Journal of Chromatography A, 1368 (2014) 1–17

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

Review

Applications of liquid-phase microextraction techniques in natural product analysis: A review Yunyan Yan, Xuan Chen, Shuang Hu, Xiaohong Bai ∗ School of Pharmacy, Shanxi Medical University, Taiyuan 030001, PR China

a r t i c l e

i n f o

Article history: Received 4 May 2014 Received in revised form 24 September 2014 Accepted 25 September 2014 Available online 2 October 2014 Keywords: Liquid-phase microextraction Natural product analysis Herbal medicine Traditional Chinese medicine In vivo In vitro

a b s t r a c t Over the last years, liquid-phase microextraction (LPME) as a simple, rapid, practical and effective samplepreparation technique, coupled with various instrumental analytical methods, has been increasingly and widely used to research and determine trace or ultra-micro-levels of both inorganic and organic analytes from different matrix-complex samples. In this review, different kinds of LPMEs such as single drop liquid-phase microextraction, dispersive liquid–liquid microextraction, and hollow fibre liquid-phase microextraction are summarized and recent applications of LPMEs in trace compounds in vivo and in vitro from different natural product matrice analysis such as tea, vegetables, seeds, herbs, and galenical are also discussed. Finally, future developments and applications of LPMEs in complex sample analysis are prospected. © 2014 Elsevier B.V. All rights reserved.

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.1. Single drop liquid-phase microextraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.2. Dispersive liquid–liquid microextraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.3. Hollow fibre liquid-phase microextraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Concluding remarks and future prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

1. Introduction Natural product analysis constitutes one of the most important fields of the science at present. The analysis of organic and inorganic compounds in natural products has gained considerable interest in recent years [1–3]. This interest is particularly focused on herbal medicines (HMs), including herbs, herbal preparations, herbal materials and finished herbal products, which contain active ingredients of natural products, other natural product materials or combinations [4]. HMs, such as traditional Chinese medicines (TCMs), have performed crucial functions in clinical therapy for

∗ Corresponding author. Tel.: +86 13935105965; fax: +86 351 4690114. E-mail address: [email protected] (X. Bai). http://dx.doi.org/10.1016/j.chroma.2014.09.068 0021-9673/© 2014 Elsevier B.V. All rights reserved.

numerous diseases and have provided valuable and easily obtainable healthcare resource in many oriental countries for thousands of years [5]. During the last few decades, the use of TCMs has expanded globally, both as primary health care of the poor in developing countries and as the predominant medical treatment in the national health care system [6]. With the widespread use of traditional medicine, safety and effectivity, as well as quality control of HMs and traditional preparation procedure, have become important concerns for both health authorities and the public. However, little is known about the chemical compositions, pharmacokinetics, pharmacodynamics and metabolomics of TCMs to date. In addition, data about identification, efficacy and safety of TCMs are far from sufficiently meeting the criteria to support their use worldwide. This state is mainly due to the following considerations: (1) TCMs are complex matrix, multi-component

2

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

integration and collaborative systems; (2) TCMs contain low active ingredient content; and (3) pretreatment technology or method is not adequate or acceptable for TCMs research and evaluation. Target compound analysis in complex samples generally includes two steps of sample pretreatment (extraction, separation, purification and concentration) and target compound determination (identification, qualitative and quantitative analysis). Given the evolution of computer, information, and instrument technologies, analysis methods and technologies have been greatly promoted and developed. However, sample pretreatment technologies or methodology have yet to adapt instrument development and meet target analyte requirements in identification, qualitative and quantitative analysis. Therefore, increased efforts are currently being focused on sample pretreatment improvement. In this sense, current trends are moving towards (1) reducing the number of steps required for the procedure, (2) reducing or totally eliminating solvents required for extraction, (3) extending the adaptability to field sampling, and (4) automation [7]. Such enhancement poses a challenging task, especially in the field of natural product analysis, in which the sample is a high complex, multi-component and has low levels of active ingredients systems. Therefore, it is essential to establish a sample pretreatment technology that can retain the effective ingredient at the extreme, as well as to separate and remove ineffective ingredient and impurities. Sample preparation is the basic and crucial step in the success for any analytical method. Sample preparation is the basic and most crucial step in the success of any analytical method. A successful sample pretreatment method typically has three major objectives: (1) sample matrix simplification and/or replacement, (2) analyte enhancement or concentration, and (3) sample clean-up [7]. Over the last years, several new miniaturized solvent-based extraction procedures, which are known as liquid-phase microextraction (LPME) techniques, have been introduced and applied with success [8–14]. LPME emerged from liquid–liquid extraction, which is probably the most widely used sample extraction and separation procedure despite its clear disadvantages such as high consumption of time and strong toxicity of solvent, as well as its tedious application [15]. LPME normally takes place between several microliters of water-immiscible solvent extraction phase or acceptor phase (AP) and an aqueous sample phase or donor phase (DP), which contains the target analytes of interest. LPME can be classified into three main categories [7,16]: single drop liquid-phase microextraction (SD–LPME) [5,17,18], dispersive liquid–liquid microextraction (DLLME) [19–22] and hollow fibre liquid–phase microextraction (HF–LPME) [23–26]. Fig. 1 shows a schematic diagram of several LPME modes. Several variations have also recently been introduced in each mode, which clearly demonstrate the methodology’s versatility. The LPME technique combines sampling, extraction, separation and concentration in one step, meanwhile a relatively high enrichment factor (EF) of analyte is obtained since its low V AP volume (Va ) and high DP volume (Vd ) (that is ˇ = Vda , phase ratio to be the higher), and the AP is easily introduced into a chromatographic or electrophoretic system. LPME performs a primary function in the extraction of inorganic and organic compounds from water samples, such as inorganic substance (fluorine [27], vanadium [28], arsenic, stibium, bismuth, plumbum, stannum and mercury [29], iodate [30], and arsenic (III) and arsenic (V) [31]); as well as organic substance (polycyclic aromatic hydrocarbon (PAH) [32,33], organochlorine pesticides [34–36], chloroacetanilide herbicide [37], sulphur compounds [38]) and so on. However, the matrix complexity, as well as the varied, low level contents of components in natural samples is a considerable drawback that makes the application in this area difficult. Even so, multiple developments and applications highly focused on the TCMs analysis field have been proposed, which represent

the start of the expansion of LPME in TCMs or complex samples analysis. Therefore, this review will primarily focus on the different applications of LPME techniques, such as SD–LPME, D–LLME, HF–LPME and so on, coupled with various instrumental analytical methods, such as high performance liquid chromatography (HPLC), gas chromatography (GC), liquid chromatography/mass spectrometry (LC/MS), ultra-high performance liquid chromatography/mass spectrometry (UHPLC/MS), gas chromatography/mass spectrometry (GC/MS), capillary electrophoresis (CE), ultraviolet–visible spectrophotometry (UV) and atomic absorption spectrophotometry (AAS), in the field of natural product analyses during the past few years. The advances of LPME technique, in vivo and in vitro, for the analysis of organic and inorganic compounds from different natural product types, such as tea, vegetables, seeds, herbs and galenical, which will be summarized and discussed below, clearly demonstrate the potential of the LPME technique as a powerful sample preparation tool in complex sample analysis. To the best of our knowledge, the present paper is the first review article dealing with the specific application of LPME techniques in natural product analysis.

