Redefining dilute and shoot: The evolution of the technique and its application in the analysis of foods and biological matrices by liquid chromatography mass spectrometry

Redefining dilute and shoot: The evolution of the technique and its application in the analysis of foods and biological matrices by liquid chromatography mass spectrometry Brett Greer a , * , Olivier Chevallier a , Brian Quinn a , Luis M. Botana b , Christopher T. Elliott a a Institute for Global Food Security, School of Biological Sciences, Queens University Belfast, UK b Departamento de Farmacología, Facultad de Veterinaria, Universidade de Santiago de Compostela, 27002, Lugo, Spain article info Article history: Available online 5 April 2021 Keywords: Dilute-and-shoot Liquid chromatography Mass spectrometry Sample preparation Food analysis Bioanalysis Mycotoxin Multi-class abstract With laboratories seeking to expand analytical capabilities and create multi-class, multi-analyte methods, there has been a shift toward generic sample clean-up techniques such as “dilute-and-shoot”. Advantages of this methodology include its simplicity, minimal analyte losses, high sample throughput and number of analyte classes included. The evolution of dilute-and-shoot has permitted its use across a variety of matrices including food and biological and in various scientificfields such as food, forensics and environmental. The versatility of the technique permits the expansion of current fields of research without the usual laborious method development. There can be issues with matrix effects and robust quantitation as analyte number increases. This review provides an overview of the technique combined with liquid chromatography mass spectrometry, highlighting its power in facilitating multi-class analysis. Coupled with increases in instrument performance there is potential to employ this methodology in expanding analytical capabilities in many areas of life science research. ©2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction In analytical chemistry and in particular in the field of small molecule analysis (approx. <1200 Da) using liquid chromatography mass spectrometry (LC-MS), the use of “dilute-and-shoot”has gained much traction over the last ten to fifteen years, with it now employed in many types of multi-residue LC-MS analysis [1]. The term itself appears quite self-explanatory, referring simply to: “Dilution of a sample or sample extract before ‘shooting’into an analytical instrument for analysis.” As a stand-alone methodology, the term implies that this technique is the only clean-up step used before sample analysis. However, dilute-and-shoot (DnS) is sometimes performed after an initial deproteinisation step dependent on the matrix being assayed, with it stated that the protein precipitation step is considered a clean-up process [2]. If this is the case and deproteinisation is carried out before dilution of the sample (supernatant), if could be argued that such methodology cannot be considered DnS. Furthermore, matrices that are not liquid must first undergo an extraction step to facilitate movement of analytes from the matrix to a liquid phase before dilution and analysis on some form of mass analyser. These include matrices such as animal feed, tissues and various (solid) foodstuffs. In these cases, a suitable extraction solvent must be chosen to facilitate migration of analytes from a solid to a liquid phase which is then diluted prior to analysis. One such example is in the recent study conducted by Sulyok et al. (2020) measuring over 500 secondary microbial metabolites in various food matrices using DnS coupled to LC-MS/MS [3]. Therein lies the issue as to what extent a sample clean-up method can be truly classified as DnS. With many publications claiming use of the ‘dilute-and-shoot’approach but which include additional steps such as solid-liquid extraction or deproteinisation, the use of or definition of the technique appears to have evolved. Therefore, due to the evolution of the technique in the literature over the last decade, we have expounded on the simple definition stated above to include solid matrices. “The direct dilution of a sample or dilution of a sample extract before ‘shooting’into an MS platform for analysis. This also allows for the addition of a solvent to a solid matrix to extract *Corresponding author. E-mail address: [email protected] (B. Greer). Contents lists available at ScienceDirect Trends in Analytical Chemistry journal homepage: www.elsevier.com/locate/trac https://doi.org/10.1016/j.trac.2021.116284 0165-9936/©2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Trends in Analytical Chemistry 141 (2021) 116284 untargeted or targeted compounds into a liquid phase before dilution and analysis. Filtration and centrifugation are included in our definition, as they are physical methods that remove solid materials from liquid samples, thereby making the sample more suitable for LC-MS analyses”. In using DnS as a sample clean-up technique, the main purpose of the dilution step is to reduce matrix components which can interfere with the chromatography and ionisation of the target analytes, thereby facilitating greater sensitivity or performance of the analytical method. With that in mind, the dilution factor used is dictated by many factors including; concentration of the analytes expected, the matrix analysed, the selectivity and sensitivity of the MS platform utilised and whether the analysis is quantitative or qualitative, with dilution factors typically ranging from 1:1 up to 1:100. Some examples include a dilution factor of 1:1 (v:v) in the analysis of secondary microbial metabolites in food matrices; a dilution factor of 1:10 analysing metabolites of the mycotoxins deoxynivalenol (DON) and zearalenone (ZEN) in human urine; a dilution factor of 1:50 in the analysis of drugs-of-abuse (DoA) and performance enhancing drugs in human urine; a 1:100 dilution of samples in the analysis of prosthesis-related metals in whole blood [3e6]. One caveat of the last two examples is that the analysis of DoA by Alc antara-Dur an et al. (2018) employed nanoflow LC coupled to an Orbitrap MS which improved method sensitivity and permitted a high dilution factor, whereas the study by Bolea- Fernandez et al. (2016) was of prosthesis-related metals, i.e. elements as opposed to intact molecules and which was carried out on Inductively Coupled Plasma MS (ICP-MS). In the main and for the purpose of this review, the focus will be on the analysis of small molecules by way of LC-MS. Over recent years there has been a shift from single class methods with low analyte numbers toward multi-class, multianalyte methods. This