMBAC

benzyltrimethylammonium chloride

1. INTRODUCTION

Crocins, mainly including crocin‐1 and crocin‐2 (Figure 1), a group of water‐soluble carotenoid ester occurring in gardenia fruit (Gardenia jasminoides Ellis) and saffron (Crocus sativus L.), have long been used as a natural colorant and a traditional herbal remedy. The most costly spice in the world is saffron, which is known as “red plant gold” and costs over $1000 per kg. In contrast, gardenia fruit costs about $1 per kg and is widely grown throughout Asia [1, 2, 3]. Because of its high yield, low cost, and high crocin content, the fruit of G. jasminoides Ellis is commonly used as a raw material to extract crocin for industrial manufacturing. Crocetin is an aglycone of crocin naturally occurring in saffron and is produced in biological systems by hydrolysis of crocins as a bioactive metabolite [4]. Crocins and crocetin have major differences in terms of their nutritional, nutraceutical, and pharmacological benefits, which are attributed to their different chemical structures [5]. The development of amyloid fibrils, which are connected to Alzheimer's disease, may also be suppressed by crocetin, and it significantly enhances cell cycle arrest via p53‐independent mechanisms followed by P21 activation [6]. The content of crocetin in commonly available gardenia yellow products is much lower than that of crocins, hence, several studies have been conducted to purify crocetin after obtaining it by converting crocins [4, 7]. Crocetin has a polyunsaturated conjugated acid structure, four side‐chain methyl groups, and seven conjugated double bonds, comprising cis‐form and trans‐form, their structures are illustrated in Figure 1. In plants, cis‐crocetin is less common than its more stable trans‐isomer. This is most likely because cis‐crocetin is unstable due to the vast spatial locations between the huge groups around the double bond that offer a greater spatial resistance. However, it was found that after consumption of one cup of saffron tea, crocetin underwent isomerization in humans, with cis‐crocetin detectable in 75% of healthy individuals and the cis‐configuration accounting for 25−50% of total crocetin in blood samples [8]. It implies that the cis‐isomer may have higher pharmacological potency and bioavailability. Therefore, it is of great interest to develop a method to separate trans‐crocetin and cis‐crocetin, which can provide pure isomers and serve as a foundation for the study of the bioactivities.

FIGURE 1.

The structures of crocin‐1, crocin‐2, trans‐crocetin, and cis‐crocetin.

A number of methods have been used for preparative isolation of crocins, including molecularly imprinted polymer solid‐phase extraction [9], macroporous resin [10], aqueous two‐phase extraction [11], centrifugal partition chromatography (CPC) [12], expanded bed adsorption [1], and high‐speed countercurrent chromatography (HSCCC) [13]. Because the variation in structure between crocins is attributable to the distinct sugars to which the glycosidic linkages are connected, with a relatively significant difference in polarity, separation and preparation are easier than for crocetin, which may be separated using a simple technique. In addition, due to the complete lack or trace levels of crocetin in the products, there have been few reports of crocetin isolation from the complete lack or trace levels of crocetin in the products. Countercurrent chromatography (CCC) based on the partition of substances in two immiscible liquid phases, a novel liquid–liquid partition chromatography separation technology, has been used widely in the separation of active compounds from complicated mixtures. The advantages of CCC over traditional liquid–solid separation technology include short separation times, quick and efficient preparation, and quantitative material recovery. CCC also improves sample loading capacity for quick and large‐scale preparation by removing some drawbacks associated with irreversible sample adsorption loss onto the solid support [14, 15]. Furthermore, the ease of industrial production may be achieved with merely a low‐cost solid‐phase column, low‐cost solvents, and mild operating conditions [16]. The isolation of trans‐crocetin by CCC after alkali hydrolysis of crocins has been reported previously [17]. However, the excessive separation time (450 min) caused partial conversion of the trans‐crocetin, resulting in only 85% purity. Chyau et al. [18] also employed CPC for the high‐purity preparation of enzyme‐transformed trans‐crocetin. However, before the crude sample was loaded into the CPC system, it needed to be prepurified by the macroporous resin (HPD‐100). Furthermore, despite the preparation of a high‐purity trans‐crocetin, the cis‐isomer of crocetin was still missing. As a result, both previously described separation methods yielded just trans‐crocetin, and the separation procedure was not straightforward or quick. Thus, solvent systems with higher resolution are required to separate trans‐ and cis‐crocetin by CCC.