2. Applications 2.1. Single drop liquid-phase microextraction One of the modes of the LPME that was first reported by Liu and Dasgupta [39] is termed SD-LPME, in which the extraction medium is in the form of a single drop (Fig. 1A and B). The technique is based on the distribution of analytes between a microdrop of extraction solvent (usually few microliters) at the tip of a microsyringe needle and an aqueous sample phase containing the analytes. After extraction, the microdrop is retracted back into the microsyringe and injected into a chromatographic or electrophoretic system for further analysis. Jeannot and Cantwell [40,41], as well as He and Lee [11], performed the technique in combination with chromatographic analysis based on a previous work of Liu and Dasgupta [39,42], in which a single drop was used as the analyte collector. SDLPME can be classified into two-phase [11,40,43] and three-phase mode [7,44]. In the two-phase mode, such as direct immersion SD-LPME (DI-SD-LPME) (Fig. 1A) and continuous flow SD-LPME (CF-SD-LPME), analytes are extracted from the sample solution (DP) into the organic solvent (AP) as a microdrop suspended from a microsyringe needle. Considering that suspended particles or impurities in sample solution may disturb the microdrop or even make that drop highly unstable and easily fall off, the two-phase mode is more suitable for simple matrix sample. In the three-phase mode, such as headspace SD-LPME (HS-SD-LPME) [45] (Fig. 1B) and drop-to-drop SD-LPME (DD-SD-LPME) [46], analytes are firstly extracted from the DP into the organic solvent or the headspace, and then back-extracted into the single drop aqueous AP. Compared with the two-phase mode, the three-phase mode is more suitable for the analysis of volatile components in complex samples. In the process, the microdrop remains relatively stable and nearly not influenced by the stirring, impurities or sample matrix interference. To reduce evaporation risk during the extraction period and obtain the desired results, several important factors of extraction solvent, such as relatively high boiling point or relatively low vapour pressure, density, high viscosity, suitable chromatographic behaviour and high extraction efficiency (EE) or EF for the target analytes, must be considered [47]. Based on these factors, the most common extraction solvents used are toluene [48], hexane, octane [49], dodecane [50] and xylene [5,45]. Except that, certain ionic liquids (ILs) [51], such as 1-butyl-3-methylimidazollium hexafluorophosphate ([BMIM][PF6 ]) [14], 1-hexyl-3-methylimidazolium

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

3

Fig. 1. A schematic diagram of some of these microextraction modies.

: magnetic stirrers;

: stirrer;

: sample pool;

ble acceptor phase (inside) and extraction solvent (outside); : microinjector;

: centrifuge tube; and

: extracting solvent;

: sample solution;

: hollow fibre tube with extraction solvent;

: hollow fibre tube with water solu-

: hollow fibre film with extraction solvent;

: spoon.

hexafluorophosphate ([HMIM][PF6 ]) [36] and 1-octyl-3methylimidazolium hexafluorophosphate ([OMIM][PF6 ]) [52], can also provide satisfactory results with better reproducibility [53] This performance is due to the ionic liquids’ high viscosity and surface tension, which helps to form a stable drop of a considerably larger volume [54] to prolong the extraction time [55–57]. In addition, ␤-cyclodextrine [58], surfactants [59] and supramolecular solvents [60] as extractants were also proposed in the field of SD-LPME. However, in actual work, the extraction solvent should be selected according to particular samples and target analytes. For example, in the case of DI-SD-LPME mode, if the extraction solvent with a significantly different density from the analysed sample is selected, the extraction drop will most probably be broken away from the needle tip. In certain cases, high boiling point is not such significant factor for the extraction solvent. On the contrary, low boiling point could be advantageous because of the possible coelution of this solvent and the target analytes in the sample. In addition, the most effective syringe for SD-LPME is a standard GC microsyringe (e.g., skew needle). The drop must cling to the tip without wicking up the exterior of the syringe needle, which requires the maximum needle tip surface area. The standard Hamilton #2 curved bevel syringe tip provides the greatest surface area, and approximately 90%–95% of the drop can be withdrawn into the syringe after extraction [54]. The mass transfer of analytes in sample phase is the rate-determining step in this extraction, so a relatively high stirring speed, high extraction temperature and long extraction time can increase and improve the EF or EE of target analyte. However, the drop volume varies during the extraction process, particularly under extreme extraction conditions (high stirring speed, high temperature and long extraction time), which could affect drop stability [5] and analytical precision [61]. This problem can be avoided or overcome when an appropriate, quantitative internal standard is introduced into the drop during extraction [18,39,62].

A detailed description of these SD-LPME modes, as well as their theoretical aspects, can be found in specific review articles [5,17,18,56,63–65] and monographs [7,44]. In short, SD-LPME has become one of the most popular sample preparation techniques because of its (1) simplicity and speed, (2) low cost and low environmental pollution, (3) wide range of solvent types being selected and used, (4) applicability to different complex matrices, (5) extraction and separation of both organic and inorganic compounds, and (6) compatibility with chromatographic or electrophoretic injection systems. Despite SD-LPME attracting considerable attention in recent years for the extraction of both organic and inorganic species, this approach is best suited for the separation and enrichment of nonpolar or moderately polar analytes from relatively clean matrix environmental samples (e.g., tap water or ground water) [66–69]. In contrast, natural product sample (particularly TCM) and biological sample applications have received limited attention, which is probably due to the complex and dirty nature of the sample matrixes. The first application of DI-SD-LPME for natural product analysis was reported in 2004 by Paramita Das et al. [70], who employed DI-SD-LPME combined with GC/MS for the determination of free iodide in aubergine, lotus seeds, lotus stem and potato samples. In the work, with N,N-dimethylaniline as an iodine derivatization reagent at pH 6.4, a straightforward conversion of iodide into 4-iodo-N,N-dimethylaniline, which was rapid and unaffected by the presence of many ionic substances, occurred. When 4-iodoN,N-dimethylaniline was formed, DI-SD-LPME was employed to extract the 4-iodo-derivative, and then GC/MS was used for determination of iodide content. 4-Iodo-N,N-dimethylaniline was a good candidate for extraction by DI-SD-LPME, followed by sensitive determination by GC/MS. Dried vegetables were wet-combusted with peroxydisulphate to liberate covalently bound iodine as iodate, which was reduced before derivatization. A rectilinear calibration graph was obtained between 0.1 ␮g L−1 to 10 mg L−1 for iodide using the DI-SD-LPME method. The correlation coefficient

