3.1. Solid Phase Extraction
Briefly, classical solid-phase extraction (SPE) process consists of percolating the sample solution through a cartridge (or disc) containing the solid sorbent that retains the target analytes, whereas the rest of the sample is discarded. After a cleaning step, an elution solvent is passed to desorb and to retrieve the analytes.
As it can be seen in
Table 1, different nanomaterials have been packed in SPE-cartridges for the determination of cosmetic-related compounds. In this sense, Márquez-Sillero et al. [
28] employed MWCNTs for the determination of four parabens in cosmetic products, previously lixiviated with water. Wang et al. [
29] developed a GO sponge for the determination of six benzotriazole compounds in sewage and cosmetic samples. The use of MIPs in SPE for cosmetic analysis was first proposed by Zhu et al. [
30], who used MIP-coated silica nanoparticles for the determination of bisphenol A (BPA) in shampoos and bath lotions, which were previously lixiviated with toluene before introducing them into the cartridge. Later, Wang et al. [
31] functionalized MWCNT with a prednisone-template MIP for the determination of this glucocorticoid. Zhong et al. [
32] employed carboxylated GO with polyvinyl chloride (PVC) as sorbent to determine different sulphonamides as contaminants in cosmetics products, and Abdolmohammad-Zadeh et al. [
33] created a LDH cartridge with nickel and zinc for the analysis of p-aminobenzoic acid in cosmetic samples, which was dissolved in a proper water or ethanol amount before the extraction.
It should be noticed that SPE is not in fact a microextraction technique, but the use of nanomaterials as sorbents allows to achieve low LODs (from ng mL−1 to ng L−1), which are suitable for the trace analysis of cosmetic-related compounds in the different matrices considered.
3.2. Solid Phase Microextraction
Solid phase microextraction (SPME) was developed by Arthur and Pawliszyn in 1990 [
34]. In this technique, analytes are retained on a fibre coated with the sorbent material. The extraction can be performed by direct immersion into the sample or, if the analytes are volatile enough, by setting the fibre in the head space. After the extraction, analytes are, usually, thermally desorbed, although in a minor extent, liquid desorption in an appropriate solvent has also been used. Several methods based on SPME have been employed for the extraction of cosmetic-related compounds from different matrices. They are all listed in
Table 2.
With that aim, different works for determination of parabens in different matrices have been reported. Ara et al. [
35] modified mesoporous silica nanoparticles with polyaniline (PANI) and p-toluene sulphonic acid to coat the fibre for the determination of three of these target compounds in various cosmetics creams and wastewater. Yazdi et al. [
36] determined the same parabens in wastewater samples employing AgNPs embedded on polypyrrole. First of all, pyrrole was polymerized on the hollow fibre, and then, it was introduced in a suspension of AgNPs for bounding.
For the determination of UV filters in environmental samples, different titanium oxide-based fibres have been used due to their excellent properties, such as high chemical and thermal stability, low cost and toxicity and good biocompatibility. In this sense, Du and coworkers used PANI-coated titania nanotubes (NTs) [
37], ZrO
2-based fibre [
38] and TiO
2 NPs functionalized with phenyl groups [
39] for the analysis of different UV filters in river water and wastewater. The same authors also used electrodeposited AuNPs onto a stainless-steel wire followed by a coating step with 1,8-octanedithiol [
40] for the same purpose. Moreover, Mei et al. [
41] synthesized a polymeric ionic liquid (PIL) with MNPs to enhance the extraction capability of diamagnetic UV filters employing magnetic field gradients. This method was applied to lake and river waters and wastewater.
In addition to parabens and UV filters, extraction of other cosmetic-related compounds has been also performed by SPME. Wu et al. [
42] employed a graphitic carbon nitride (g-C
3N
4) modified with rGO for the analysis of six polycyclic aromatic hydrocarbons (PAHs) in cosmetic products previously diluted in water. Tong et al. [
43] synthesized a polymeric monolith by copolymerization of butyl methacrylate (BMA) and ethylene dimethacrylate (EDMA), followed by the addition of rGO nanosheets for the analysis of nine glucocorticoids (GCCs). In this methodology, GCCs were first extracted with acetonitrile (ACN) and then, SPME was performed. Finally, Wang et al. [
44] used hydroxyapatite (HAP) NPs to coat a titanium fibre for the analysis of different chlorophenols, BPA and triclosan (TCS) in river water and sewage.