has been achieved by moving away from use of traditional sample clean-up techniques such as solid phase extraction (SPE) and Immunoaffinity Columns (IAC) to more generic techniques such as QuEChERs (Quick, Easy, Cheap, Effective, Rugged and Safe), with the latter methodology routinely applied in multi-residue pesticide analysis in agricultural commodities [7]. However, even techniques such as QuEChERs still suffer from losses or incomplete recovery of the target analytes when performing multi-class analysis. These analyses typically utilise chromatographic separation to aid detection and can be performed on high performance liquid chromatography (HPLC). Currently however, chromatography is routinely performed on ultra-high performance liquid chromatography (UHPLC or UPLC) coupled to some form of mass analyser. The switch to UHPLC facilitated shorter run times while allowing the same number of analytes to be resolved. Of late, method run times have become longer to facilitate the increase in the number of analytes included in a single run [3,8]. In regards to the mass analyser used, these include low resolution mass spectrometers (LRMS), mainly triple quadrupole instruments (QqQ or MS/MS), and high-resolution instruments such as ion trap or orbital ion traps (Orbitrap), time-of-flight (TOF) and the new generation of hybrid systems such as quadrupole-ToF (QToF) and quadrupole-Orbitrap (Q- Orbitrap), with these HRMS systems offering different resolving powers and therefore selectivities and sensitivities. LRMS instruments are used mainly in routine testing laboratories for high throughput quantitative analysis, whereas HRMS systems are more typically employed for research purposes. Thus, the instrumentation used will be dependent on the type of analysis required, whether it be fully quantitative, qualitative for screening purposes or simply untargeted for “omics”analysis. The purpose of this review is to highlight the evolution in the use of dilute-and-shoot in analytical chemistry which has now been employed across many different matrices, whilst also indicating that the definition of the term should be extended to reflect this. In addition, we look at how its use is expanding particularly in the area of targeted analysis, mainly due to the advancement of the detection platforms which has allowed it to transcend many different scientific disciplines. 2. Dilute-and-Shoot 2.1. An overview of the technique DnS refers to the dilution of a sample matrix with a suitable dilution solvent (diluent) for analysis. Therefore, depending on the matrix analysed it can either be a straightforward dilution of a biological fluid such as urine or oral fluid which conform to the ‘original’definition of the technique. However, other more complex fluids such as serum, plasma or milk can require an initial protein precipitation step, or dilution of the extraction solvent used in extracting analytes of interest from a solid matrix such as foodstuffs, with the latter in particular included in the extended, updated definition of the technique outlined above. Compared with other traditional sample clean-up techniques which remove unwanted co-extracted matrix components, DnS simply decreases matrix effects while eliminating time-consuming and expensive extraction procedures and can therefore be considered the most straightforward and fastest clean-up method available. One caveat of this is in the application to solid foodstuffs whereby an extraction step is required to facilitate migration of analytes from a solid to a liquid phase, with examples detailed in Table 1. Use of DnS in analytical chemistry has become quite commonplace due to its simplicity, limited and fast sample preparation, low solvent usage and associated low-costs, as well as reduced waste production, with the benefit of being more environmentally friendly than most other sample clean-up techniques. As well as being a relatively environmentally friendly technique, DnS lends itself well to high throughput routine analysis which has led to its use across various disciplines such as in drugs-of-abuse (DoA) and anti-doping (performance enhancing drugs), forensics and food safety, and can be used across a range of “omics”studies such as metabolomics. It also lends itself to multi-class, multi-analyte approaches as there are little to no analyte losses and its simplicity minimises the chance of errors during the extraction stage. However, there can be issues with the repeatability, reproducibility, accuracy and precision of this technique, especially in relation to the validation of multi-residue methods in foodstuffs using DnS. These aforementioned issues have been addressed somewhat in the studies carried out by Malachova et al. (2015) and Sulyok et al. (2020) [3,9]. Examples of where the DnS methodology has been employed alongside either LC-MS/MS or LC-HRMS are given in Table 1, with examples as to why DnS is such a powerful analytical tool is discussed in more detail later. 2.2. Issues with dilute-and-shoot: terminology and methodology In terms of the expounded definition of DnS, whereby an extraction solvent is required to facilitate the migration of analytes from a solid matrix into a liquid phase before dilution for analysis, one must be aware that the choice of extraction solvent directly impacts the range of analytes that can be extracted, i.e. the polarity of the extraction solvent used will determine the polarity and therefore range of analytes extracted. It is clear that certain matrices require an extraction step before dilution and an example of this is addressed in the study by Sulyok et al. (2020). The B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 2 Table 1 Application of dilute-and-shoot alongside LC-MS/MS across various scientific disciplines and sample matrices. Compound type (number of analytes) Matrices Methodology LC-MS Platform Analytical conditions Reference Mycotoxins (39) Wheat &maize 0.5 g matrix extracted with 2 mL of extraction solvent (ACN:H 2 O:CH 3 COOH, 79:20:1 v/v/v), 350 m L supernatant diluted with 350 m Lof (ACN:H 2 O:CH 3 COOH, 20:79:1 v/v/v) Dilution factor 1:2 QTRAP®4000 equipped with a Turbo Ion Spray ESI source (AB SCIEX) coupled to a 1100 series HPLC (Agilent) ESI þ &ESI  : MRM Gemini C 18 column (150 mm 4.6 mm, 5 m m) þC 18 security guard cartridge (4 mm 3mm,5 m m) @ 25  C, flow rate 1 mL/ min, injection volume 5 m L. MPA: MeOH:H 2 O:CH 3 COOH (10:89:1 v/v/v), MPB: MeOH:H 2 O:CH 3 COOH (97:2:1 v/v/v), both containing 5mMNH 4 CH 3 COO  . [27] Bacterial &fungal metabolites (295) Apple puree, hazelnut, maize &green pepper 0.5 g matrix extracted with 2 mL of extraction solvent (ACN:H 2 O:CH 3 COOH, 79:20:1 v/v/v). 