4

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

and the limit of detection (LOD) for iodide were 0.9998 and 10 ng L−1 , respectively. The article points out that, compared with solid phase microextraction (SPME), DI-SD-LPME appeared to be a more efficient technique in terms of better LOD and precision. Moreover, the technique is simple to perform, naturally free from memory effects and cost-effective. To date, DI-SD-LPME has also been successfully applied to analyse the active alkaloids ingredients of China in human urine [71] and adenine in green tea sample [72]. HS-SDME is more often applied to analytes with relatively high vapor pressure. The first application of HS-SDME for natural product analysis was reported in 2005 by Deng et al. [73]. In the work, 27 compounds of the essential oil (for example camphor, borneol, borneol acetate and so on) in the TCM, Fructus amomi, were extracted using pressurized hot water extraction, followed by extraction and concentration with HS-LPME and detection by GC/MS. The concrete steps involved the following: (1) withdraw 1.0 ␮L of cyclohexane (containing 20 ␮g L−1 menthol, internal standard) into the microsyringe, (2) insert the microsyringe needle into headspace vial and keep the needle suspended over aqueous extract, and (3) withdraw 5.0 ␮L of gaseous sample at 1.0 ␮L s−1 , then depress the plunger back the original mark immediately and hold for 5.0 s. The same process was repeated for 20 times. Finally, the microsyringe needle was removed from the headspace vial and injected into GC/MS for analysis. The HS-LPME parameters (extraction solvent, plunger withdrawal rate, dwelling time and extraction cycle) with respect to the EE were optimized and investigated, and method repeatability was studied. Twenty-seven compounds in the essential oil were identified, and only three active compounds camphor, borneol and borneol acetate in the F. amomi sample were quantitatively analyzed by internal standard method. The proposed method required simple sample preparation and small sample mass, and provided good repeatability and short analysis time. Since then, HS-SDME has also been reported for the analysis of the essential oil in Zanthoxylum bungeanum Maxim [74], dried S. aromaticum (L.) Merr. et Perry, Cuminum cyminum L [75] and C. cyminum L [76]. The technique was also used to analyse the volatile components in Artemisia capillaris Thunb [45], Myrtus communis L [77], Foeniculum vulgare Mill [78] and Curcuma wenyujin Y. H. Chen et C. Ling [79], thymol and carvacrol from Thymus transcaspicus [80]. This last case, which corresponded to the HSSDME work of Yang et al. [81], in which hydrodistillation coupled with HS-SDME with IL as the extraction solvent, followed by GC analysis technique, was successfully developed to determine the volatile and semi-volatile compounds in the seeds of C. cyminum L. Table 1 summarizes these applications of SD-LPME in natural product analysis, including sample, target analyte, optimum extraction condition, EFs, detection methods and quantitative analysis parameters. Among these applications, given the possibility of in situ complexation or derivatisation, less sample pollution and strong purification ability, HS-SDME was the technique that has been most used in this sense. 2.2. Dispersive liquid–liquid microextraction Dispersive liquid–liquid microextraction [82–86] (DLLME, Fig. 1C) as a novel sample preparation method was first introduced by Assadi and co-workers in 2006 [87]. Conventional DLLME is based on a ternary component solvent system, including disperser solvent, extraction solvent (AP) and aqueous phase sample (DP) containing the analyte of interest. When an appropriate mixture of the extraction solvent and dispersive solvent is injected into an aqueous phase sample, a cloudy solution is formed. The hydrophobic target analytes are enriched in the AP, which is dispersed into the bulk aqueous solution. After extraction, the AP gathers and precipitates at the bottom of DP or floats on its surface, the analytes enhanced in the AP are analysed by conventional techniques.

In traditional D-LPME, extraction solvent, dispersing solvent and assistant-dispersive measure are the major effect factors. In generally, the extraction solvent typically uses microliter volumes of solvent including low-density [88–90] or high-density [28,90–104] solvents more than water. Common high-density extraction solvents include chlorobenzene, dichloromethane, carbon tetrachloride, tetrachloroethylene, ILs and so on. Several low-density extraction solvents, for instance, long-chain alcohols like 1-undecanol, 1-dodecanol, 1-octanol, n-hexanol, etc., are used. Common disperser solvents are both water dissolved and extraction solvent dissolved, small molecule organic solvent, such as acetone, methanol, acetonitrile, ethanol, tetrahydrofuran, etc. In addition, the researcher [105,106], used magnetic nanoparticles (MNPs) as the disperser in the DLLME process and developed new DLLME technology-based MNPs. These shaking[55,107,108], stirring-, temperature- [109–111], vortex- [112–114] and ultrasound-assisted [115–117] steps are used as assisting measure for hastening AP dispersion in the DP and enlarging the contact between the AP and DP to rapidly attain extraction equilibrium. Surfactant solutions [118–120] and ILs, which are also regarded as environment-friendly solvents that are easily commercially available, have been proposed as alternatives to organic solvents because of their low vapour pressure, high viscosity, good thermal stability, miscibility with water or organic solvent and greater use of larger, reproducible extracting volume [64]. These solutions are easily synthesised or commercially available. Tian et al. [55] developed and introduced organic solvent DLLME (OS-DLLME) and ionic liquid DLLME (IL-DLLME) coupled with HPLC for analysing an active ingredient, emodin, and its metabolites (aloe-emodin, anthraquinone-2-carboxylic acid, rhein, danthron, chrysophanol and physcion) in urine samples. Under optimal conditions, the EFs for emodin and its metabolites by OS-DLLME and IL-DLLME were within the range of 90 to 295 and 63 to 192 respectively. The relative standard deviations (RSDs, n = 3) for intra-day and interday precisions were lower than 7.2% and 8.7% by OS-DLLME and lower than 5.7% and 6.4% by IL-DLLME. Finally, recoveries ranged from 87.1% to 108% for OS-DLLME and from 87.5% to 106% for ILDLLME. The group verified that no significant deviations existed between the two methods for the determination of emodin and its metabolites. From the results of HPLC/UV of the urine sample after DLLME, the metabolites of aloe-emodin, rhein, chrysophanol and physcion were identified by comparing the retention times with the standards. From the results of HPLC/MS of urine sample after D-LPME, anthraquinone-2-carboxylic acid and danthron as unreported metabolites of emodin were first found. Anderson and coworkers [121,122] introduced an in situ metathesis IL-DLLME method. In this approach, a hydrophilic IL is completely dissolved in the aqueous sample solution to promote interaction between the IL and analytes. Then, an ion-exchange reagent is introduced to carry out the in situ metathesis reaction. A turbid solution with fine IL micro-droplets was formed, and a hydrophobic IL was produced, which can be directly analysed by HPLC. A significant advantage of this method lies in that the metathesis reaction and extraction are accomplished in one step, making the method very rapid and amendable to high-throughput analysis. Elimination of the dispersion solvent is an approach that has been developed in DLLME and applied in TCMs analysis, because the solvent can decrease the partition coefficient of the analytes into the extraction solvent. Ultrasound is frequently used to assist the extraction to form an appropriate oil-in-water emulsion that would disperse the extraction solvent into the aqueous phase. The approach called ultrasound assisted DLLME (USA-DLLME) increases the interfacial area between DP and AP without the aid of a dispersive solvent. Yan et al. [123] developed an USADLLME coupled with HPLC and introduced it for the extraction

