Journal of Chromatography B, 953–954 (2014) 115–119
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
An evaluation of washing and extraction techniques in the analysis of ethyl glucuronide and fatty acid ethyl esters from hair samples L.C.A.M. Bossers a,∗ , R. Paul a , A.J. Berry a , R. Kingston b , C. Middendorp c , A.J. Guwy d a
University of South Wales, ARW Building, Pontypridd CF374 AT, UK Lextox, The Maltings East Tyndall Street, Cardiff CF24 5EA, UK University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands d University of South Wales, 4 Forest Grove, Pontypridd CF37 1DL, UK b c
a r t i c l e
i n f o
Article history: Received 31 October 2013 Received in revised form 25 January 2014 Accepted 29 January 2014 Available online 13 February 2014 Keywords: Fatty acid ethyl esters Ethyl glucuronide Alcohol Sample preparation Washing extraction
a b s t r a c t Ethyl glucuronide (EtG) and fatty acid ethyl esters (FAEEs) are alcohol metabolites measured in hair and are after a decade of research thought to be the best markers in hair to indicate alcoholism and abstinence Forensic Sci. Int. 218 (2012) 2. A great body of work concerning EtG and FAEEs detection in hair has been performed. However, no recent extensive comparison has been made concerning washing and extraction procedures. This work shows that the washing procedure of dichloromethane followed by a methanol rinse of the hair sample removes more than 16% of the FAEEs and 50% of the total EtG that is present in and on the hair. A review of ten washing protocols (where the removal is categorised: high, medium or low) showed that a relatively high percentage of FAEEs was removed and “medium” amount of EtG compared to the other washing protocols. This work shows promising results for the extraction of the FAEEs and the combined extraction of FAEEs and EtG by using 30 min of sonication with methanol. More FAEEs were recovered from hair with methanol than with any other extraction solvent including the commonly used dimethyl sulfoxide/heptane mixture. When the sonication time was increased a higher percentage of transesterification of the FAEEs was observed, the extraction was “dirtier” as solids and a colour change was observed whereas the extraction efficiency did not increase. Therefore, washing the hair sample with dichloromethane and methanol followed by an addition of 1 ml of methanol and sonication for 30 min to extract the FAEEs and EtG from hair is recommended for FAEEs as well as for the combined analysis of EtG and FAEEs. A linear calibration curve (r2 > 0.99) was obtained for all analytes. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Though alcohol consumption is not illegal in many parts of the world, establishing someone’s alcohol consumption can be useful
Abbreviations: EtG, ethyl glucuronide; FAEE, fatty acid ethyl esters; SoHT, Society of Hair Testing; BSTFA, NO-bis(trimethylsilyl)trifluoroacetamide; DMSO, dimethyl sulfoxide; DCM, dichloromethane; GC, gas chromatography; MS, mass spectrometry; HS SPME, head space solid phase micro extraction; PDMS/DVB, polydimethylsiloxane/divinylbenzene; SIM, selected ion monitoring; SPE, solid phase extraction; E1, an ester formed from ethanol and a saturated (0 double bonds) 14 carbon long acid; 14:0, an ester formed from ethanol and a saturated (0 double bonds) 14 carbon long acid; E16:0, an ester formed from ethanol and a saturated (0 double bonds) 16 carbon long acid; E18, 1an ester formed from ethanol and a unsaturated (1 double bonds) 18 carbon long acid; E18:0, an ester formed from ethanol and a saturated (0 double bonds) 18 carbon long acid. ∗ Corresponding author. Tel.: +41438178006; fax: +44 1443 482285. E-mail addresses: [email protected], [email protected], [email protected] (L.C.A.M. Bossers), [email protected] (R. Paul), [email protected] (A.J. Berry), [email protected] (R. Kingston), [email protected] (C. Middendorp), [email protected] (A.J. Guwy). 1570-0232/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2014.01.049
in the legal and medical field. In child custody cases, for instance, it is important to determine if allegations of chronic heavy drinking are true. Knowledge about someone’s drinking behaviour is required when diseases occur, like fetal alcohol spectrum disorder, that can only be diagnosed and treated when it is known that someone has been drinking. Alcohol hair testing can aid in the confirmation of someone’s drinking behaviour in such cases [1]. Currently, the consensus of the Society of Hair Testing (SoHT) [1,2] describes that concentrations of the minor metabolites EtG (ethyl glucuronide) and FAEEs (fatty acid ethyl esters) in hair are expected to be above a cut-off value (30 pg/mg and 0.5 ng/mg) for chronic heavy drinkers. Several washing procedures and solutions are currently used for hair prior to the EtG and FAEEs analysis. Most FAEEs methods include a non-polar solvent as was suggested by Politi et al. [3] to remove the greasy layer from hair, which may otherwise hinder the extraction. However, it is clear that no general consensus is reached especially not for EtG analysis. The washing protocol described for EtG analysis differ in (i) the polarity of the solvents that are used, (ii) the volume of the solvents, (iii) the amount of washes and (iv) the application of sonication. Kronstrand et al. [4],