Deep eutectic solvents (DESs), which operate as environmental friendly substitutes for conventional organic solvents, have enormous potential for the production of many types of novel functional materials, facilitating easy access to cutting‐edge materials for sample preparation [19]. In several chromatographic techniques [20, 21, 22, 23, 24], DESs have been employed as suitable mobile phases or mobile phase additions. With the advantages of nontoxicity, biodegradability, low cost, solute stability, and simplicity of preparation [25, 26], DESs were first introduced in 2003 by Abbott and co‐workers [27], at a time when the overuse of traditional organic solvents constitutes a serious hazard to human health and the ecological environment. Synthetic organic solvents exhibit great extraction capacity depending on the polarity of themselves, however, DESs composed of hydrogen bond acceptor (HBA) and hydrogen bond donor (HBD) molecules in a certain molar ratio can obtain feasible structural designs by adjusting the combination of HBD and HBA [28]. Thus, DES can be designed and added to the system to alter target distribution and increase the separation effect. Several investigations have been conducted on DES‐based systems used in countercurrent chromatographic separation. Tong Wang et al. used DES (menthol‐lactic acid, 1:2, mol/mol)‐based solvents as HSCCC solvent systems to separate the hydrophobic compounds in Rosmarinus officinalis leaves, Ginkgo biloba leaves, and Salvia miltiorrhiza roots [29], and adopted the HSCCC solvent system DES (choline chloride/L‐malic acid, 1:1, mol/mol)‐water‐ethyl acetate (1:1:1, v/v/v) to separate nine target compounds from chrysanthemum extract [30]. In another study [31], five flavonoids were separated from Malus hupehensis using DES‐based HSCCC solvent system comprised of DES (choline chloride/glucose, 1:1, mol/mol)‐water‐ethyl acetate (1:1:2, v/v).

In this work, a novel biphasic system based on DES was first developed for the separation of cis‐ and trans‐crocetins. The composition of the basic solvent system, DES composition, including types of HBD and HBA and the mole ratios of them and sample loading capacity, were investigated and optimized. High purity of trans‐ and cis‐crocetins was prepared through a single separation procedure.

2. MATERIALS AND METHODS

2.1. Chemicals and reagents

Gardenia jasminoides Ellis was supplied by Beijing Tong Ren Tang Technology Development Ltd. The samples were pulverized by an automatic high‐speed blade mill, and passed through an 80‐mesh (0.2 mm aperture) sieve, and sealed and stored at room temperature. Acetamide (Ac, 99%), levulinic acid (LA, 98%), glycolic acid (GA, 99%), benzyltrimethylammonium chloride (TMBAC, 98%), benzyltriethylammonium chloride (TEBAC, 99%), crocetin (98%), and sodium carbonate (Na₂CO₃, 99%) were purchased from Shanghai Macklin Biochemical Technology Ltd. Ethanol, ethyl acetate, n‐butanol, heptane, and formic acid (analytical grade) were bought from Beijing Chemical Works. Acetonitrile and acetic acid (HPLC grade) were bought from Meridian Medical Technologies Ltd. (USA). Pure water was manufactured by a Milli‐Q water purification system (Millipore).

2.2. Apparatus

CCC separation was conducted on a type‐J CCC planet centrifuge with two spiral tube columns (STCCC‐30, self‐made in our laboratory). It was equipped with two 15‐mL column (0.8 mm id tubing). The CCC system was connected with LC3000 HPLC system (Beijing Tong Heng Innovation Technology Ltd.), which included two P‐3000 pumps, an UV‐3000 detector, and a CXTH3000 chromatography workstation. Analyses of samples containing targets were performed on an Agilent HPLC system (Agilent 1260 Series, America) equipped with a diode array detector (DAD), a column thermostat, a quaternary pump, and an autosample injection systems.

2.3. Sample preparation

According to the previous study [4], the gardenia fruit powder (0.5 g) was mixed with 50 mL of ethanol aqueous solution (ethanol: water: 1:1, v/v) and was sonicated for 1 h. Then, 100 mL of 0.1 M NaOH aqueous solution was added to the supernatant, and the mixture was sonicated for 30 min to convert crocins to crocetin, which was finally neutralized with 0.1 M HCl aqueous solution. The obtained precipitate was filtered, and washed with water to eliminate any remaining HCl in the aqueous solution. Thus, the crude sample of crocetin was obtained.