350 m L supernatant diluted with 350 m Lof (ACN:H 2 O:CH 3 COOH, 20:79:1 v/v/v) Dilution factor 1:2 QTRAP®5500 equipped with a TurboV ESI source (AB SCIEX) coupled to a 1290 series UHPLC (Agilent) ESI þ &ESI  : sSRM Gemini C 18 column (150 mm 4.6 mm, 5 m m) þGemini C 18 guard column (4 mm 3mm,5 m m) @ 25  C, flow rate 1 mL/ min, injection volume 5 m L. MPA: MeOH:H 2 O:CH 3 COOH (10:89:1 v/v/v), MPB: MeOH:H 2 O:CH 3 COOH (97:2:1 v/v/v), both containing 5mMNH 4 CH 3 COO  [28] Mycotoxins &metabolites (>500) Wheat, maize, figs, dried grapes, walnuts, pistachios &almonds. 0.25 g matrix extracted with 1 mL of extraction solvent (ACN:H 2 O:CH 3 COOH, 79:20:1 v/v/v). 300 m L supernatant diluted with 300 m Lof (ACN:H 2 O:CH 3 COOH, 20:79:1 v/v/v) Dilution factor 1:2 QTRAP®5500 equipped with a TurboV ESI source (AB SCIEX) coupled to a 1290 series UHPLC (Agilent) ESI þ &ESI  : sMRM NB: Two separate runs for ESI þ &ESI  Gemini C 18 column (150 mm 4.6 mm, 5 m m) þC 18 security guard cartridge (4 3mm i.d.) @ 25  C, flow rate 1 mL/min, injection volume 5 m L. MPA: MeOH:H 2 O:CH 3 COOH (10:89:1 v/v/v) MPB: MeOH:H 2 O:CH 3 COOH (97:2:1 v/v/v), both containing 5 mM NH 4 CH 3 COO  . [3] Mycotoxins (10) and pesticides (6) Rice 5 g of rice was extracted and diluted with 10 mL ACN:H 2 O (8:2 v/v) containing 2% CH 3 COOH (v/ v) then filtered. Dilution factor 1:3 QTRAP®6500 (AB SCIEX) coupled to a 1290 series UHPLC (Agilent) ESI þ &ESI  : MRM Zorbax RRHD C 18 column (50 mm 2.1 mm, 3 m m) with pre-column @ 30  C, flow rate 350 m L/min, injection volume 5 m L. MPA:H 2 O, MPB: MeOH both containing 0.5% CH 3 COOH (v/v) [29] Pesticides (185) and mycotoxins (12) Wine (red) A 500 m L aliquot of wine was diluted with 500 m L of MeOH:ACN (1:1, v/v) and filtered. Dilution factor 1:2 Xevo®TQ-S (Waters) coupled to an ACQUITY H- Class UPLC®with quaternary solvent manager (Waters) ESI þ : MRM ACQUITY UPLC®BEH C 18 column (100 mm 2.1 mm, 1.7 m m) @ 60  C, flow rate 450 m L/min. Pesticide analysis: MPA: 5 mM NH 4 HCOO  (aq) ,MPB: MeOH Mycotoxin analysis: MPA:H 2 O and MPB: ACN, both containing 0.1% HCOOH (v/v) [17] Prescription drugs, DoA and metabolites (37) Urine (human) 100 m L of centrifuged urine diluted with 700 m LH 2 O and 200 m LIS mix . Dilution factor 1:10 TQ-S Micro (Waters) coupled to an ACQUITY UPLC (Waters) ESI þ : MRM CORTECS C 18 column (2.1 mm 50 mm, 1.6 m m) @30  C. NB: Injection volume not stated. MPA: Water, MPB: MeOH (both containing 0.1% HCOOH v/v). [16] Mycotoxins and metabolites (37) Urine (human) 100 m L of centrifuged urine diluted with 900 m L of diluent (H 2 O:ACN:HCOOH, 94:5.:1 v/v/v) Dilution factor 1:10 QTRAP®6500 (AB SCIEX) coupled to a 1260 Infinity series HPLC (Agilent) ESI þ &ESI  : sMRM NUCLEODUR®C 18 Pyramid column (150 mm 2 mm, 3 m m) with a C 18 guard column (2 mm 4mm),flow rate 600 m L/min, injection volume 10 m L. MPA: ACN, MPB: Water (both containing 0.1% HCOOH v/v). [70] Statins (8) Urine (human) 500 m L of urine spiked with 60 m LIS mix and made up to 1.5 mL total volume with H 2 O. Dilution factor 1:3 API 3200 Q-Trap equipped with a Turbo V ESI source (AB SCIEX) coupled to an Agilent 1200 series. ESI þ : MRM Kinetex C 18 column (50 mm 3.0 mm, 2.6 m m) with a C 18 guard column (4 mm 3 mm) @ 20  C, flow rate 350 m L/min, injection volume 10 m L. MPA: Water (0.1% CH3COOH), MPB: ACN. [92] (continued on next page) B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 3 Table 1 (continued) Compound type (number of analytes) Matrices Methodology LC-MS Platform Analytical conditions Reference Neurotransmitters and metabolites (10) Urine (human) Dilute 1 mL of urine with 1 mL of ACN. Mix, centrifuge then dilute 1 mL supernatant with 0.25 mL IS mix and 1.25 mL 0.2 M acetic acid. Dilution factor 1:5 6490 TQ MS (Agilent) coupled to a 1290 infinity series LC system (Agilent) ESI þ : MRM Agilent Poroshell 120 Bonus-RP LC column (2.1 mm 100 mm, 2.7 m m), flow rate 180 m L/ min. NB: Injection volume not stated. MPA: Water (0.2% HCOOH, v/v). MPB: ACN (0.1% HCOOH, v/v). [31] (i) Mycotoxin DON and its glucuronide metabolites (ii) Mycotoxins and metabolites (15) (iii) Mycotoxins DON, ZEN and metabolites NB: All of the above studies use the same extraction and analysis. Urine (human) Dilute 100 m L of centrifuged urine with 900 m L dilution solvent (H 2 O:ACN, 90:10 v/v). Dilution factor 1:10 QTRAP®5500 equipped with a Turbo V ESI source (AB SCIEX) coupled to a 1290 series UHPLC (Agilent) ESI þ &ESI  : SRM Atlantis®T3 column (150 mm 3mm,3 m m) with a C 18 security guard cartridge @ 35  C, flow rate 600 m L/min, injection volume 5 m L. MPA: Water, MPB: ACN (both containing 0.1% CH 3 COOH) [4,10,11] ZEN and metabolites (9*) *known metabolites Urine (porcine) Urine samples diluted to 0.2 mM creatinine with H 2 O then centrifuged. Dilution factor unknown QTRAP®6500 equipped with an Ion Drive Turbo V ESI source (AB SCIEX) coupled to a 1290 series UHPLC (Agilent) ESI  : SRM Kinetex Biphenyl column (150 mm 3 mm, 2.6 m m) @ 30  C, flow rate 400 m L/min, injection volume 5 m L. MPA: Water, MPB: ACN (both containing 0.1% CH 3 COOH) [72] Novel Psychoactive Substances (826) Urine (human) 100 m L of urine spiked with IS mix and diluted with 400 m LH 2 O. Dilution factor 1:5 Enzymatic hydrolysis: 100 m L of urine mixed with 100 m L glucuronidase solution. Dilution factor 1:2 Agilent 6460 QqQ MS with Jet Streaming Technology coupled to a 1290 Infinity series LC (Agilent) ESI þ : dMRM Zorbax RR HD Eclipse Plus C 18 column (100 mm 3 mm, 1.8 m m) with pre-column @ 40  C, flow rate 300 m L/min, injection volume 5 m L. MPA:5mMNH 4 HCOO  (aq) (0.1% HCOOH, v/v), MPB: MeOH (0.1% HCOOH, v/v). [30] Chemical warfare agent metabolites (6) Urine (human) 500 m L of urine mixed with 500 m LofH 2 O. Dilution factor 1:2 QTRAP®4000 (AB SCIEX) coupled to an Ultimate 3000 LC (Dionex) ESI þ &ESI  : MRM Acclaim 120C 18 column (150 mm 2.1 mm, 2.2 m m) @ 30  C, flow rate 450 m L/min, injection volume 20 m L. MPA: Water, MPB: ACN (both containing 0.5% HCOOH) [83] Cannabinoids and acid precursors (8) Oral fluid, serum and urine (human) 100 m L of sample mixed with 100 m LM3® reagent, 200 m LIS mix and 200 m L acetone:ACN (8:2 v/v) Dilution factor 1:6 Xevo®TQ-S (Waters) coupled to an ACQUITY UPLC®I-class (Waters) ESI þ : MRM ACQUITY UPLC®BEH C 18 column (50 mm 2.1 mm, 1.7 m m) @ 50  C, flow rate 400 m L/min, injection volume 10 m L. MPA: Water containing 50 mM NH 4 HCOO  pH3, MPB: MeOH [73] Psychoactive drugs (13) Oral fluid (human) 500 m L of oral fluid was mixed with 10 m LIS mix and diluted further with 1 mL H 2 O. Dilution factor 1:3 Xevo®TQ-S (Waters) coupled to an ACQUITY UPLC®I-class (Waters) ESI þ : MRM ACQUITY UPLC®BEH C 18 column (75 mm 2.1 mm, 1.7 m m) @ 50  C, flow rate of 350 m L/min, injection volume 2 m L. MPA: Water, MPB: ACN (both containing 0.1% HCOOH) [15] Broad spectrum drugs of abuse (61) Urine (human) 200 m L aliquot of urine diluted with 400 m LH 2 O, 300 m L b -Glucuronidase solution and 100 m L IS mix . Dilution factor 1:2 Xevo®G2 ToF (Waters) coupled to an ACQUITY UPLC (Waters) ESI þ Resolution: 20 000 FWHM @ m/z 400 Mass tolerance: ±5 and 20 ppm ACQUITY UPLC®BEH C 18 column (150 mm 2.1 mm, 1.7 m m) @ 50  C, flow rate 400 m L/min, injection volume 20 m L. MPA:5mMNH 4 HCOO  (aq) (pH 3), MPB: ACN (0.1% HCOOH, v/v) [93] B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 4 Anabolic steroids and metabolites (32) Urine (human) 200 m L aliquot of urine diluted with 200 m LIS mix in MeOH. Dilution factor 1:2 Exactive Orbitrap® (Thermo Scientific) coupled to an Accela 1250 LC (Thermo Scientific) ESI þ &ESI  Resolution: 50 000 FWHM @ m/z 100 to 2000 Mass tolerance: ±5 ppm Varian Omnispher™C 18 column (100 mm 2 mm, 3 m m) þChromSep guard column (10 mm 2 mm, 5 m m) @ 35  C, flow rate 250 m L/min, injection volume 25 m L. MPA:H 2 O and MPB: MeOH, both containing 1mMNH 4 CH 3 COO  and 0.1% CH 3 COOH v/v. [94] Sports drugs and metabolites (27) Urine (human) 270 m L aliquot of urine diluted with 100 m Lof 100 mM NH 4 HCOO  , 700 m L ACN and 30 m L IS mix . Dilution factor 1:11 Q Exactive Hybrid Orbitrap®coupled to an Accela LC (Thermo Scientific) and a 1100 series LC (Agilent) ESI þ &ESI  Resolution: 17 500 FWHM @ m/z 100e500 Da Mass tolerance: ±5 ppm Nucleodur HILIC column (100 mm 2mm, 1.8 m m) þNucleodur HILIC trapping column (20 mm 2 mm, 3 m m), flow rate 250 m L/min, injection volume 20 m L. MPA:H 2 O, MPB: ACN and MPC: 200 mM NH 4 CH 3 COO  (aq) (0.15% glacial CH 3 COOH, v/v) [95] Sport drugs (46) Urine and plasma (equine) Plasma: 75 m L aliquot of deproteinated plasma diluted with 75 m L MPA. Dilution factor 1:3 Urine: 20 m L aliquot spiked with 20 m LIS mix and diluted with 960 m L MPA. Dilution factor 1:50 Q-Exactive™Orbitrap (Thermo Scientific) coupled to an ACQUITY UPLC® (Waters) ESI þ Resolution: 140 00 FWHM @ m/z 200 Mass tolerance: ±5 ppm ACQUITY UPLC®BEH C 18 column (100 mm 2.1 mm, 1.7 m m) @ 55  C, flow rate 350 m L/min, injection volume 5 m L. MPA: 5 mM NH 4 HCOO  (aq) (pH 3), MPB: ACN (0.1% HCOOH, v/v). [96] Pesticide and veterinary drugs (>350) Meat (beef, pork &chicken) 2.5 mL of H 2 O added to 2.5 g of homogenised meat then diluted with 7.5 mL ACN (1% HCOOH, v/v). Dilution factor 1:4 Exactive™Orbitrap (Thermo Scientific) coupled to a Transcend 600 UHPLC (Thermo Scientific) ESI þ &ESI  Resolution: 25 000 FWHM @ m/z 70 to 1000 Mass tolerance: ±5 ppm Hypersil GOLD aQ C 18 column (100 mm 2.1 mm, 1.7 m m) @ 30  C, flow rate 300 m L/min, injection volume 10 m L. MPA: 4 mM NH 4 HCOO  (aq) (0.1% HCOOH, v/v), MPB: MeOH (0.1% HCOOH, v/v). [80] Pesticides, mycotoxins, process-induced toxicants &packaging Contaminants (32) Tea (brew and leaves) Tea brew: 2 mL diluted with 8 mL H 2 O:ACN (3:1, v/v) acidified with 0.1% HCOOH (v/v) then 0.5 mL of the mixture filtered (0.2 m m). Dilution factor 1:5 Tea leaves underwent a “dilute, shoot &evaporate approach”. Xevo®G2-S ToF (Waters) coupled to an ACQUITY H- Class UPLC®with quaternary solvent manager (Waters) ESI þ &ESI  Resolution: 30 000 FWHM @ m/z 200 Mass tolerance: not given C 18 PFP column (150 mm 2.1 mm, 2 m m) @ 30  C, flow rate 400 m L/min, injection volume 10 m L. MPA:H 2 O, MPB: ACN both containing 0.1% HCOOH (v/v) and MPC: MeOH [97] DoA and sport drugs (81) Urine (human) Aliquot of urine diluted with H 2 O:ACN (95:5 v/ v) and filtered. Dilution factor 1:50 Q-Exactive™Orbitrap (Thermo Scientific) coupled to an EASY-nLC 1000 nano- LC system (Thermo Scientific) ESI þ &ESI  Resolution: 70 000 FWHM @ m/z 195 EASY-Spray PepMap®C 18 column (150 mm 75 m m, 3 m m) @ 25  C, flow rate 200 nL/min, injection volume 1 m L. MPA:H 2 O, MPB: ACN both containing 0.1% HCOOH (v/v) NB: Nano-flow used as opposed to LC [5] ACN: Acetonitrile,CH 3 COOH: Acetic acid, DoA: Drugs of Abuse, DON: Deoxynivalenol, dMRM: Dynamic Multiple Reaction Monitoring, ESI: Electrospray Ionisation, ESI - : Electrospray Ionisation (negative mode), ESI þ : Electrospray Ionisation (positive mode), FWHM: Full Width at Half Maximum (resolution), HCOOH: Formic acid, HILIC: Hydrophilic Interaction Liquid Chromatography, HPLC: High Performance Liquid Chromatography, IS mix :Internal Standard mixture, LC: Liquid Chromatography, MeOH: Methanol, MPA: Mobile Phase A, MPB: Mobile Phase B, MPC: Mobile Phase C, MRM: Multiple Reaction Monitoring, MS: Mass Spectrometry, MS/MS: Triple Quadrupole Mass Spectrometry (QqQ), m/z: Mass to Charge Ratio, NH 4 CH 3 COO  : Ammonium acetate, NH 4 HCOO  : Ammonium formate, ppm: Parts Per Million, QTRAP®: Triple Quadrupole Linear Ion Trap, QuEChERS: Quick, Easy, Cheap, Effective, Rugged, and Safe, sMRM: Scheduled Multiple Reaction Monitoring, SPE: Solid Phase Extraction, SRM: Selected Reaction Monitoring, UHPLC: Ultra-High Performance Liquid Chromatography, UPLC®: Ultra-Performance Liquid Chromatography, ZEN: Zearalenone. Dilution factor expressed as sample:total (volumes), i.e. 500 m L sample mixed with 500 m L diluent equates to a dilution factor of 1:2. B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 5 workflow employed was highly fit for purpose and being classed as DnS is a good example of the ambiguity of the classic DnS definition. In this study, the authors looked at several food matrices with varying degrees of complexity, from wheat and maize to several classes of nuts. A 250 mg aliquot was extracted with 1 mL of extraction solvent and after mixing and centrifugation, a 300 m L aliquot of supernatant was diluted 1:1 (v/v) with dilution solvent giving a dilution factor of 1:2, after which a portion was injected for analysis. Compare this to the DnS protocol employed by Warth et al. (2012) whereby a 100 m L aliquot of sample matrix, in this case human urine, was simply diluted with 900 m L of dilution solvent after centrifugation, giving a dilution factor of 1:10 before analysis [10,11]. Although both methods employ sample dilution, the simplicity of the latter leans more toward the type of application that fits with the DnS concept, i.e. no other extraction or sample clean-up technique was used, with the key differences outlined in the workflows shown in Fig. 1. There are also publications that mention use of DnS as their clean-up approach, however on examination of the methodology employed, some have used an additional clean-up step to reduce matrix interferences further including QuEChERS or SPE and are therefore not DnS [12e14]. A potentially more rigorous