116
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
for instance, used ultrasonication and 3 ml per wash whereas Albermann et al. [5] uses 1 ml and Morini et al. [6] did not use sonication. In our study solvents are used with a range of polarities and distinct hair swelling properties; for compounds that are protic are said to have the ability to swell hair and thereby facilitate the diffusion and thus extraction of the analytes from hair [7]. In this study the polarity and protic ability of the solvent washes were varied to investigate the influence of these characteristics on the wash and extraction efficiency of EtG and FAEEs. It is also interesting to see the effect on the two alcohol markers since their chemical and physical properties are different. EtG is polar and non-volatile whereas FAEEs are non-polar and semi-volatile. As stated before non-polar solvents can remove a greasy layer that may hinder the extraction of the analytes and protic solvents cause hair swelling and facilitate extraction. In our research is looked at these effects and the occurrence in the analysis of both markers. To our knowledge this has not yet been presented in another study whereas it is important because a difference in washing procedure can influence the reported concentration of the alcohol markers in hair. Hence, using the same cut-off values to come to a negative or positive test result for methods with different washing procedures may not be appropriate.
Table 1 Summary of washing solvents used in this research.
2. Material and methods
temperature gradient was applied of 30 ◦ C/min to 300 ◦ C, taking the total run time to 25.17 min. A chemical ionisation selected ion monitoring (SIM) program was used (88, 101, M + 1 for the FAEEs and for FAEEs-D5 93, 106 and M + 1) and the sample was introduced via HS SPME as described by Pragst et al. [8]. For EtG a shorter temperature program was used for this analysis and tandem MS was used as described by Paul et al. [9]. However, in this research no solid phase extraction (SPE) was used. Analysis for both EtG and FAEEs were performed with the helium carrier gas flow at 1 ml/min and the data was acquired and analysed using the MS workstation 6.9.2.
2.1. Solvents and reagents FAEEs-D5 and EtG-D5 were purchased from LGC Standards (Teddington, UK). N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) was purchased from Stratlab Ltd (Macherey-Nagel, Germany). Lab reagent grade dimethyl sulfoxide (DMSO), acetone, ethyl acetate, and dichloromethane (DCM); technical grade reagent sodium chloride; analytical grade acetic acid and diethyl ether; HPLC grade heptane, hexane were purchased from Fisher. HPLC grade methanol and toluene were bought from Sigma-Aldrich (Gillingham, UK). Methane of a purity of 99.9995%, Argon of a purity of 99.9999% and Helium of a purity of 99.999% were supplied by Air Liquide (Birmingham, UK). Ultra-pure water was prepared with a Purite Neptune system equipped with NCP 8 cartridges (Thame Oxon, UK). 2.2. Instrumentation and analysis Sample preparation was required before analysis. Various extraction solvents were used of which those with DMSO required an additional step to separate the DMSO from the extraction solvent heptane or hexane. These samples containing DMSO were frozen and the alkane layer was decanted to be analyzed. All samples were divided in two equal aliquots one for the determination of EtG and one for the analysis of FAEEs. The EtG samples were evaporated in a GC vial, derivatised with 10 l BSTFA in the presence of 10 l ethyl acetate at 80 ◦ C. The samples were subsequently injected directly on the gas chromatography mass spectrometry (GCMS) system. The FAEEs samples were evaporated in a 10 ml head space vial, extracted and injected onto the GCMS system by head space solid phase micro-extraction (HS SPME) with a polydimethylsiloxane/divinylbenzene (PDMS/DVB) fibre as was done by Pragst et al. [8]. Analysis was performed on a Varian CP3800 gas chromatograph equipped with a 320 series triple quadrupole mass spectrometer. The column used was a Varian Factor IV 5-MS (30 m × 0.25 mm × 0.25 m film thickness). For the FAEEs, the injector and ion source temperatures were set to 250 ◦ C and 150 ◦ C, respectively. The column temperature was initially set at 50 ◦ C for 1 min before increasing to 140 ◦ C at a rate of 20 ◦ C/min. The temperature was then increased to 220 ◦ C at 5 ◦ C/min before a final
Washing solvents
Analytes
References
De-ionised water and dichloromethane n-Heptane De-ionised water and n-heptane Dichloromethane 2× Diethyl ether and acetone
FAEEs
[10]
FAEEs EtG and FAEEs
[3,11–17] [15,18]
EtG EtG
De-ionised water, acetone and methanol Methanol
EtG
[19] [20] Used a similar protocol, but with ether [21]