All the analyses reported with SPME show a great sensitivity, proving to be one of the most appropriate techniques for the analysis of traces. As shown in
Table 2, the lowest LODs are achieved for those methods focused on the analysis of environmental samples (mostly waters). Since cosmetic matrices are usually difficult matrices, a clean-up step with organic solvents is usually required, which reduces the sensitivity of the method due to the dilution effect, but in any case, the achieved LODs are low enough to analyse the cosmetic samples.
3.4. Dispersive Solid Phase Extraction
Dispersive solid phase extraction (DSPE) has become a widely used extraction technique since its proposal by Anastassiades et al. in 2003 [
50]. Traditionally, the sorbent is introduced and dispersed into the sample. When the extraction is completed, the sorbent is recovered by means of centrifugation and decantation. However, nowadays, this technique has gained more interest due to the introduction of magnetic materials as sorbents, allowing an easy recovery of the sorbent by employing an external magnetic field, which considerably reduces the analysis time.
As can be seen in
Table 4, 38 articles employing DSPE for the determination of cosmetic-related compounds have been reported, and only 6 of them resort to nonmagnetic sorbents, which shows the high impact that magnetic materials have caused in this extraction technique. In this sense, Rocío-Bautista et al. [
51] used the MOF HKUST-1 in vortex-assisted DSPE for the extraction of a group of seven parabens in cosmetic creams, urine and environmental waters. Rashvand et al. [
52] also analysed two parabens in wastewater samples by employing a GO-PANI composite. Li et al. [
53] dispersed the MOF MIL-101 (Cr) in toner samples for the determination of different benzophenones. Gao et al. [
54] synthesized a TCS-based MIP on CNTs in order to extract this analyte from lake and river waters. Zhai et al. [
55] developed a method for the determination of hormones employing the MOF MIL-101 that was dispersed into the cosmetic sample after its dilution in a saline solution. Finally, Liu et al. [
56] achieved the extraction of Hg(II) from cosmetic samples by measuring the fluorescence of CDs obtained from grass carps after their interaction with the analyte.
On the other hand, several magnetic composites have been reported, especially focused on the study of parabens and TCS in different matrixes and UV filters in environmental samples. With that aim, Tahmasebi et al. [
57] used PANI-coated Fe
3O
4 MNPs for the determination of three parabens in wastewaters, cosmetic creams and toothpaste. Ghambari et al. [
58] employed recycled polystyrene (PS) to synthesize a composite with CoFe
2O
4 MNPs to determine a group of four parabens in river, creek and tap waters by using vortex to disperse the composite. Abbasghorbani et al. [
59] used a magnetic composite of aminopropyl (AP) and Fe
3O
4 MNPs for the determination of five parabens in different aqueous samples. Ariffin et al. functionalized the Fe
3O
4 with different surfactants, such as Sylgard 309 [
60] and DC193C [
61], for the extraction of different parabens in lake, river and sea waters. Casado-Carmona et al. [
62] created a hybrid material based on MNPs and an IL (i.e., MIMPF
6) for the determination of four parabens along with some benzophenones and BPA in pool waters. The extraction was performed by dispersing the Fe
3O
4@MIMPF
6 by ultrasounds and employing vortex agitation to achieve the adsorption of the analytes. Mehdinia et al. [
63] immobilized self-doped PANI on a Fe
3O
4-rGO composite for the determination of various parabens in different cosmetics (sunscreen, toothpaste and moisturizing cream) pretreated with MeOH. Later, the same authors [
64] compared different silica-based magnetic nanocomposites for the extraction of parabens from various cosmetic samples. Feng et al. [
65] also worked with rGO for determination of two parabens in cosmetic samples. In this case, Fe
3O
4 MNPs were embedded into the rGO surface, and then, it was covered by layers of mesoporous silica (mSiO
2) with phenyl-functionalized pore walls. Ultrasounds were employed for the dispersion of the material.
Jalilian et al. [
66] modified the MOF MIL-101 surface with Fe
3O
4 MNPs and MWCNTs for the determination of two parabens along with three phthalates in both cosmetic creams and tap water. Cosmetic products were previously dissolved in MeOH:H
2O before the extraction. The use of a COF as sorbent was proposed by Shavar et al. [
67], who functionalized Fe
3O
4 MNPs with a covalent triazine-based COF for the determination of a group of four parabens in water, cosmetic products and breastmilk. Yusoff et al. [
68] synthesized a magnetic composite with Fe
3O
4 MNPs coated with the IL 1-butyl-3-methylimidazolium chloride. This sorbent was applied to the extraction of four parabens in river, pond and lake waters and in MeOH pretreated cosmetic creams. Pastor-Belda et al. [
69] precipitated Fe
3O
4 MNPs on MWCNTs surface for the analysis of several parabens in water and urine. Ghasemi et al. [
70] employed γ-Fe
2O
3 MNPs coated with HAP to determine six parabens in soils, water and urine assisted by ultrasounds. Before DSPE procedure, soil samples were lixiviated in water.