definition of DnS is that it is the only technique used prior to analysis by LC-MS, thus the technique only really applies to liquid matrices such as urine, oral fluids and liquid foodstuffs such as wine which can simply be diluted before analysis [4,5,11,15e17]. However, as outlined above, the definition must be extended to include dilution of a sample extract such as that performed in the study by Sulyok et al. (2020). Furthermore, other biological matrices such as blood (whole blood, serum or plasma) must first undergo a step to facilitate removal of proteins due to the fact they can block the column, injector and/or components of the LC system. This protein precipitation step is typically performed by addition of an organic solvent such as acetonitrile, methanol, ethanol or acetone which precipitates the protein content as well as partitioning the analytes of interest from the matrix into the organic solvent layer. However, this protein precipitation step can be considered an additional clean-up process. After deproteinisation and centrifugation, an aliquot of supernatant is diluted for analysis, or no further dilution is performed due to the solvent volume added at the protein precipitation step being sufficient. Therefore, in this case we have to extend the definition of DnS to include this additional step. Examples of this methodology include the analysis of pesticides and carotenoids in human serum using HPLC-DAD and HPLC-MS/MS respectively [18,19]. However, use of DnS can be problematic for detection of blood metabolites found in low concentrations, due to the dilution caused by addition of solvent required for deproteinisation. This can be remedied by drying the supernatant under a gentle stream of nitrogen before reconstitution in an appropriate volume to facilitate detection of lower concentration metabolites. This technique has been referred to as a ‘dilute, evaporate and shoot’approach [20]. One such example of this methodology was employed in the study by Devreese et al. (2012) in the analysis of pig plasma for mycotoxins and metabolites and which is routinely employed in this type of analysis [20e23]. One note of caution in this methodology is that the drying step can lead to some loss of analytes, particularly in complex matrices such as blood. Therefore, this methodology does not fall under the expounded DnS definition, and as such should be termed ‘dilute, evaporate and shoot’. 2.3. Instrumentation: low resolution (LRMS/QqQ) versus high resolution mass spectrometry (HRMS) The lack of selectivity associated with DnS is due to its generic sample clean-up and this can be offset by the selectivity afforded by triple quadrupole (QqQ) mass spectrometers that generally afford Fig. 1. Example of the best practice workflows across different matrices such as foodstuffs (solid and liquid) and biological for both quantitative and qualitative (screening) analysis using LC-MS (MS/MS or HRMS). 1 Dilute and shoot approach; 2 Dilute, evaporate and shoot approach; *Store in darkness overnight to allow solvent evaporation and analyte interaction with matrix; *Representative solvent; IS mix : Internal standard mix. B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 6 higher sensitivity than HRMS systems. There are many MS instruments on the market compatible with HPLC/UHPLC and which have been comprehensively reviewed in relation to their suitability for the intended analytical method [24e26]. As shown in Table 1 whereby DnS was employed alongside LC-MS/MS, the majority of analysts used an AB SCIEX QTRAP [27e29]. With DnS diluting rather than removing matrix, as well as diluting the analytes of interest therein, these systems appear to reduce the influence of coeluting matrix compounds and offer better sensitivity due to the enhanced selectivity afforded by the QTRAP technology and source design. This is not to state that these are the only QqQ instruments capable of being used alongside DnS, with other QqQ instruments also used successfully alongside DnS such as the Waters Xevo TQXS and Agilent 6460 and 6490 Triple Quadrupole LC-MS systems [15,17,30,31]. With improvements in QqQ instrumentation electronics facilitating faster cycle and dwell times, the number of analytes that can be included in a single run can be increased. One example of this is in the analysis conducted by Sulyok et al. (2020) in which more than 500 secondary microbial metabolites were analysed in a single run using a QTRAP 5500 (AB SCIEX), with chromatography performed on HPLC as opposed to UHPLC resulting in a run time of approximately 20 min. One caveat of this method and others with multiple analytes such as that of Steiner et al. (2020), is that with increasing analyte number and subsequent MRM transitions, cycle and dwell times have to be reduced in order to acquire enough data points per analyte for robust quantitation. Further to this, although polarity switching can be performed as rapidly as 5 msec according to some vendors, this along with the settling time would be detrimental to the number of data points acquired at any given time, and therefore, two injections may be required. Another multianalyte method was that conducted by Kimble and DeCaprio (2019), analysing more than 800 psychoactive substances using an Agilent 6460 MS/MS with Jet Streaming Technology coupled to UHPLC. As with the aforementioned methods, due to the dwell and cycle times afforded by the instrument, the actual number of analytes in a single run was approximately 400, with samples analysed twice to cover all 800-plus compounds. The analysis time for each run was 20 min and therefore 40 min overall [30]. Thus, due to the enhanced selectivity and stability afforded by QqQ platforms, these instruments are generally used for quantitative analysis. A new generation of hybrid HRMS systems has improved their capabilities, such as the quadrupole-ToF (QToF) and quadrupole- Orbital ion trap and which can now compete with QqQ instruments. These latest HRMS instruments have significantly better sensitivity specifications than previous HRMS instruments, achieved by advances in detection technology and increased resolving power, ultimately leading to a reduction in interferences with matrix compounds. The addition of the quadrupole (Q) to these instruments allows them to perform both qualitatively and quantitatively, with one such study using one of the new generation hybrid HRMS systems carried out by Martinez-Dominguez et al. (2016). In this multi-class, multi-analyte method, they analysed 257 pesticides and mycotoxins using a Q-Exactive Orbitrap both qualitatively and quantitatively. Although their methodology included an extra clean-up step