EtG: FAEEs: EtG EtG
[9] [22] [23,24] [3–6,25–28]
EtG
[14,18,28–36]
EtG:
Used as part of the decontamination protocol by [8,14,15,18,21,28–37] Used as part of the decontamination protocol by [10,15,18]
Methanol and acetone Dichloromethane and methanol De-ionised water and acetone De-ionised water
FAEEs:
2.3. Obtaining hair samples from alcoholics Studies were performed on hair obtained from alcohol abusers via collaboration with a local clinic. These volunteers admitted to have been drinking more than 60 g of alcohol per day by answering the first authors questions concerning frequency of drinking and quantity per day, history of their consumption how often they washed their hair and what kind of hair care products were used. A bundle of hair of about a pencil thickness was then cut close to the scalp at the posterior vertex (back of the head). The hair of each subject was cut with a pair of scissors in 1 cm segments and mixed prior the washing step, dried and cut in mm pieces prior to the extraction. Either the 1 cm segments or the mm pieces were weighed in the washing or extraction vial to obtain 30 mg that was used as sample size in these experiments. 2.4. Comparison of washing solvents 1 ml of a washing solvent was applied to 30 mg hair; the mixture was shaken briefly after which the solvent was removed. 5 l of 200 ng/ml EtG-D5 and 20 l of a mixture of 2000 ng/ml ethyl oleateD5 and 1000 ng/ml of the other three deuterated FAEEs were then added to the washing solvent which were subsequently analysed for EtG and FAEEs. 1 ml methanol and the same amounts of deuterated standards were added to each of the washed hair samples for extraction. All experiments were performed in duplicate. The washing procedures that were investigated (from Refs. [3–6,8–37]) are listed in Table 1. One millilitre of each washing solvent was used and when more than one washing solvent was used in a washing procedure the solvents were used after each other.
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119 Table 2 Summary of extraction solvents used for FAEEs and EtG in this research. Ratios are expressed in volume. Extraction solvent
Analytes
References
Methanol Acetone n-Heptane: DMSO (4:1) n-Hexane: DMSO (4:1) de-ionised water Methanol: water (1:1) Methanol: DCM (1:1))
EtG FAEE FAEE FAEE EtG EtG
[32] – [5,12–18,27] [8] [6,9,14,15,19,23–29,31–36,38] [32] –
These washing solvents were selected based on the variation in polarity and have been used by several laboratories (see Table 1). To assess the washing efficiency of the solvents, analyte concentrations in both the extract and corresponding washing solvent were set to 100% thereby normalizing the results. The amount of analyte found in the washing step was then set out in a percentage of this total. This was described as the washing-out effect. 2.5. Comparison of extraction solvents For these studies 30 mg hair was washed with 1 ml DCM and subsequently washed with 1 ml methanol. Deuterated standards were added to the 1 ml extraction solvent as previously described. For extraction each of the samples were treated with one of the following solvents: acetone, water, methanol, methanol + water (1:1 v/v), methanol + DCM (1:1 v/v), heptane + DMSO (4:1 v/v) or hexane + DMSO (4:1 v/v). The extraction was facilitated by 30 min or 4 h of sonication. All experiments were performed in duplicate. Note that these extraction solvents have been previously discussed in the literature (see Table 2 where Refs. [5,6,8,9,12–19,23–29,31–36,38] were used) and provide a range of properties from protic polar to non-polar. 2.6. Transesterification For the extraction of the FAEEs in methanol by sonication it is known that transesterification can take place [8]. To investigate the extent of transesterfication, 30 and 500 ng/ml FAEEs in methanol were sonicated for 15, 30, 60, 120, 240 min up to 14 h of sonication. The methyl esters were detected using GCMS. The same conditions were used as for the ethyl esters, but other m/z values were selected based on the full scan (74, 87, M+). With this the ratio of methyl to ethyl esters was calculated. For this experiment two different sonication baths were used: a Fisherbrand FB11022 and an Ultrawave U500 of which the first generated more heat during sonication. This was done to investigate whether this additional heat would affect the amount of transesterification. 3. Results and discussion 3.1. Comparison of washing solvents Most solvents removed roughly half of the EtG during the washing step and half during the 30 min of sonication with methanol. This is in contrast with what Morini et al. [6] found in which the loss of EtG in the washes was considered to be negligible. This may be related to the hair samples that were used in their work. For our study, hair was analysed from people who claimed to regularly wash their hair (i.e. from once a day up to once a week) and more importantly also claimed to consume alcohol nearly constantly. High concentrations of EtG may be expected on the surface of hair if the main route of incorporation is via sweat [7,39]. This incorporation mechanism and possibly the difference in the concentration of EtG in sweat due to the high alcohol consumption of the subjects