Table 1 Applications of SD-LPME in natural product analysis. Analyte

Mode

Extraction solvent

Donor phase

Extraction time/ temperature

Analytical Limit of detection method

Recoveries (%)

Enrichment factor

Ref.

Artemisia capillaris Thunb

␣-Pinene ␤-Pinene ␤-Caryophyllene (E)-␤-Farnesene Capillene Cubenol ␤-Eudesmol Total iodine

HS-LPME

Dodecane

The sample (2.0 g)

4 min

GC/MS

–

–

–

[45]

SD-LPME

iso-Octane

Aqueous sample solution

15 min

GC/MS

10 ng L−1

96.5–107.0

–

[70]

Palmatine Berberine Tetrahydropalmatine

SD-LPME

n-Octanol

5 min /30 ◦ C

MEKC␣

0.5 ng mL−1 0.2 ng mL−1 1.5 ng mL−1

88.5–95.0 105.5–107.7 93.1–115.5

3556 2114 1583

[71]

Green tea

Adenine

Ethyl acetate

10 min

CE

0.002 ␮g mL−1 99.4

550

[72]

Fructus amomi

Camphor Borneol Borneol acetate 4-Methyl-1-(1-methylethyl)bicyclo[3.1.0]hexene (E)-3,7-Dimethyl-1,3,6-octatriene linalool 3,7-Dimethyl-2,6-octadien-1-ol 3-Methyl-6-(1-methylethyl)-2cyclohexen-1-one Terpinyl acetate Essential oil

Single-drop liquid–liquid–liquid microextraction HS-LPME

4.0 mL sample solution containing 500 mM NaOH Aqueous sample solution

Cyclohexane

Aqueous extract

10 min/60 ◦ C

GC/MS

–

–

–

[73]

HS-LPME

n-Heptadecane

100 mg of sample powders and 5 mL of water

10 min

GC/MS

–

–

–

[74]

HS-LPME

Decane

0.4 g of sample

5–7 min/100 ◦ C

GC/MS

–

–

–

[75]

Pharmaceuticals iodized salt milk powder vegetables Human urine

Zanthoxylum bungeanum Maxim.

Syzygium aromaticum (L.) Merr. et Perry Cuminum cyminum L.

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

Sample

5

6

Table 1 (Continued ) Analyte

Mode

Extraction solvent

Donor phase

Extraction time/ temperature

Analytical Limit of method detection

Cuminum cyminum L.

␤-Pinene ␥-Terpinene p-Cymene Cuminal Cuminalcohol ␣-Pinene Limonene 1,8-Cineole Linalool Linalyl acetate ␤-Myrcene Limonene ␥-Terpinene Fenchone Estragole trans-Anethole 4-Methoxy-benzaldehyde ␤-Elemene, Curzerene, Curzerenone, Germacrone, Curcumol, Isocurcumenol, Curcumenol Thymol Carvacrol ␤-Pinene p-Cymene ␥-Terpinene Cuminal Cuminalcohol

HS-LPME

n-Heptadecane

50 mg of sample powder and 3 mL of water

5 min

GC/MS

HS-LPME

n-Octadecane

The powdered sample (0.6 g, 30 mesh)

30 min /40 ◦ C.

HS-LPME

Benzylalcohol

The powdered sample (1.0 g, 120 mesh)

HS-LPME

n-Dodecane

HS-LPME HS-LPME

Myrtus Communis L

Foeniculum vulgare Mill

Curcuma wenyujin Y.H. Chen et C. Ling

Thymus transcaspicus Cuminum cyminum L.

Micellar electrokinetic chromatography.

Recoveries (%)

Enrichment factor

Ref.

–

–

[76]

GC/MS

14.8 pL L−1 6.67 pL L−1 10.1 pL L−1 – – –

–

–

[73]

20 min /70 ◦ C.

GC/MS

–

–

–

[78]

The powdered sample (4.5 g, 120 mesh)

20 min /70 ◦ C.

GC/MS

–

–

–

[79]

n-Pentadecane

1 g sample in water

2 min

GC/FID

[80]

The sample (2.0 g) and water (30 mL)

30 min

GC/MS

89–101 95–116 –

–

[OMIM][PF6 ]

1.87 mg L−1 0.23 mg L−1 –

–

[81]

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

Sample

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

7

Fig. 2. Extraction mechanism of phenylpropionic acids by IL-DLLME. A, the [HMIM][PF6 ] molecular-ordered organized assembly; B, the carboxyl-imidazole charge transfer superamolecule. : [HMIM][PF6 ];

: Phenylpropionic acid.