after performing DnS, by way of florisil SPE cartridges, they reported similar results for DnS alone and only added SPE to maintain instrument performance [13]. One drawback in the use of QToF analysers is that polarity switching is not possible or practical and although this can be achieved with an Orbitrap, it is offset by a reduction in the instruments resolving power. However, with use of UHPLC as opposed to HPLC providing faster run times, this allows injection of each sample in positive and negative ion modes with the benefit that ionisation can be optimised for both, therefore increasing the number of detected compounds [32]. Overall, the number of methods employing DnS alongside LC-MS/MS compared to LCHRMS over the last 10 years is higher as indicated in the number of examples shown in Table 1. One thing to note is that not all articles using the DnS approach mention the methodology in the title, abstract or keywords. Therefore, the number of publications employing the technique are higher than indicated when a literature search is performed with the keywords ‘DnS’combined with either ‘tandem MS’,‘high resolution MS’or ‘Orbitrap’. Use of QqQ over HRMS is in part due to the enhanced selectivity afforded by QqQ instruments as well as their sensitivity which compensate for the generic sample clean-up technique. With the matrix being reduced rather than removed, the selectivity of QqQ instruments allows analytes to be identified based on the differentiation between analyte and unwanted matrix, although this is also due to the sensitivity as selectivity alone does not afford adequate detection limits. However, with many of the technical limitations around HRMS instruments having been resolved with advancements in technology, as well as the introduction of the hybrid quadrupole HRMS systems, there may well be a rise in the number of analysts using HRMS for routine laboratory tests. 3. Matrix effects In order to extract analytes from a liquid food matrix or biological matrices such as urine or blood, liquid-liquid extraction (LLE) can be used. Alternatively, solid-liquid extraction is used on solid matrices such as feed and food. In either case, the process using a single or multiple component liquid extraction mixture (organic solvents, water, acids, bases, etc.) serves to migrate the analytes of interest from the matrix into the added liquid phase. The addition of water to the organic (extraction) solvent can help the organic solvent penetrate the food matrix in the case of solid foodstuffs, while also facilitating extraction of some of the more polar analytes. The addition of acid serves to help break bonds between the analytes and some of the matrix components such as proteins and sugars, aiding their extraction [33]. When an extraction technique is chosen, there will always be co-extraction of matrix components giving rise to matrix effects, with the amount dependent on the extraction solvent and complexity of the sample. Some of the most challenging issues in analytical chemistry are matrix effects caused by co-extracted matrix compounds. These are unpredictable and can cause various issues such as: ion suppression (or in some cases enhancement), poor chromatography, false positives due to the presence of isobaric compounds, poor analyte recovery due to chemisorption and interferences with the ion ratio of the product ions in targeted analysis [34,35]. Use of internal standards is one of the most commonly used ways to compensate for matrix effects, however, as various analytes can exhibit different behaviours in the same matrix, ideally an internal standard is required for each analyte. This is not feasible due to the fact they are expensive and also that commercial availability is limited. Standard addition is another way to compensate for matrix effects and is often used in routine analysis. However, it is more suited to singleanalyte methods rather than multiple-analyte determination due to it being laborious and costly because of its high consumption of standards [1]. Some of these issues can be overcome by improved chromatography, however, this has limitations as there has to be a benefit of improved chromatography against the need for fast run times, especially in high-throughput laboratories. Even with the advancement in detection capabilities afforded by LC-MS, there may still be a need for sample clean-up to remove unwanted matrix compounds using more traditional techniques such as SPE, IACs, LLE or QuEChERS when method sensitivity is paramount [26]. Use of these techniques is really only applicable for B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 7 targeted methods where the efficacy of the clean-up technique can be assessed. This has limitations in regards to the number of analytes that can be incorporated and can get difficult as the number of analytes increase due to several factors, which include but are not limited to: Differing physiochemical properties, selectivity/specificity of the clean-up technique, instability of some analytes in the working standard solution, ionisation efficiency, availability of reference material and suitability of the clean-up platform chosen. An example of some of the issues encountered when employing DnS for a multi-class, multi-analyte method was highlighted in the study conducted by Malachova et al. (2014), with validation data acquired for 295 of the 331 target analytes due to some of the aforementioned issues. One of the main issues was that of signal suppression/enhancement (SSE) due to matrix effects and which was observed for all four matrices studied. Green pepper was considered the most complex matrix of the four, with only 10% of analytes not suffering from SSE [28]. In this example, samples were run in both positive (ESI þ ) and negative ionisation modes (ESI  ). A recent study by Steiner et al. (2020) investigated matrix effects of 80 fungal metabolites, 11 pesticides, and 9 pharmaceutical active agents in complex feed using DnS and LC-MS/MS. Although the majority suffered signal suppression, with polar analytes in ESI þ particularly susceptible to this, apolar analytes in ESI  were prone to signal enhancement, both caused by co-extracted matrix compounds [8]. Other matrices such as black tea (highly complex and dry) and rocket (high water and chlorophyll content) have shown to be somewhat problematic in pesticide analysis (39) using dilution of a sample extract [36]. Black tea and rocket indicated that only 4% and 7% of the analytes respectively were free of matrix effects at a 10- fold dilution, whereas the majority were completely free of matrix effects at a 100-fold dilution, indicating that a “dilute-and- shoot”approach