117
Table 3 The washing efficiency of various washing solvents for EtG analysis from hair. Washing-out effect
Solvent(s)
High Medium
DCM, n-heptane H2 O, H2 O + DCM, H2 O + acetone, H2 O + acetone + methanol, methanol, methanol + DCM, methanol + acetone, diethyl-ether + acetone Methanol, H2 O + n-heptane
Low
used may be the reason why more EtG was washed-out in these experiments compared to what was observed by Morini et al. [6]. The effect of the washing step on EtG from hair was divided in three groups: high, medium and low. When at least twice as much analyte was found in the washing solvent compared to the extraction solution, the washing solvent was categorised as having a ‘high washing-out effect’ for EtG. For amounts that are washed away between 30% and 60% of the total analysed in the washing and extraction solvents, the washing procedure was categorised as ‘medium’, and for amounts less than 25%—‘low washing-out effects’. The results showed that most washes were comparable and had a medium washing-out effect for EtG (see Table 3). It can be observed from Table 3 that of all the solvents tested DCM and heptane showed superior washing-out effect and a lower than average was demonstrated by pure methanol as well as a combination of water and n-heptane. An explanation for this is that DCM and n-heptane may wash away more EtG than solvents that are comparable in polarity to EtG, due to (i) better hair swelling property which facilitates the extraction of the analytes or (ii) a more efficient grease removal. It may be possible that grease has encapsulated EtG, and therefore a non-polar solvent needs to be used first to liberate it. When water is used prior to a non-polar solvent, the water may form a hydrophilic layer which may hinder the hydrophobic solvent from dissolving the greasy layer on the hair shaft and is thereby unable to liberate the encapsulated EtG. This may explain why n-heptane alone washes better than water plus n-heptane. This also suggests that the hair sample should be dry before another wash solution is applied when the solution is immiscible with the previous wash solution. Water alone was as effective as most other solvents as a wash solvent but lower than DCM. Water is protic and thus has hair swelling properties. It is also an excellent solvent for EtG. However, as suggested earlier, the greasy layer may need to be removed first to liberate EtG. Normal hair hygiene including shampoo that removes this greasy layer, for instance, is thought to explain the significant decrease of EtG in the more distal segments of hair from chronic drinkers [40]. Another observation was that the results for pure methanol were variable; two samples showed that methanol had a ‘medium’ washing-out effect and two others samples showed that methanol had a ‘low’ washing-out effect. This may indicate that the washing-out effect may be difficult to determine. The amount that was removed during the washing procedures was relatively low for the FAEEs compared to what was observed for EtG. From the ratio between the response for the washes and extraction amount it was established that approximately 10% was removed, whereas for EtG this was about 50%. This observation is in agreement with the published hypothesis that normal hair hygiene may remove EtG and has a relatively low impact on the concentration of the FAEEs in hair [17]. In addition, the use of methanol + acetone or methanol + DCM removed more than 20% of the total FAEEs measured in the washing and extraction solutions. The use of these protocols were considered for the washing step as they showed a comparatively high washing-out effect (see Table 4). The efficiency of the washing solvents compared to the FAEEs extraction was divided in three groups: ‘high’, ‘medium’ and ‘low’
118
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
Table 4 The washing efficiency of various washing solvents for FAEEs analysis from hair. Efficiency
Washing solvents
High Medium
Methanol + acetone, DCM + methanol H2 O + DCM, DCM 2×, H2 O + acetone + methanol, methanol, H2 O + acetone, H2 O n-Heptane, H2 O + n-heptane, diethyl ether + acetone
Low