Regarding the analysis of UV filters in environmental samples, Wang et al. [
71] performed the extraction of three benzophenones in soils with Fe
3O
4 MNPs combined with MOF-1210 (Zr/Cu). Piovesana et al. [
72] employed graphitized carbon black (GCB) prepared with MNPs for the extraction of 10 UV filters in different surface waters. Cheng et al. [
73] used polydopamine-coated Fe
3O
4 MNPs for the analysis of 11 UV filters in wastewaters. Román-Falcó et al. [
74] covered the CoFe
2O
4 MNPs surface with oleic acid. The extraction and subsequent determination of six UV filters was accomplished in tap, river and sea waters. Giokas et al. [
75] developed a method for the determination of four UV filters, consisting a cloud-point (CP) extraction followed by a DSPE step in the micellar phase using core–shell Fe
2O
3@C coated with polysiloxane (PSx).
Regarding to TCS determination, Yang et al. [
76] performed the microextraction in toothpastes previously lixiviated in MeOH. For the DSPE step, MIL-101 MOF was functionalized with Fe
3O
4. Li et al. [
77] analysed TCS and triclocarban (TCC) in biological samples employing a magnetic COF formed by the condensation of 1,3,5-tris(4-aminophenyl) benzene (TAPB) and terephthaldicarboxaldehyde (TPA) on the surface of the MNPs. Li et al. [
78] employed GO embedded with magnetic iron nanowires for the analysis of TCS in lake water and wastewater along with BPA, and Jiang et al. [
79] synthesized a Fe
3O
4-PANI composite for the extraction of TCS, BPA and 2,4-dichlorophenol from water samples.
Besides those compounds mentioned before, other analytes have been also determined using nanomaterials. Three works have been reported on the analysis of GCCs in cosmetic products. Du et al. [
80] employed Fe
3O
4 coated with a MIP for the determination of dexamethasone in skincare products. Liu et al. [
81] prepared a magnetic composite based on MNPs coated with a dual template MIP for the determination of hydrocortisone and dexamethasone from different cosmetic products (lotions, masks and toners), which were previously treated with a saturated NaCl solution and acetonitrile (ACN). Finally, Li et al. [
82] determined five GCCs in facial masks previously sonicated in ultrapure water, employing magnetically functionalized g-C
3N
4 bonded to MIL-101 MOF.
Moreover, the determination of the dye rhodamine B in different matrices has been also performed. In this regard, Khani et al. [
83] worked with γ-Fe
2O
3 MNPs coated with imino-pyridine on hand washing soaps. Before the DSPE step, the samples were dissolved in water. Bagheri et al. [
84] used Fe
3O
4 MNPs functionalized with poly(aniline-naphthylamide) (PAN) for its determination in shampoos, eye shadows and hand washing products.
Tarigh et al. [
85] worked with lipstick samples for the determination of lead and manganese employing a composite of Fe
3O
4 MNPs and MWCNT. Before the extraction, samples were mineralized at 450 °C, and subsequently, the ashes were dissolved with nitric acid. Xia et al. [
86] determined whitening agents working with Fe
3O
4 MNPs coated with a polymeric COF based on benzidine and 1,3,5-triformylphloroglucinol. Liu et al. [
87] synthesized a MIP-coated Fe
3O
4 MNPs for the determination of metronidazole in cosmetic creams, lotions and powders, previously lixiviated with MeOH. Finally, Maidatsi et al. [
88] prepared a magnetic composite of Fe
3O
4 MNPs and rGO functionalized with octylamine to determine different musks, allergens and phthalates in water samples. More recently, Zhang et al. [
89] employed halloysite nanotubes (HNTs) that where first filled with CoFe
2O
4 MNPs and later assembled with Au-NPs on its surface using APTES. This composite was applied for the determination of 4,4′–thioaniline in hair dyes.
As described in
Table 4, LODs between μg mL
−1 and ng L
−1 are achieved in DSPE-based methods, although as expected, the instrumental technique has a huge impact on this parameter. In this sense, despite LC-UV has been extensively used, it might be not enough sensitive for the determination of trace levels of some of the cosmetic-related compounds. For this reason, other options, such as LC-MS/MS, have been preferred.