may be successful. However, depending on the concentration of analyte(s) present, this dilution factor may prove too much for analyte detection. Therefore, a compromise could be sought, with a 25-fold dilution proving suffice if matrix effects were no greater than 20%. Another pesticide study in fruit and vegetables indicated that with no dilution, the matrices varied in the percent of analytes showing no suppression, defined as being <20%. Leek proved the most problematic, with only 9% of pesticides not supressed and 37% of analytes highly supressed (>50%). Overall, a dilution factor of 15-fold was enough to eliminate matrix effects, permitting quantification with solvent standards for the majority [37]. Therefore, matrix effects and analytical sensitivity will be a major factor in whether DnS can be employed, as well as the number and classes of analytes to be incorporated in the method. In some instances, the complexity of the sample matrix can lead to severe matrix effects and depending on the sensitivity required, DnS may not be suitable and other sample clean-up strategies may need to be employed, such as those listed in section 4. Taking into account complex matrices such as black tea, rocket, leek and green pepper, as well as other matrices studied such as urine and foodstuffs, all of the aforementioned can lead to signal suppression of the analytes therein, caused by co-elution of unwanted matrix components and the resulting contamination of the MS source and ion optics. Thus, although use of DnS can give excellent results with little to no sample preparation, MS detectors may well require regular venting and cleaning of the source components after multiple injections to maintain instrument sensitivity leading to unwanted instrument downtime. Therefore, when using DnS, the downtime is dependent on the MS system used and source design, with ABSCIEX QqQ instruments appearing to be the instrument of choice for targeted analysis using DnS as indicated in Table 1. Of late, there has been improvements in source design with various vendors implementing different designs or technologies to improve sensitivity while also reducing the necessity for regular maintenance. Examples of this include differential mobility spectrometry (DMS) with SCIEX introducing the SelexION DMS technology and the Agilent Jet Stream Ion Source with the addition of VacShield. These technologies serve to enhance the systems robustness with the bonus of spending less time on maintenance and increasing sample throughput. 4. Sample clean-up techniques to reduce matrix effects 4.1. Solid phase extraction (SPE) SPE is a useful tool for reducing matrix effects across various complex matrices such as foodstuffs, animal feed, aqueous samples and other biological matrices, and usually affords low LOQs. However, a drawback in its suitability for sample clean-up is the number of analytes incorporated is usually limited due to the selectivity of the stationary phase, with several factors to consider when choosing this such as: varying physiochemical properties of the analytes and the loading, washing and elution steps which can lead to losses or incomplete recovery of analytes. All of the aforementioned can lead to time consuming and laborious method development. Some examples in the use of SPE include the analysis of mycotoxins in pig serum, plasma and urine and in the analysis of freshwater toxins in aquaculture ponds and fish tissues [38e41], the analysis of mycotoxins in baby food, analysis of nerve agents in human urine and the analysis of veterinary drug and pesticide residues in animal feed [12,42,43]. Furthermore, two or more SPE columns can be used in tandem to analyse several compound classes, or an analyte (ZEN) and its metabolites across various matrices [44]. 4.2. Micro-extraction techniques: sorbent-based Similar to SPE, micro-extraction techniques based on solid phases or sorbents are now commonplace for use in sample preparation and clean-up. These techniques include solid-phase microextraction (SPME), stir-bar sorptive extraction (SBSE) and micro-extraction by packed sorbent (MEPS) [45]. SPME is based on partitioning of analytes between a coated fibre and sample, achieved by placing the SPME fibre above a solution (matrix) to capture the volatile vapour. However, SPME is more routinely employed alongside GC-MS analysing volatile organic compounds (VOCs) or volatile metabolites (VOMs). Use of direct immersion SPME (DI-SPME) has facilitated its compatibility with LC-MS in the analysis of non-volatiles [46]. One main drawback is its use of stationary phases coating the fibre [47], and along with DI-SPME showing poor recoveries in the analysis of polar molecules, the capacity of this technique in the creation of multi-class, multi-analyte methods for analysis by LC-MS is limited. Similar to SPME, SBSE is suited to analysis of medium to apolar molecules across various matrices using GC-MS. When combined with a sorptive phase, multi-residue methods have been developed for the analysis of endocrine disruptors (77) and pesticides (15) in water samples using GC-MS and LC-MS respectively [48]. Analysis of polar molecules by SBSE requires derivitisation which is laborious, although new materials could be developed in order to directly analyse polar molecules by LC-MS, making them comparable to SPE cartridges [48]. Once again however, the limitations for SBSE are similar as those mentioned for both SPME and SPE. MEPS has been developed as a miniaturised version of SPE and is fully compatible with LC-MS, with the added benefits of being fully automated, rapid and reduced solvent usage in comparison to SPE. Due to its size, MEPS are perfectly suited to small volumes of biological fluids such as plasma/serum, oral fluids and urine [45]. With B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 8 their ability to be utilised for small sample volumes and low concentration of analytes therein, they concentrate the analytes of interest whilst simultaneously removing unwanted matrix compounds, giving rise to more sensitive analytical methods. However, as with the other aforementioned techniques, the selectivity of the sorbent limits the number of analytes incorporated. 