washing-out effect. The decision for the efficiency was based on the ratio between quantity detected in the washing solution(s) and the quantity extracted from hair after this washing procedure. The values in the range of: the average (11%) ± the standard deviation of 4% obtained from all solutions tested was used as ‘medium’ washingout effect. ‘Low’ and ‘high’ washing-out effects were below 8% and above 16% of the total that was removed from hair during the washing and extraction procedure. A wash with a solvent that swells the hair fibre and is nonmiscible with heptane (e.g. water) prior to the heptane wash reduced the washing-out effect of the heptane wash (see Table 4). Water may form a layer and thereby hinders the non-polar heptane wash to remove the FAEEs. Since this was not observed for the combination of water and DCM, some other factors such as drying time between the washes may be crucial. The use of methanol and DCM as wash steps for both EtG and FAEEs seems to be a good compromise. This procedure had a medium washing-out effect (approximately 50%) for EtG and a high washing-out effect for FAEEs (lower than 50%). 3.2. Comparison of extraction solvents The effect of extraction time was investigated. Two different extraction times were used: 30 min and 4 h. Results show that an extended period of 4 h did not result in an increase in the signal for EtG nor in a higher response (i.e. ratio area IS/area DIS). SPE was not performed on samples prior to final analysis using GCMS to avoid the introduction of other potential variables. Using methanol, EtG was detected after 30 min of extraction by sonication, but was not detected after 4 h. This was due to the extraction of other matrix constituents, thereby increasing the background in the chromatogram and degradation of this analyte may have contributed to this too. Therefore in this work 30 min was used as a guideline for the extraction time. From the extraction results for EtG it is clear that water is the best solvent (see Fig. 1). The second-best extraction solvents for EtG are those containing methanol. This is supported by the work of Jurado et al. [32] in which a lower recovery of EtG for methanol was found than water. For the alkane/DMSO mixtures EtG was not detected in the heptane or hexane probably due to the higher solubility of EtG in DMSO. EtG was not measured in DMSO because the evaporation process after extraction, necessary for subsequent
Fig. 2. FAEEs extraction with different solvents of which the heptane and hexane layers were measured but the extraction was performed in the presence of DMSO. The values were normalised to highest concentration for the fatty acid detected.
derivatisation, requires too much time. The EtG level in DMSO could possibly be tested when another method would be deployed such as SPE to redissolve EtG in another solvent which is beyond the scope of this article. In conclusion, in the current sample preparation process the second-best extraction solvent for EtG are those containing methanol. FAEEs were normalised based on the highest peak area found for ethyl myristate (E14:0), ethyl palmitate (E16:0), ethyl oleate (E18:1) and ethyl stearate (E18:0). This makes it easier to interpret all of the FAEEs in one figure since for these hair samples the ethyl oleate concentration in hair was much higher than the concentration of the other FAEEs. That ethyl oleate has a high concentration in hair compared to the other FAEEs is in itself interesting since ethyl oleate may be incorporated in hair to a lesser extent than the three other FAEEs who have a more linear structure and therefore can cross the membrane more easily. As was suggested by Potsch et al. [41] when discussing the general factors that influence the incorporation of an analyte in hair. This may be countered by the phenomenon described by Auwarter et al. [12] in which higher proportion of ethyl oleate were found in the sebum that was deposited on hair compared to the three other FAEEs. From the normalised chart (see Fig. 2) it is clear that hardly any ethyl myristate was detected for the methanol/water mixture. Ethyl myristate is thought to be low for the extraction with water and methanol due to the increased evaporation time after the extraction. Note the evaporation is performed after the extraction to perform HS SPME GCMS and FAEEs are semi-volatiles of which ethyl myristate has the lowest boiling point of the four that are analysed. That the low signal for ethyl myristate is caused by evaporation is supported by the undetected deuterated form of ethyl myristate. This phenomenon was also observed for the extraction with water. Moreover, water had the lowest extraction efficiency. This may be explained by the low solubility of the FAEEs in water and the relatively long evaporation step after the extraction. The first is supported by Pragst [10] who states that the extraction by aqueous buffer is inefficient for these highly lipophilic substances. Methanol (with and without DCM) is an interesting candidate for the extraction of FAEEs from hair (see Fig. 2). Methanol has been used in the past for both EtG and FAEEs extraction from hair. In this research methanol was established to be a reasonable candidate for EtG since it is the second-best extraction solvent of those that are tested. However, since water has a superior extraction efficiency this would be preferred for tests that solely analyse EtG. 3.3. Transesterification during the extraction with methanol
Fig. 1. EtG extraction with different solvents of which the heptane and hexane layers were measured but the extraction was performed in the presence of DMSO. The values are a percentage of the EtG that was extracted with water.