Extraction times are similar regardless of the use of magnetic materials or not. However, the use of the magnetic ones avoids centrifugation steps to recover the sorbent in the extraction and desorption steps, which redounds in the reduction of the total time of analysis.
Compared with other techniques, DSPE combined with nanosorbents allows excellent LODs, many times comparable with SPME and SBSE, but with the advantage of shortest extraction times, usually under 20 min.
3.5. Stir Bar Sorptive-Dispersive Microextraction
A hybrid approach combining DSPE and SBSE, termed stir bar sorptive-dispersive microextraction (SBSDME), was introduced in 2014 by Benedé et al. [
90]. In this technique, a magnetic sorbent coats the stir bar by means of magnetic interactions. When the stirring rate is high enough, the sorbent is dispersed in the sample until the stirring is stopped; at that moment, the magnetic composite containing the analytes is retrieved by the stir bar. This approach has been also employed for the determination of cosmetic-related compounds in different matrices (
Table 5). Benedé et al. developed different strategies for determination of UV filters in environmental samples employing CoFe
2O
4 MNPs coated with oleic acid for the analysis of eight hydrophobic UV filters in environmental samples [
90,
91,
92]. Later, the same authors developed a method based on CoFe
2O
4 MNPs embedded on nylon-6 polymer for the determination of six hydrophilic UV filters [
93].
Recently, SBSDME has been applied for the determination of other types of analytes in different matrices. Grau et al. [
94] applied this technique for the study of triphenyl phosphate (TPP) and its metabolite, diphenyl phosphate (DPP), in urine samples by means of CoFe
2O
4 incrusted onto a weak anion exchanger (Strata X-AW). Miralles et al. [
95] functionalized MIL-101 MOF with CoFe
2O
4 for the determination of eight N-nitrosamines in cosmetic products. In this methodology, N-nitrosamines were first pre-extracted in hexane and then preconcentrated with SBSDME. Finally, Vállez-Gomis et al. [
96] determined 10 PAHs in cosmetic creams employing rGO covered with CoFe
2O
4 MNPs, where samples were previously extracted with hexane, and then, SBSDME was performed on the hexane solution.
As can be seen in
Table 5, similar LODs are obtained for SBSDME and DSPE with comparable extraction times. The main difference between these two approaches relies where the magnet is positioned, i.e., outside the solution in DSPE and inside the solution in SBSDME; thus, this last one does not require an external magnet to retrieve the sorbent. This fact alleviates losses of the extractant material in the different steps due to the reduction of sorbent and sample manipulation.
3.6. Other Sorbent-Based Microextraction Approaches
Besides the most used microextraction techniques described above, other extraction approaches using nanomaterials for the determination of cosmetic-related compounds have been published. These methods are summarized in
Table 6. Makkliang et al. [
97] proposed rotative SPME using a multistir-rod microextractor based on MWCNT functionalized with carboxyl groups, which was applied for parabens determination in cosmetic samples previously dissolved in MeOH. Alcudia-León et al. [
98] also determined parabens in pool and sea waters by magnetically confined hydrophobic nanoparticles microextraction. In this method, a magnetic device made of a magnet, a PTFE septum and a magnetic nanocomposite (Fe
3O
4@C18) is employed for the microextraction. Wang et al. [
99] determined five parabens and TCS in biological samples, using a magnetic μSPE chip connected directly with the chromatographic system. Fresco-Cala and Cárdenas [
100] synthetized a composite based on carbon nanohorns inside pipette tips for the determination of a group of four parabens in urine. Wang et al. [
101] developed a device with GO packed into polyamide organic membrane for the analysis of different parabens in water, and finally, Montesdeoca-Esponda et al. [
102] analysed benzotriazole UV stabilizers in sewage water by fabric phase sorptive extraction (FPSE), with a PDMS nanocomposite bonded on a polystyrene support as extraction device.
As summary, we would like to emphasize that, regarding to the nanomaterials-based microextraction techniques used for the determination of cosmetic-related compounds, those based on the dispersion of the sorbent (i.e., DSPE and SBSDME) represent more than half of the published articles, as it is shown in
Figure 2a. As commented before, the reduction of the extraction time, most probably, is the reason behind this trend.
With regard to the target analytes, authors paid attention during many years to the determination of parabens and UV filters, which gather about half of the published articles, as it is shown in
Figure 2b.
Finally, with regard to the use of nonmagnetic or magnetic materials,
Figure 2c shows that they are practically on par, but the observed trend is an increase in the use of magnetic materials in the last years.