and analysis of anthraquinone compounds like aloe-emodin, rhein, emodin, chrysophanol and physcion in TCMs samples. The most favourable conditions, such as an AP of 60 mg of [HMIM][PF6 ], DP of pH 3.0, ultrasound at 40 ◦ C for 7 min, and centrifugation at 2500 rpm for 6 min, were obtained. Under the optimal conditions, good linearities of the five anthraquinone compounds were obtained with correlation coefficients higher than 0.9939; the LODs ranged between 0.01 and 0.09 ␮g L−1 ; the RSDs (n = 3) were less than 9.8% and the EFs were in the range of 80–197 fold. In addition, temperature-assisted DLLME [109] has also been proposed in 2010 for the determination of anthraquinones in Radix et Rhizoma Rhei samples. Afterwards, Tian [107] and Wang [124] respectively analysed and discussed the extraction mechanism of IL–DLLME. Wang et al. thought that in the IL–DLLME, [HMIM][PF6 ] as AP was first diffused and orderly arranged in the DP, and then formed the IL molecular ordered organized assembly (MOOA), which consisted of hydrophilic cations (imidazole group) in the DP and a hydrophobic alkyl chain constituting the core (Fig. 2A). In the acid medium, conjugated electrons in phenylpropionic acids move to the carboxyl because of a carboxyl electron-withdrawing inductive effect, thus making the carboxyl electron-rich. When extracted by IL–DLLME, the phenylpropionic acids encounter imidazole cation in IL’s MOOA, and the electron-rich carboxyl easily transfers electrons to the imidazole cation and forms a more steadily and highly ordered carboxyl-imidazole charge-transfer supramolecular (CICTSM). Thus, the phenylpropionic acids were included in IL’s MOOA and extracted by IL (Fig. 2B). To verify the CICTSM extraction mechanism, the author compared and analysed the ultraviolet spectrum of phenylpropionic acids before and after IL-DLLME. The results showed that the absorbances of the target analytes after IL-DLLME are significantly higher than before IL-DLLME, and the maximum absorption wavelength underwent a blue shift by 10 nm. ˇ and Bursová [125] developed a new In addition, Cabala microextraction technique for equilibrium, non-exhaustive analyte preconcentration. The key point of the method is the application of a specially designed, optimised bell-shaped extraction device. The technique has been applied to the preconcentration of selected volatile and semi-volatile compounds from aqueous solutions into organic solvents lighter than water and determined by GC/MS in spiked water samples. A novel automated approach, low density solvent based/solvent demulsification DLLME coupled to analysis by GC/MS, was

developed [126]. In this procedure, all extraction steps and analysis were continuously carried out repetitively, completely and automatically with the use of a CTC CombiPAL autosampler. Another kind of DLLME named solidification of floating organic drop microextraction (SFOME) (Fig. 1D) was introduced by KhaliliZanjani [127] in 2007. To date, the method has been further expanded [128–132]. A solvent droplet with a melting point (M.P.) of 10 ◦ C–30 ◦ C and a density lower than that of water is added (DP), and extraction is performed. After extraction, the sample vial is placed in an ice bath or a refrigerator freezer, and the AP solidifies rapidly on the DP surface. After the solid is removed and melted at room temperature, the analyte enhanced in the AP is determined [28]. 1-undecanol (M.P. 13 ◦ C–15 ◦ C) [128,133], 1-dodecanol (M.P. 22 ◦ C–24 ◦ C) [134], hexadecane (M.P. 18 ◦ C) [135], undecanoic acid [30] etc. are often used. In addition, 1-bromo hexadecane (M.P. 17 ◦ C–18 ◦ C), 1,10-dichorodecane (M.P. 14 ◦ C–16 ◦ C), 2-dodecanol (M.P. 22 ◦ C–27 ◦ C) and 1-hexadecanethiol (M.P. 18 ◦ C–20 ◦ C) are also reported. Toraj and coworkers [136] employed SFOME combined with HPLC/UV to determine the low concentration of the complex matrix of real samples taken from addicted persons and patients under morphine and codeine treatment. Xue et al. [137] introduced SFOME combined with HPLC for the determination and evaluation of five low abundance lignan compounds in Schisandra Chinensis (Turcz.) Baill from different origins. The researches also believed that in SFOME, the extraction solvent n-dodecyl diffused rapidly in water under a certain stirring speed and orderly arranged to form dodecanol MOOA. In the MOOA, the hydrophilic -OH pointed to the external water system and hydrophobic alkyl group gathered to constitute the hydrophobic inner core. The analyte molecules with different polarities, which either were distributed in different positions or entered into the hydrophobic inner core of the dodecanol’s MOOA, were extracted. Under the condition of below freezing point of the extraction solvent, the dodecanol’s MOOA forms fruity and dense solid supermolecular aggregation by hydrogen bond and float on the top of the aqueous sample phase. Recently, Zhang et al. [138] combined two DLLMEs of both low- and high-density extraction solvents in one step, developed a novel LPME technique termed ionic liquid-water-organic solvent three phase microextraction (IL-W-OS-3p-ME) and introduced the method for the simultaneous preconcentration and determination of flavonoids and anthraquinones in TCMs. The procedure was performed in one step using a syringe and two extraction solvents (the low- and high-density APs are not only insoluble but also

8

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

non-emulsified in DP; however, both APs are immiscible). Firstly, 3.0 mL of the DP was placed in a 5 mL syringe. Next, 100 ␮L of heptanol (AP1 ) and 100 ␮L of [HMIM][PF6 ] (AP2 ) were injected separately into the sample solution with a syringe. A cloudy emulsion was formed after 25 s of manual shaking. The emulsion was allowed to stand for 3 min. Soon afterwards, the emulsion readily separated into three clear phases: the AP1 mainly contained the flavonoids floating on the surface of the DP, the aqueous sample DP and the AP2 that mainly contained the anthraquinones settling at the bottom of the syringe. With the syringe held pointing upward, the syringe plunger was pressed to transfer the upper layer AP1 into a small EP tube. By rotating the syringe so that it pointed downward and allowed to settle for 1 min, the lower AP2 was recovered into the above-mentioned EP tube from the syringe tip by slightly pressing the plunger. Next, the solution was diluted with 20 ␮L of HPLC-grade methanol. Then, the mixed solution was injected into the HPLC system for analysis. Under optimized conditions, the EFs of the flavonoids were ranged between 102 and 230, and those of the anthraquinones were from 102 to 211, the LODs were below 0.05 ng mL−1 , and the precisions and recoveries were 5.2% to7.8% and 80.1% to 105.8%, respectively. In addition to the above mentioned, DLLME and SFOME have been widely supplied in vivo and in vitro from the following natural product, such as phenylpropionic acid compounds [124], anthraquinones [123,138], flavonoids [138], furocoumarin [139], oleanolic acid and ursolic acid [140] in TCMs, as well as volatile constituents in tea [141], essential oil constituents of the plant Oliveria decumbens vent [142], auxins in plants [143], endogenetic plant growth regulators in common green seaweeds [144], emodin and its metabolites in urine samples [55] and opium alkaloids in human plasma [136]. In addition, several metal compounds have also been analysed, such as total and water soluble copper [145] and vanadium [28]. All the specific parameters are listed in Table 2. The main advantages of the technique include simplicity, low cost, short extraction time, simple apparatus and minimum organic solvent consumption [69]. However, the extraction solvent is susceptible to contamination by sample impurities and the matrix because of its exposure to the sample solution during extraction. In addition, the precision of analytical result is limited, and the two can be mentioned as the technology’s drawbacks. 2.3. Hollow fibre liquid-phase microextraction To enhance single drop stability and reduce AP pollution from impurities or substrate in DP in the above mentioned SD-LPME, hollow fibre liquid-phase microextraction (HF-LPME) was introduced in 1999 by Pedersen-Bjergaard and Rasmussen as a simple and effective LPME [46]. Based on the AP in/on hollow fibre (HF) type, the HF-LPME can be classified into two-phase HF-LPME (2p-HF-LPME) [146,147] and three-phase HF-LPME (3p-HF-LPME) [46,148,149]. In 2p-HF-LPME (Fig. 1E and F), the organic solvent is immobilized in the pores and filled in the HF lumen, and the analytes are extracted by passive diffusion from the DP into the organic AP in the HF lumen [150]. A weak acid in acidic DP medium, weak alkali in alkaline DP medium or neutral substances compounds with high solubility in non-polar organic solvent AP can be extracted by a two-phase system. In the 3p-HF-LPME (Fig. 1G), three liquid phases participate in analyte extraction: aqueous sample DP containing the target analyte, the water-immiscible organic extraction phase immobilised in the wall pores of the HF and the aqueous AP in the HF lumen [149]. The target analytes in DP are first extracted by an organic solvent immobilised in the HF pores, and then further extracted into the aqueous AP in the HF lumen. Acid, alkali or strong dielectric compounds with poor solubility in organic solvent can be back-extracted to the aqueous AP of the three-phase system [25]. The enriched-target analytes in organic AP or aqueous AP are