4.3. Immunoaffinity Columns (IACs) IACs work on the basis that the stationary phase contains an antibody to capture the analyte(s) of interest and can facilitate analysis at low levels, generally affording low LOQs. Although IACs are very selective, this has drawbacks such as lack of specificity for analytes of different classes or analogues, and they can exhibit some non-specific binding with other proteins. IACs are mainly used in the analysis of one or two compounds due to the cross reactivity profile of the antibody, such as in the analysis of chloramphenicol in food or vitamin B12 in green algae [49,50]. IACs can be utilised in the analysis of several compounds such as in the study of microcystins (15) and the structurally related toxin nodularin, with a comparison between IAC and SPE indicating a lower method detection limit using IAC [51]. They have also been developed with the capacity to bind several mycotoxins with varying physicochemical properties across a range of biological matrices, such as the 11þMyco MS-PREP IAC (R-Biopharm: https://food.r-biopharm. com/products/11myco-ms-prep/). IACs can facilitate quantification with solvent calibration curves, although this is not always the case as some feed matrices have been shown to cause SSE even with use of IACs [52]. In order to analyse a wider range of analytes, IACs with different specificities can be used in tandem [53], and they can also be used in conjunction with SPE, with the main driver in doing so to increase analyte number [54,55]. However, the selectivity of these techniques do not lend themselves to creation of multi-analyte methods. 4.4. QuEChERs QuEChERs was introduced in 2003 and has rapidly expanded from use in pesticide analysis to applications in food, pharmaceutical, environmental and biological analysis. Due to its success, two international standard organisations adopted versions of the original technique for quantitation of pesticides in both fruit and vegetables; the European Committee for Standardisation (CEN) and the Association of Official Analytical Chemists (AOAC) International [56]. The method combines liquid extraction and partitioning using salts, followed by a clean-up step using dispersive SPE (dSPE). As well its simplicity, it is relatively cheap in comparison to other techniques as well as being rapid. QuEChERs has been used extensively in pesticide residue analysis in fruit and vegetables, with 451 pesticide residues analysed along with a Q-Exactive Orbitrap, permitting both identification and quantitation [7]. Work has been conducted to evolve the method to areas other than pesticide analysis in agricultural commodities, with these discussed in two comprehensive reviews looking at its evolution, applications and trends [56,57]. Suffice to say, it is now used in the analysis of mycotoxins, antibiotics, veterinary drugs and herbicides across a variety of matrices. One such example is the analysis of 630 multi-class food contaminants using UHPLC-QToF including, veterinary drugs, food-packaging contaminants and mycotoxins, although the majority were pesticides (426) [58]. Although a very effective and evolving technique, it like other clean-up techniques has disadvantages due to incomplete or low recoveries of the suite of analytes due to the choice of dSPE chosen, as well as being mainly pesticide-centric. 4.5. Liquid-liquid extraction (LLE) Used as an extraction technique for some liquid matrices, LLE is considered to be a sample clean-up step. Apart from facilitating protein precipitation in some matrices, the partitioning separates compounds based on their solubility in two different immiscible liquids, such as water and an organic solvent, with analytes typically extracted in the organic layer leaving matrix components in the other, with some matrix co-extracted in the organic layer and vice versa. As mentioned previously, the choice of extraction solvent dictates the polarity of the analytes extracted and therefore presented for analysis. This is particularly important for untargeted qualitative analysis or metabolomic studies using LC-HRMS. In general, most of these extractions are conducted on biological matrices with examples given below. 4.5.1. LLE applied to blood samples (plasma or serum) This involves addition of an organic solvent for both deproteinisation and migration of analytes from matrix to the solvent, which may be acidified. This technique has limited advantages due to only partial removal of matrix components which are responsible for signal suppression [21]. Furthermore, in most instances instead of dilution, the sample is dried and reconstituted to concentrate the analytes of interest, with the caveat that it also concentrates any matrix present too. This technique, referred to as ‘dilute, evaporate and shoot’has benefits similar to DnS where the same extraction procedure can be employed for targeted and untargeted analysis due to minimal or no analyte losses. Examples include the studies carried out by Catteuw et al. (2019), Devreese et al. (2012) and Lauwers et al. (2019) on pig plasma [21,22,59] and others conducted in human plasma [60,61]. 4.5.2. LLE applied to urine samples Unlike blood, urine can simply be diluted for analysis as it does not usually require deproteinisation. In some instances, the addition of an organic solvent is needed to facilitate migration of the analytes from the polar, aqueous urine phase into the organic phase, whilst removing some matrix, before drying and reconstituting the organic layer. Examples of this include the analysis of mycotoxins where the urine was extracted with acetonitrile/water/ formic acid (52/45/3, v/v/v) or ethyl acetate in human and pig urine, respectively [23,62]. In both cases, use of both targeted and untargeted methods can be employed, however, some caution has to be observed for the latter extraction if analysing by HRMS for metabolites or potential biomarkers as some analytes had low recoveries in ethyl acetate at varying pH levels. 4.6. Dilute-and-shoot In comparison to the aforementioned techniques bar LLE, DnS is a generic analytical method with its advantages and disadvantages outlined in Table 2. The evolution of DnS has facilitated the creation of multi-analyte methods which were previously limited due to the differing physiochemical properties of the analytes in question and the sample clean-up technique chosen. One can note that with the evolution of DnS to include foodstuffs, the initial solid-liquid extraction can lead to SSE due to co-extraction of matrix, as well as low recoveries of some of the target analytes based on the solvent chosen and possible chemisorption. However, due to the absence of stationary phase and chemistry therein, the technique is considered not selective and permits the creation of multi-class, multi-analyte methods. Although in comparison techniques such as SPE, microextraction (SPME, SBSE and MEPS) and IAC can generally achieve lower detection limits, they are hampered by their respective selectivities and specificities, limiting the number B. Greer, O. Chevallier, B. Quinn et al. Trends in Analytical Chemistry 141 (2021) 116284 9