2.4. Synthesis and characterization of DESs

The strong interaction between benzylammonium chloride‐based‐DES and crocetin was found through our previous investigations (to be published). Thus, LA, Ac, and GA were chosen as HBDs, linked by hydrogen bonds with various benzylamine chlorides of different alkyl substituents as HBAs, and their structures are shown in Figure 2. All eutectic mixtures were heated at 80°C and stirred for 1 h until a stable, transparent liquid was formed. The thermal characteristics of DESs were tested using a differential scanning calorimeter (DSC, TA Instruments Q2000). The melting point for the DES was determined by DSC (accuracy ± 0.1°C; precision ± 0.02°C) in the temperature range from 25 to 300°C under nitrogen cleansing of 50 mL/min and a heating and cooling rate of 10°C/min. The samples were measured by Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR‐FTIR, IRTracer‐100, Shimadzu). The spectra were obtained for the transmission mode, with the wavenumber range from 4000 to 400 cm⁻¹ with a resolution of 2 cm⁻¹.

FIGURE 2.

The structures of five compounds used for the preparation of deep eutectic solvents.

2.5. Selection of the two‐phase solvent system

The biphasic solvent system was selected by measuring the partition coefficient (K) of trans‐ and cis‐crocetin. The K values were determined as follows: 1 mg of crude sample was dissolved in 1 mL of lower phase from the pre‐equilibrated biphasic solvent system. Then, 1 mL of the upper phase was added to the prepared solution and was thoroughly mixed. After the equilibration was established, the upper and lower phases were all determined by HPLC, and the peak areas were recorded as A _upper and A _lower respectively. The partition coefficient was calculated as K = A _upper/A _lower. The separation factor (α value) was calculated as the ratio of K values of targets.

2.6. Separation procedure

Solvent systems of n‐heptane, n‐butyl alcohol, and Na₂CO₃ in water or DES were used to separate trans‐ and cis‐crocetin from crude sample. First, the upper phase (stationary phase) was pumped into the spiral tube at a rate of 2 mL/min. When the upper phase had completely filled the column, lower phase (mobile phase) was pumped at a flow rate of 1 mL/min while the CCC apparatus was rotated at 1150 rpm. After achieving dynamic equilibrium, 1 mg of crude sample dissolved in 1 mL of the lower phase was injected into the column. The column effluent was continually monitored at 440 nm and each fraction was collected at same interval and then analyzed by HPLC.

2.7. Recovery of targets

After adding 10 μL of formic acid to the collected CCC fractions (1 mL), Na₂CO₃ in mobile phase was neutralized, and crocetin (cis/trans) insoluble in water was precipitated, and then pure crocetin (cis/trans) was obtained after washing with water. Then, the remaining collected fractions were subsequently analyzed by HPLC. The mass of isolated crocetin (cis/trans) is calculated as the amount of crocetin detected in the collected fractions minus the amount of crocetin detected in the collected fractions following acid addition. In addition, as the content of DES accounts for only 5% of the system, and the raw material for the synthesis of DES is relatively cheap, there is no need to use additional reagents to recover DES.

2.8. HPLC analysis and structural elucidation

Based on the previous research [4], the condition was fine‐tuned, and the final method is presented below. All samples were analyzed by HPLC with a C18 column (Yilite®, 200 ×4.6 mm, 5 μm) at 30°C. The mobile phase consisted of acetonitrile (A) and 1% (v/v) acetic acid solution (B). The isocratic elution was carried out by 70% A and 30% B for 6 min. The flow rate was 1 mL/min and the injection volume was 10 μL. The samples were filtered through a 0.22 μm PTFE syringe filter before injection. The chromatograms were recorded at 440 nm.

The purified trans‐ and cis‐crocetin was characterized by ¹H NMR. An Agilent DD2 operating at 600 MHz was used at 25°C with deuterated DMSO (DMSO‐d₆ ) as the solvent.