directly or indirectly detected by UV, GC, AAS, HPLC, CE [151,152], UHPLC/MS and HPLC/MS [153]. In HF-LPME, extraction solvent is also crucial for achieving good selectivity and high EE of the target compound. The extraction solvent should have enough dissolving capacity for the analyte, similar polarity to the fibre and no reaction with any of the compounds in the sample solution. Only one kind of extraction solvent is used in general, but mixed solvent is used at times to obtain the best EE [154,155] is used. When Xi et al. [154] studied the effect of 20 organic solvents on the extraction capacities of four hyoscyamines compounds in TCMs, the group found that 7:3 (v:v) of xylene and n-heptanol mixed solvent had the best extraction effect on the four hyoscyamines. HF is another important factor for achieving optimal extraction and concentration of the target analytes. In addition to serving as carrier to protect the AP and micro-strainer to purify sample solution [156,157], HF also participate in target analyte extraction and improve the analyte EE or EF [158]. Wang et al. [158] used HF-LPME combined with HPLC for the preconcentration and quantitation of phenylpropionic acid active constituents in TCM preparations. While investigating the different HFs as carriers to extract phenylpropionic acids, they found that, under the same condition, polyvinylidene difluoride (PVDF) had the highest phenylpropionic acids EEs, so the researchers discussed and analysed the HF-LPME mechanism based on PVDF. In the paper, they proposed that a strongly electronegative fluorine (F) exists in the PVDF structure, making the F electron-rich in PVDF surface. Meanwhile, the benzene ring in phenylpropionic acids is electron-deficient given its conjugated with the attractive electronic carbanyl group. In the extraction process, the electron-deficient benzene ring in the target analyte and electron-rich PVDF surface easily form the orderly charge-transfer supramoleculars (Fig. 3), which make the phenylpropionic acids concentration increases on the HF surface and enhance their EE or EF. Zhou et al. [159] developed a 2p-HF-LPME with magnetofluid coupled to HPLC with ultraviolet detection and successfully applied it to separate and determine four phenylethanoid glycosides (PhGs) (echinacoside, tubuloside B, acteoside and isoacteoside) in rat plasma after oral administration of Cistanche salsa extract. Under optimised conditions, the EFs for PhGs exceeded 625. The calibration curve for PhGs was linear in the range between 0.1 and 100 ng mL−1 , with correlation coefficients greater than 0.9996. The intra-day and inter-day precisions (RSDs) were below 8.74%, and the LODs for the four PhGs ranged between 8 and 15 pg mL−1 . To date, 2p- or 3p-HF-LPME has been successfully employed for analysing and researching the amounts of compounds in natural product samples. 2p-HF-LPME has been used for analysing and researching the anthraquinones compounds in Radix Polygoni Multiflori [160], emodine and its metabolites in rat urine and plasma [161], protein binding properties of furocoumarins [162] and flavonoids compounds [163] to bovine serum albumin, osthole pharmacokinetics in cerebral ischemia hypoperfusion rat plasma [164], and permeable biomembrane ingredients in TCMs [165,166]. 3p-HF-LPME has been employed to preconcentrate and determine oxymatrine and matrine [167], phenylpropionic acids [158], cinnamic acid and its derivative [168], valerenic acid [169], flavonoids [151], magnolol and honokiol [156], aristolochic acid [170], anthraquinones [171] in Rhubarb, studies on protein binding rates [172], extraction mechanism [173,174], the preferred conformation of ephedrine and pseudoephedrine [174], as well as correlativities between the EF of the target analyte and its pKb [175] or pKa [176]. The presence of charged compounds used in liquid extraction procedures led several authors to propose the use of electrical fields to enhance and manipulate HF-LPME [177–184, 195]. For the first time in 2006, Pedersen–Bjergaard and Rasmussen [185], proposed

Table 2 Applications of DLLME in natural product analysis. Sample

Analyte

Mode

Extraction solvent

Disperser solvent

Donor Phase

Extraction time/ temperature

Analytical method

Limit of detection

Recoveries (%)

Enrichment factor

Ref.

Tap Radix et Rhizoma Rhei

Vanadium Aloe-emodin, Rhein, Emodin, Chrysophanol, Physcion Emodin, Aloe-emodin, Rhein, Danthron, Chrysophanol, Physcion

SFODME DLLME

1-Undecanol [HMIM][PF6 ]

Acetone Methanol

pH = 3 aqueous solution pH 2.0 H3 PO4 solution

10 s/270 ◦ C 10 min/ 60 ± 1 ◦ C,

ETAAS HPLC, DAD

96–105 95.2–108.5

CHCl3

THF

pH = 1 aqueous solution

4 min

HPLC/UV

87.1–105

184 176 174 209 192 213 90–295

[28] [109]

DLLME

7 ng L−1 1.23 0.50 2.02 1.87 1.20 ␮g L−1 70–400 pg mL−1

Aloe-emodin, Rhein, Emodin, Chrysophanol, Physcion Cinnamic acid, Ferulic acid, Caffeic acid, p-hydroxycinnamic acid, ferulic acid Morphine, Papaverine, Codeine, Noscapine Psoralen, Oxypeucedan, Imperatorin, Isoimperatorin, Oleanolic acid, Ursolic acid

DLLME

[HMIM][PF6 ] [HMIM][PF6 ]