Methanol has the perceived disadvantage that the ethyl esters may transform into methyl esters, as was described by Pragst et al. [8]. Therefore, in our study transesterfication was tested. No difference was observed between the two sonication baths. At the
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
lower concentration (30 ng/ml), more in the range of values that can be detected in real hair samples, no methyl esters were detected. With the higher concentration sample (500 ng/ml), only two markers were observed to convert to methyl esters: ethyl stearate and ethyl palmitate. These markers showed 2% conversion from ethyl to methyl at sonication times ranging from 15 min to several hours. Sonication for 14 h gave up to 10% conversion and should therefore be avoided. 3.4. Practical implications The previous sections showed that with an extraction of 30 min (i) the intensity of the signal is higher than for several hours and (ii) only 2% transesterification occurs for the concentration of 500 ng/ml FAEEs. Under these conditions linear calibration curves (r2 > 0.99) were obtained with the sonication in the presence of methanol, indicating that due to the added deuterated internal standards FAEEs can be extracted in methanol by sonication and quantification can be used thereafter. Hence, an extraction with methanol by sonication is possible when the extraction time is 30 min. Some practical implications however, do arise from the washing and extraction experiments. Methanol seemed to be a good common ground for which both EtG and FAEEs were extracted, although to a lesser extent than the extraction solvents currently used in the industry for the single methods (water and a DMSO/heptane mixture, respectively). It may, however, be necessary to establish a new cut-off value, due to the difference in extraction efficiency for in particular EtG compared to the standard water extraction method. Possibly the percentage that is extracted with for instance MeOH may depend on the concentration of the alcohol metabolite present. Hence receiver operating characteristic curves need to be processed and analysed for this extraction method to determine the new cut-off values. It may also be required for autopsy cases to use an additional washing step prior to the washing procedure discussed in this article, especially when the hair is heavily contaminated with blood for instance. Yegles et al. [15] suggest a first wash with de-ionised water for this purpose. 4. Conclusion
solvents for hair alcohol testing may lead to discrepancies in the final reported concentrations of EtG and FAEE. Acknowledgements The authors thank all volunteers for their participation in the study and social workers from Huggard, Cardiff and Drug Aid Cymru for their support. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]
This work shows that a difference exists between the evaluated wash and extraction solutions. It was demonstrated that of the ones tested for EtG and FAEEs a very efficient washing protocol consists of a wash with DCM followed by methanol. An interesting finding was that the most efficient washing solution was neither the solution in which the analyte dissolves the best nor the solution that was known to swell the hair fibre. Methanol, with and without DCM, is an interesting possibility for the extraction of FAEEs and EtG. In particular the FAEEs analysis would benefit from extraction with methanol compared to the currently used DMSO/heptane mixture. EtG, on the other hand, shows a lower extraction for methanol than for the standard solution currently used (water). For the extraction with methanol 30 min of sonication is preferred over a longer extraction time. First, 14 h sonication results in the formation of a higher percentage of methyl esters. Second, 4 h sonication with methanol did not increase the signal for EtG compared to 30 min of sonication. In summary, 30 min of sonication with methanol is suitable for the extraction of FAEEs and for the combined extraction of EtG and FAEEs. This research may indicate that the lack of a common approach amongst commercial hair testing laboratories regarding washing and extraction