3. RESULTS AND DISCUSSION

3.1. Analysis of crude sample

Significant differences of total crocins, primarily including crocin‐1 and crocin‐2, were ascertained among cultivars. Crocin‐1 was determined to be the major crocin in gardenia fruit, ranging from 1.03 to 2.29% dry weight, and crocin‐2 follows it as the second abundant crocin, ranging from 0.09 to 0.31% dry weight [32]. The crude sample prepared above was analyzed by HPLC prior to CCC separation by external standard method. The contents of trans‐ and cis‐crocetin was 1.9 and 0.5%, respectively. It implied roughly that crocins were completely hydrolyzed into crocetin. It can be seen from Figure 3 that cis‐crocetin shows a significant absorption peak at 324 nm different from trans‐crocetin without wavelength absorption, which is an important indicator to distinguish trans‐ and cis‐form [33]. The total crocetin content in the crude sample to be separated was only 2.4%, while the amount of trans‐crocetin was over three times greater than the cis‐crocetin, which made the separation more challenging.

FIGURE 3.

HPLC chromatogram of crude sample at 440 nm and UV‐visible absorption spectra of trans‐ and cis‐crocetin.

3.2. Characterization of DESs

The synthesized DESs were identified here by DSC and FTIR characterization, as reported in Figure 4A. The melting point of the eutectic mixture is substantially lower than the melting points of the two components, indicating their specific strong interaction. As shown in Figure 4B, consistent with the description of the characteristic absorption of acetamide in previous study [34], the C–H stretching appears at 2815 cm⁻¹ and N–H stretching appears at 3146 and 3296 cm⁻¹. Moreover, C = O stretching vibrations is observed to be overlapped with the N–H bending vibration at 1661 cm⁻¹. When it comes to TMBAC, the peaks are assigned to C–H sp³ (3001 cm⁻¹), C−N (1211 cm⁻¹), C = C aromatic skeleton vibration (1474 cm⁻¹), C−H_arom wagging (776 and 737 cm⁻¹) of TMBAC. It is obvious that the peaks of the NH group in acetamide shift from 3296/3146 cm⁻¹ to 3325/3158 cm⁻¹. These demonstrate the successful formation of the proposed DESs [35].

FIGURE 4.

Characterization of deep eutectic solvent. (A) Experimental solid‐liquid equilibrium phase diagram; (B) Fourier transform infrared spectra (FTIR) of benzyltrimethylammonium chloride‐acetamide (TMBAC‐Ac: 1:4) and their pure sources.

3.3. Selection of a solvent system

According to the Pfizer company solvent selection guide [36], water, n‐butanol, and ethyl acetate were the preferred solvents, therefore, these three were picked as the basic components of the systems. Since, crocetin is insoluble in water and previous research has implied that sodium crocetinate has a more moderate difference in solubility between the cis‐ and trans‐forms than crocetin [37], Na₂CO₃ was added to the water. Furthermore, K values for various substances range from 0.5 to 2.5 and α ≥ 1.5 could offer an adequate resolution between targets [38]. As seen in Table 1, the solvent system ethyl acetate/1 mmol/L Na₂CO₃ in water (1:1, v/v) shows reasonable K values but insufficient separation of the cis‐ and trans‐crocetin, so this system was not appropriate for use. As for the solvent systems n‐butyl alcohol/250 mmol/L Na₂CO₃ in water (1:1, v/v), cis‐ and trans‐crocetin are only distributed in the upper phase. This is probably because n‐butyl alcohol is too polar to fairly partition the targets between the upper and lower phases. To cope with it, n‐heptane was added to the system, as a usable solvent specified in the Pfizer solvent selection guide, to modify the K value. Meanwhile, the addition of it shortened the settling time of the two phases (implicating a better retention rate of stationary phases). So the solvent system n‐heptane/n‐butyl alcohol/Na₂CO₃ in water were better suitable to be used as a basic system. Then, initially, DESs composed of Ac, GA, and LA as HBDs and TMBAC and TEBAC as HBAs, with an HBA:HBD mole ratio of 6:1 were tested to improve separation. It is obvious that using Ac as the HBD greatly enhanced separation, and TEBAC was less conducive to separation than TMBAC, which had a shorter chain length. Therefore, the biphasic system n‐heptane/n‐butyl alcohol/Na₂CO₃ in water/Ac‐TMBAC (4:6:9:1, v/v) was selected and further optimized in CCC.

TABLE 1.

Partition coefficient (K) values of targets and their resolution (α) by different solvent systems.