Acetonitrile Methanol

pH = 2 aqueous solution

7 min/40 ◦ C

HPLC

100–1000 pg mL−1 0.01–0.09 ␮g L−1

94.8–103 81.7–110.9

63–192 80–197

[123]

DLLME

[HMIM][PF6 ]

THF

pH = 3, 127 g L−1 NaCl aqueous solution

3 min

HPLC/UV

0.01–0.13 ng mL−1

86.9–112.6

56.0–159.3

[124]

SFODME

1-Undecano

Acetone

0.5 min

HPLC/UV

0.5–5 mg L−1

88–110.5

110.4–165

[136]

DLLME

CCl4

Acetonitrile

pH = 9, 1% (w/v) NaCl aqueous solution 2.5%(w/v) NaCl aqueous Solution

3 min

HPLC/UV

1.00–3.0 mg L−1

97.5–114.8

12 6–38.5

[139]

DLLME

Chloroform

Methanol

pH = 2, 10% (w/v) NaCl aqueous solution

2 min

HPLC/UV

0.02 mg mL−1

85.7–116.2

933, 1378

[140]

DLLME

Chloroform

Methanol

7.4% (w/v) NaCl aqueous solution

21 min/ 32 ◦ C

GC/MS

0.3 mg L−1

–

4.0–42.6

[141]

DLLME

Chlorobenzene

Acetonitrile

Aqueous Solution

Room temperature

GC

0.2–29 ng mL−1

89–93

–

[142]

DLLME

CHCl3

Acetone

pH = 4, 7.5% (w/v) NaCl aqueous solution

0.1 min

HPLC/FLD

0.02–0.1 ng mL−1

94.7–116

10–60

[143]

DLLME

CHCl3

Acetone

0.1 min

HPLC

0.2–1.0 ␮g mL−1

80–120

–

[144]

DLLME

CCl4

Ethanol

pH = 4, 7.5% (w/v) NaCl of aqueous solution 1.0% (w/v) NaCl of solution

A few seconds

CE

32 ng mL−1

88.3–94.4

–

[145]

Urine samples

Radix et Rhizoma Rhei

Human plasma (persons addicted to opium) Radix Angelicae

Hedyotis, Diffusa, Eriobotrya, Japonica Tea plant

Oliveria decumbens Vent

Chlorella vulgaris, Duranta

Green seaweeds

Rhizoma coptidis

Limonene, Cumaldehyde, Caffeine Beta-Pinene, Beta-Myrecene, p-Cymene, Limonene, gamma-Terpinene, Thymol, Carvacrol, Myristicine Indole-3-acetic acid, Indole-3-butyric acid, Indole-3-propionic acid, 1-Naphthylacetic acid Gibberellic acid, Abscisic acid

Copper

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

Rhizomatyphonii

[55]

9

10

Y. Yan et al. / J. Chromatogr. A 1368 (2014) 1–17

Fig. 3. Schematic of PVDF and phenylpropionic acids forming charge transfer supermolecule.

electromembrane extraction (EME). In essence, the method is similar to a 3p-HF-LPME [186], but migration in EME system through an organic solvent immobilised in the HF wall (e.g. supported liquid membrane, SLM) is forced by an electrical field generated from two electrodes, one that is placed outside the HF (DP) and the another inside the HF lumen (AP). To ensure efficient electrokinetic mobility in the EME system, pH must be adjusted to provide total ionisation of the analytes in the two aqueous solutions. In a two-phase EME recently described by Davarani et al. [187], the researchers discovered that the application of an electrical potential could significantly increase mass transfer and reduce extraction time [188]. In various fields, an increasing number of applications of carbon nanotubes (CNTs) have been presented because of CNT’s high selectivity and good extraction efficiency in case of organic analytes extracted from aqueous samples [189]. These features present a potential use for CNTs as adsorbing material in reinforcing LPME. Es’haghi et al. [190] succeeded in promoting the HF-LPME technique by inserting multi-walled CNTs (MWCNTs) into the pores of a HF in the measurement of caffeic acid in echinacea purpurea herbal extracts. In the test, the MWCNTs dispersed in the organic solvent were held in the pores of a porous HF supported by capillary forces and sonication. The CNTs then remained in contact with the aqueous DP and the aqueous AP. The analyte from the DP diffuses into the MWCNTs dispersed in the organic solvent and filled the HF pores, and was then back-extracted into a small volume of aqueous AP and enriched. The method showed good linearity (0.1 ng L−1 –50 ␮g L−1 ), repeatability, LOD (0.05 ng L−1 ) and excellent enrichment (EF = 2108). Chen et al. [173] had applied MWCNTs reinforcing LPME to the separation and determination of three cinnamic acid derivatives, coffee acid, ferulic acid and cinnamon acid in pollen. The inter-day and intra-day precision values of the experiment were less than 9.8% (n = 9), the average recoveries ranged from 93.8% to 115.2% and the EFs of the three analytes ranged between 514 and 1084. In addition, Chen and his group also discussed the mechanism of CNT-reinforcing HF-LPME and thought that hydroxylation CNTs, which are rich in hydroxyls on ports and possess larger specific surface area, are excellent adsorbent; octanols, which not only being rich in hydroxies and having proper polarity, are a good extraction solvent. In a CNTs/octanol dispersed system, each CNT constitutes a SPME unit, and each octanol forms a LPME unit. When the micropore in the HF wall was inlaid or full of CNTs/octanol dispersive solution, each micropore constitutes a CNTs/octanol solid-liquid microextraction (SLME) array or unit beam. The multipath or multiply extraction mode improves the adsorption and extraction capability of the analyte, so the group proposed a HF-CNTs/octanol solid-liquid synergistic microextraction (HF-CNT-O-SLSME) mechanism and defined the EFS/L . synergistic microextraction coefficient (SMEC): RSMEC = EF +EF S