119
[26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]
P. Kintz, Forensic Sci. Int. 218 (2012) 2. G.A.A. Cooper, R. Kronstrand, P. Kintz, Forensic Sci. Int. 218 (2012) 20. L. Politi, F. Leone, L. Morini, A. Polettini, Anal. Biochem. 368 (2007) 1. R. Kronstrand, L. Brinkhagen, F.H. Nyström, Forensic Sci. Int. 215 (2012) 51. M.E. Albermann, F. Musshoff, B. Madea, Anal. Bioanal. Chem. 396 (2010) 2441. L. Morini, L. Politi, A. Groppi, C. Stramesi, A. Polettini, J. Mass Spectrom. 41 (2006) 34. P. Kintz, Analytical and Practical Aspects of Drug Testing in Hair, CRC/Taylor & Francis, Boca Raton, FL, 2007. F. Pragst, V. Auwaerter, F. Sporkert, K. Spiegel, Forensic Sci. Int. 121 (2001) 76. R. Paul, R. Kingston, L. Tsanaclis, A. Berry, A. Guwy, Forensic Sci. Int. 176 (2008) 82. F. Pragst, K. Spiegel, F. Sporkert, M. Bohnenkamp, Forensic Sci. Int. 107 (2000) 201. M.E. Albermann, F. Musshoff, B. Madea, Anal. Bioanal. Chem. 400 (2011) 175. V. Auwarter, F. Sporkert, S. Hartwig, F. Pragst, H. Vater, A. Diefenbacher, Clin. Chem. 47 (2001) 2114. V. Kulaga, J. Gareri, N. Fulga, G. Koren, Ther. Drug Monit. 32 (2010) 294. J. Gareri, B. Appenzeller, P. Walasek, G. Koren, Anal. Bioanal. Chem. 400 (2011) 183. M. Yegles, A. Labarthe, V. Auwarter, S. Hartwig, H. Vater, R. Wennig, F. Pragst, Forensic Sci. Int. 145 (2004) 167. S. Süße, C.M. Selavka, T. Mieczkowski, F. Pragst, Forensic Sci. Int. 196 (2010) 111. S. Hartwig, V. Auwärter, F. Pragst, Forensic Sci. Int. 131 (2003) 90. F. Pragst, M. Yegles, Ther. Drug Monit. 30 (2008) 255. P. Kintz, M. Villain, E. Vallet, M. Etter, G. Salquebre, V. Cirimele, Forensic Sci. Int. 176 (2008) 87. G. Skopp, G. Schmitt, L. Potsch, P. Dronner, R. Aderjan, R. Mattern, Alcohol Alcohol. 35 (2000) 283. M. Klys, K. Wozniak, S. Rojek, E. Rzepecka-Wozniak, P. Kowalski, Forensic Sci. Int. 179 (2008) e1. L. Politi, F. Mari, S. Furlanetto, E. Del Bravo, E. Bertol, J. Pharm. Biomed. Anal. 54 (2011) 1192. A. Alt, I. Janda, S. Seidl, F-M. Wurst, Alcohol Alcohol. 35 (2000) 313. I. Janda, W. Weinmann, T. Kuehnle, M. Lahode, A. Alt, Forensic Sci. Int. 128 (2002) 59. L. Morini, E. Marchei, F. Vagnarelli, O. Garcia Algar, A. Groppi, L. Mastrobattista, S. Pichini, Forensic Sci. Int. 196 (2010) 74. P. Bendroth, R. Kronstrand, A. Helander, J. Greby, N. Stephanson, P. Krantz, Forensic Sci. Int. 176 (2008) 76. F. Pragst, M. Rothe, B. Moench, M. Hastedt, S. Herre, D. Simmert, Forensic Sci. Int. 196 (2010) 101. R. Agius, P. Kintz, Drug Test. Anal. 2 (2010) 367. H. Kharbouche, F. Sporkert, S. Troxler, M. Augsburger, P. Mangin, C. Staub, J. Chromatogr. B 877 (2009) 2337. I. Tarcomnicu, A.L.N. van Nuijs, K. Aerts, M. De Doncker, A. Covaci, H. Neels, Forensic Sci. Int. 196 (2010) 121. B.M.R. Appenzeller, R. Agirman, P. Neuberg, M. Yegles, R. Wennig, Forensic Sci. Int. 173 (2007) 87. C. Jurado, T. Soriano, M.P. Gimenez, M. Menedez, Forensic Sci. Int. 145 (2004) 161. I. Kerekes, M. Yegles, U. Grimm, R. Wennig, Alcohol Alcohol. 44 (2008) 62. R. Agius, T. Nadulski, H.-G. Kahl, B. Dufaux, Forensic Sci. Int. 218 (2012) 10. L.M. Ferreira, T. Binz, M. Yegles, Forensic Sci. Int. 218 (2012) 123. F. Sporkert, H. Kharbouche, M.P. Augsburger, C. Klemm, M.R. Baumgartner, Forensic Sci. Int. 218 (2012) 97. I. Álvarez, A.J. Bermejo, M.J. Tabernero, P. Fernandez, P. Cabarcos, P. Lopez, Anal. Bioanal. Chem. 393 (2008) 1345. R. Lees, R. Kingston, T.M. Williams, G. Henderson, A. Lingford-Hughes, M. Hickman, Alcohol Alcohol. 47 (2012) 267. H. Kharbouche, N. Steiner, M. Morelato, C. Staub, B. Boutrel, P. Mangin, F. Sporkert, M. Augsburger, Alcohol 44 (2010) 507. L. Tsanaclis, R. Kingston, J. Wicks, Ann. Toxicol. Anal. 21 (2009) 67. L. Potsch, G. Skopp, M.R. Moeller, Forensic Sci. Int. 84 (1997) 25.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