Solvent systems	Volume ratio (v/v)	K 1 (trans‐crocetin)	K 2 (cis‐crocetin)	α (K2/K1 )
Ethyl acetate/1 mmol/L Na2CO3 in water	1:1	1.04	1.03	1.01
n‐butyl alcohol/250 mmol/L Na2CO3 in water	1:1	N.A.	N.A.	N.A.
n‐Heptane/n‐butyl alcohol/10 mmol/L Na2CO3 in water	1.5:1:2.5	0.68	0.73	1.06
	2:1:3	0.14	0.15	1.1
	2.5:1:3.5	2.33	1.99	1.17
	3:1:4	1.24	1.16	1.07
n‐Heptane/n‐butyl alcohol/80 mmol/L Na2CO3 in water/LA‐TMBAC (6:1)	4:5:8:1	1.12	1.19	1.06
n‐Heptane/n‐butyl alcohol/80 mmol/L Na2CO3 in water/LA‐TEBAC (6:1)		0.89	0.86	1.04
n‐Heptane/n‐butyl alcohol/80 mmol/L Na2CO3 in water/GA‐TMBAC (6:1)	6:3:8:1	1.52	1.63	1.08
n‐Heptane/n‐butyl alcohol/80 mmol/L Na2CO3 in water/GA‐TEBAC (6:1)		1.02	1.04	1.02
n‐Heptane/n‐butyl alcohol/10 mmol/L Na2CO3 in water/Ac‐TMBAC (6:1)	3:6:8:1	1.59	2.11	1.33
n‐Heptane/n‐butyl alcohol/10 mmol/L Na2CO3 in water/Ac‐TEBAC (6:1)		2.3	2.87	1.25
n‐Heptane/n‐butyl alcohol/10 mmol/L Na2CO3 in water/Ac‐TMBAC (6:1)	2:6:8:1	3.74	5.01	1.34
	4:6:8:1	1.12	1.53	1.34
	4:6:9:1	1.37	2.02	1.47
	4:7:9:1	1.42	2.12	1.49
	4:6:10:1	1.85	2.58	1.39

The preliminary shake flask test revealed that a slight variation in the concentration of n‐butyl alcohol and DES content had a significant impact on the target partitioning in the system, thus, their concentrations were optimized by comparing the overlapping area ratio of cis‐ and trans‐crocetin, which was calculated as the overlay area of two peaks divided by the area of cis‐crocetin using Origin, indicating peak separation resolution. As shown in Figure 5A‐F, the overlapping area reduces and subsequently increases as the n‐butyl alcohol concentration increases, a concentration of 33.3% was chosen to optimize the DES content. A moderate amount of DES resulted in a considerable decrease in overlapping area. As shown in Figure 6A‐F, the overlapping area may reach 96.1% when the solvent system is not supplemented with DES. However, the continual addition of DES, on the other hand, increases the viscosity of the system, resulting in poor mass transfer f targets between two phases, lowering resolution efficiency. As a result, the system n‐heptane/n‐butyl alcohol/10 mmol/L Na₂CO₃ in water/Ac‐TMBAC (6:1): 4:7:9:1 was selected.

FIGURE 5.

CCC separation of trans‐ and cis‐crocetin from crude sample with different content of n‐butyl alcohol. System: n‐heptane/n‐butyl alcohol/10 mmol/L Na₂CO₃ in water/Ac‐TMBAC(6:1): 4:x:9:1. x: 5 (A); 6 (B); 7 (C); 8 (D); 8.5 (E); (F) The effect of n‐butyl alcohol content on overlay area ratio. Ac‐TMBAC (acetamide‐benzyltrimethylammonium chloride); column volume: 30 mL; rotational speed: 1150 rpm; sample: 1 mg; flow rate: 1.0 mL/min; detection: 440 nm; stationary phase (upper) retention: 60%.

FIGURE 6.

CCC separation of trans‐ and cis‐crocetin from crude sample with different content of DES. System: n‐heptane/n‐butyl alcohol/10 mmol/L Na₂CO₃ in water/Ac‐TMBAC(6:1) : 4:7:9:x. x: 0 (A); 0.5 (B); 0.8 (C); 1 (D); 1.5 (E); (F) The effect of DES content on overlay area ratio. Ac‐TMBAC (acetamide‐benzyltrimethylammonium chloride); other conditions are the same as in Figure 5.