L

EFS/L , EFS and EFL were the EF of analyte in HF-CNT-O-SLSME, in CNTs-SPME and in octanol-LPME, respectively. In the study, the SMEC of coffee acid, ferulic acid and cinnamon acid were 4.2, 2.7 and 1.9, respectively. Another useful LPME technique based on HF is known as HF solvent bar microextraction (HF-SBME) (Fig. 1F), which was first proposed by Huang et al. [191] in 2012 and has been applied to separate and concentrate phthalate esters from aqueous matrices. Ma et al. [192] have recently used an improved HF-solvent bar (HF-SB), in which HF of PVDF (1.0 cm) was wrapped by octanol (about 2 ␮L) on its outside surface and coupled with HPLC to extract, preconcentrate, and simultaneously determine lipid soluble active ingredients and water–soluble ingredients in Radix Salvia miltiorrhiza. Under the most favourable conditions, the EFs of the analytes were 0.7–612, and the LODs were below than 1.11 ng mL−1 . Hao and coworkers [140] used the improved HF–SBME combined with HPLC to rapidly determine oleanolic acid (OA) and ursolic acid (UA) in TCMs, investigate the protein-binding rates between OA or UA and human serum albumin (HSA), and calculate and obtain the numbers of binding site and binding constants of OA and UA. Compared with HF-LPME or the conventional HF-SBME, the improved HF-SBME did not require injecting and removing the AP using a microsyringe before and after the extraction, as well as prevented sealing both ends of HF-SB. However, the method has certain limitations, such as susceptibility of the AP to pollution by sample impurities or matrix and limited precision because the system is unprotected. A novel microporous membrane/solvent microextraction (MPMSME, as Fig. 1H–J) was recently developed [193,194] and introduced for extraction and preconcentration of active compounds in TCMs. In this approach, filter paper (or membrane) scrap was cut into small square pieces (1 cm2 ), which was immersed into the extraction solvent for 10 s. Then, the filter paper were removed from the solvent using forceps and placed on absorbent paper to remove the excess extraction solvent on the membrane surface. Afterwards, a solvent-impregnated membrane piece was attached to a clean microsyringe needle that was pre-inserted vertically through the centre of a rubber stopper. The rubber stopper was placed on a vial containing DP, and the membrane was submerged in the DP. Extraction was performed using a magnetic bar at 900 rpm. After extraction, the membrane was removed from the needle and placed in double-distilled water for 5 s to wash the impurities or sample matrix. The enriched analytes on the membrane were dissolved in methanol, and the eluate was analysed. Xing et al. [193] selected the cinnamic acid derivatives as model analytes to evaluate and assess MPMSME (Fig. 1H). Parameters affecting MPMSME, such as type of extraction solvent, membrane area (or volumes of extraction solvent), pH and ionic strength in DP, extraction stirring rate and time, and DP volume

Table 3 Applications of HF-LPME in natural product analysis. Sample

Analyte

Pine needles from the Pinus pinea 13 Polycyclic aromatic L. species hydrocarbons

Hollow fibre

Extraction Solvent

Donor phase

Acceptor Phase

Extraction time/ temperature

Analytical method

Limit of detection

Recoveries (%)

Two -phase

Polypropylene

Hexane or Toluene

Toluene

20 min

GC/MS

0.01–0.95 ng g−1

64–160

pH 9.75

80 min

HPLC/UV

40 min

CE

50 min

UHPLC/MS

0.002–0.054 ␮g L−1

0.01–0.03 mg L−1

Echinophora platyloba DC., Mentha piperita

Morin, Naringenin, Quercetin, Luteolin, Kaempferol, Apigenin

Three -phase

Polypropylene

1-Octanol

Liquid sonicated extract in a 20% (v:v) acetone solution pH 2

Urine

Strychnine, Brucine

Three -phase

Polypropylene

1-Octanol

0.5 mol L−1 NaOH

Polygonum hydropiper L.

Rutin, Hyperin, Isoquercitrin, Quercitrin, Catechin, Epicatechin, Quercetin, Kaempferol, Isorhamnetin

Three -phase

Polyvinylidene fluoride

Ethyl acetate

Raceanisodamine tablets, belladonna tablets

Anisodamine Hydrobromide, Three -phase Atropine sulphate, Scopolamine, Hydrobromide, Scopolamine, Butylbromide

Polypropylene

Dimethylbenzene /n-Heptanol (V/V = 7:3)

Magnolia Officinalis

Magnolol, Honokiol

Three -phase

Polyacrylonitrile

Caffeic acid, Ferulaic acid, p-Hydroxycinnamic acid, Methoxy cinnamic acid, Cinnamic acid Plasma Echinacoside, (after oral Tubuloside B, administration of Cistanche salsa Acteoside, Isoacteoside extract)

Three -phase

Radix Polygoni Multiflori praeparata

Plasma, Urine

100 mmol L−1 H3 PO4 50 mL of original extracting 10 mmol L−1 solution (pH 3.0) NaHCO3 solution; pH 8.5

1.5 6 4 0.5 7 3 ng mL−1 1 ng mL−1 2 ng mL−1

1.0 × 10−3 mol L−1 NH3 ·H2 O

5 × 10−3 molL−1 HCl

100 min

HPLC

n-Butanol

pH 5 HCl (25% methanol)

pH 12 NaOH

30 min

HPLC

Polyvinylidene fluoride

n-Heptanol

pH 3 HCl

pH 11.7 NaOH

60 min

HPLC

Two -phase

Polypropylene

n-Octanol

1.5% (w/v) NaCl.

Fe3 O4 magnetofluid n-octanol

5 min

HPLC/UV

Aloe-emodin, Rhein, Emodin, Chrysophano, Physcion

Two -phase

Polyvinylidene fluoride

n-Octanol

2 mmol L−1 HCl (50%methanol)

n-Octanol

60 min

HPLC

0.25–0.30 ␮g L−1

Aloe-emodin, Rhein, Danthron, Emodin, Chrysophano, Physcion Psoralen, Oxypeucedanin, Imperatorin, Isoimperatorin,

Two -phase

Polyvinylidene fluoride

n-Octanol

20 mmol L−1 HCl (plasma), n-Octanol 12 mmol L−1 HCl (urine)

40 min

HPLC

0.1–0.3 ␮g L−1

Two -phase

Polypropylene

1-Heptanol

Analytical drug containing serum or bovine serum albuminsamples

60 min

HPLC/UV

Plasma

Osthole

Two -phase

Polypropylene

1-Octanol

5 m (2 mL plasma sample)

n-Octanol

20 min/ 25 ◦ C

HPLC/ Fluorescence detection

Sophora flavescens Ait

Oxymatrine Matrine

Three -phase

Polyacrylonitrile

Isopropyl alcohol

pH 9 NaOH

pH 4 HCl

30 min

2.0 × 10−4 mol L−1 HCI

0.008 mol L−1

Shuanghuanglian peroral liquid, Danggui concentrate pellet, Guizhifuling pellet

Traditional Chinese medicines

Traditional Chinese medicines

Valeriana officinalis

Cinnamic acid, Hydroxy-cinnamic acid, p-Methoxy cinnamic acid

Three -phase

Polyvinylidene fluoride

Heptanol

Valerenic acid

Three -phase

Polypropylene

Dihexyl ethe

0.10 ␮g mL−1 , 0.07 ␮g mL−1 0 004–0.1 g L−1

8 15 10 12 pg mL−1

pH 9.5

Ref.

[33]

293 245 16 311 146 216

[149]

50, 35

[152]

95.2–99.8

36–83

[153]

95.0–119, 93.0–95.0

11–16

[154]

92–99

98.3–105.1