The concentration of Na₂CO₃ had minimal influence on the separation, as indicated in Supporting Figure S1, the separation can be slightly improved by adjusting it. Furthermore, the structural tunability of DES is its most important benefit; DES‐based biphasic systems with optimum separation were found by varying molar ratios of HBD to HBA. The overlapping area is minimized at a molar ratio of HBD to HBA of 4:1, as seen in Supporting Information Figure S2. This may because that DES cannot be synthesized when the ratio of HBD is less than three times the HBA, when the ratio of HBD was increased from the point just able to form DES, the interaction between HBD and HBA was boosted, as well as the interaction force with the target; however, an excessive ratio of HBD would diminish the interaction between DES and the target, which would be detrimental to the separation. Thus, the system n‐heptane/n‐butyl alcohol/13 mmol/L Na₂CO₃ in water/Ac‐TMBAC (4:1): 4:7:9:1 was selected.

3.4. Sample loading capacity

The performance of the developed solvent system was then evaluated in terms of sample loading capacity and separation resolution. However, crocetin is insoluble in water and most organic solvents, except for pyridine and dimethylsulfoxide [39]. The maximal injection volume is 6 mg since the maximum solubility of the mobile phase for the crude sample of the targets is 6 mg/mL. There is no discernible change in the overlapping area ratio with the increase in sample loading quantity from 1 to 6 mg, as illustrated in Figure S3. This indicated that the separation technique was workable, and had a potential for preparative separations on a large scale.

3.5. Recovery and characterization of targets

After addition of formic acid, precipitation, washing and drying, about 99.3 μg trans‐crocetin and 21.89 μg cis‐crocetin were separated from 6 mg of the crude sample. They were subsequently analyzed by HPLC‐DAD, as shown in Supporting Information Figure S4, two highly purified targets were obtained with no obvious interference observed in 3D chromatograms. The purities of trans‐ and cis‐crocetin at 440 nm were as high as 95.9 and 91.7%, which made up 1.6 and 0.3% of the crude sample, respectively. The recovery rates are about 84 and 60% individually. The ¹H NMR identification spectral data of trans‐ and cis‐crocetin are described below. In comparison to trans‐crocetin, cis‐crocetin shows more numerous peaks due to hydrogen atom interaction around the isomeric sites, which are in agreement with the reported literature [40].

¹H NMR (600 MHz, DMSO‐d₆ ):

Compound 1 (trans‐crocetin): δ7.17 (d, J = 11.6 Hz, 2H), 6.80 (dd, J = 8.0, 2.8 Hz, 2H), 6.70 (d, J = 15.0 Hz, 2H), 6.58 (dd, J = 15.0, 11.6 Hz, 2H), 6.47 (d, J = 10.0 Hz, 2H), 1.94 (s, 6H), 1.88 (s, 6H).

Compound 2 (cis‐crocetin): δ7.34 (d, J = 14.7 Hz, 1H), 7.27 (d, J = 11.6 Hz, 1H), 7.23 (d, J = 11.9 Hz, 1H), 7.13 (dd, J = 14.1, 11.9 Hz, 1H), 6.77‐6.71 (m, 2H), 6.60 (dd, J = 14.7, 11.6 Hz, 2H), 6.42 (d, J = 11.6 Hz, 1H), 6.32 (d, J = 11.9 Hz, 1H), 1.97‐1.91 (m, 6H), 1.88 (d, J = 4.2 Hz, 6H).

4. CONCLUSIONS

In this study, a novel biphasic system based on DES was first developed for the quick separation of trans‐ and cis‐crocetin. The composition of the basic solvent system, DES composition, including types of HBD and HBA and the mole ratios of them, and sample loading capacity were investigated. After optimization, high‐purity trans‐crocetin (95.9%) and cis‐crocetin (91.7%), which account for around 1.9 and 0.5% in the crude sample, were effectively separated using the solvent system n‐heptane/n‐butyl alcohol/13 mmol/L Na₂CO₃ in water/Ac‐TMBAC (4:1) (4:7:9:1, v/v). After recovery, about 99.3 μg trans‐crocetin and 21.89 μg cis‐crocetin were prepared from 6 mg crude samples. It has potential for preparative separations in a large scale. The separation efficiency of the DES‐containing system was noticeably better than the conventional system, indicating that DES may successfully enhance the difference in partitioning between cis‐ and trans‐isomers. It expands the range of applications for CCC in separation and opens up a fresh path for the logical design of two‐phase solvent systems integrating DES for efficient separation of cis‐ and trans‐isomers.

Supporting information

Supporting Information

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.