An evaluation of washing and extraction techniques in the analysis of ethyl glucuronide and fatty acid ethyl esters from hair samples L.C.A.M. Bossers a,∗ , R. Paul a , A.J. Berry a , R. Kingston b , C. Middendorp c , A.J. Guwy d a
University of South Wales, ARW Building, Pontypridd CF374 AT, UK Lextox, The Maltings East Tyndall Street, Cardiff CF24 5EA, UK University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands d University of South Wales, 4 Forest Grove, Pontypridd CF37 1DL, UK b c
a r t i c l e
i n f o
Article history: Received 31 October 2013 Received in revised form 25 January 2014 Accepted 29 January 2014 Available online 13 February 2014 Keywords: Fatty acid ethyl esters Ethyl glucuronide Alcohol Sample preparation Washing extraction
a b s t r a c t Ethyl glucuronide (EtG) and fatty acid ethyl esters (FAEEs) are alcohol metabolites measured in hair and are after a decade of research thought to be the best markers in hair to indicate alcoholism and abstinence Forensic Sci. Int. 218 (2012) 2. A great body of work concerning EtG and FAEEs detection in hair has been performed. However, no recent extensive comparison has been made concerning washing and extraction procedures. This work shows that the washing procedure of dichloromethane followed by a methanol rinse of the hair sample removes more than 16% of the FAEEs and 50% of the total EtG that is present in and on the hair. A review of ten washing protocols (where the removal is categorised: high, medium or low) showed that a relatively high percentage of FAEEs was removed and “medium” amount of EtG compared to the other washing protocols. This work shows promising results for the extraction of the FAEEs and the combined extraction of FAEEs and EtG by using 30 min of sonication with methanol. More FAEEs were recovered from hair with methanol than with any other extraction solvent including the commonly used dimethyl sulfoxide/heptane mixture. When the sonication time was increased a higher percentage of transesterification of the FAEEs was observed, the extraction was “dirtier” as solids and a colour change was observed whereas the extraction efficiency did not increase. Therefore, washing the hair sample with dichloromethane and methanol followed by an addition of 1 ml of methanol and sonication for 30 min to extract the FAEEs and EtG from hair is recommended for FAEEs as well as for the combined analysis of EtG and FAEEs. A linear calibration curve (r2 > 0.99) was obtained for all analytes. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Though alcohol consumption is not illegal in many parts of the world, establishing someone’s alcohol consumption can be useful
Abbreviations: EtG, ethyl glucuronide; FAEE, fatty acid ethyl esters; SoHT, Society of Hair Testing; BSTFA, NO-bis(trimethylsilyl)trifluoroacetamide; DMSO, dimethyl sulfoxide; DCM, dichloromethane; GC, gas chromatography; MS, mass spectrometry; HS SPME, head space solid phase micro extraction; PDMS/DVB, polydimethylsiloxane/divinylbenzene; SIM, selected ion monitoring; SPE, solid phase extraction; E1, an ester formed from ethanol and a saturated (0 double bonds) 14 carbon long acid; 14:0, an ester formed from ethanol and a saturated (0 double bonds) 14 carbon long acid; E16:0, an ester formed from ethanol and a saturated (0 double bonds) 16 carbon long acid; E18, 1an ester formed from ethanol and a unsaturated (1 double bonds) 18 carbon long acid; E18:0, an ester formed from ethanol and a saturated (0 double bonds) 18 carbon long acid. ∗ Corresponding author. Tel.: +41438178006; fax: +44 1443 482285. E-mail addresses: [email protected], [email protected], [email protected] (L.C.A.M. Bossers), [email protected] (R. Paul), [email protected] (A.J. Berry), [email protected] (R. Kingston), [email protected] (C. Middendorp), [email protected] (A.J. Guwy). 1570-0232/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2014.01.049
in the legal and medical field. In child custody cases, for instance, it is important to determine if allegations of chronic heavy drinking are true. Knowledge about someone’s drinking behaviour is required when diseases occur, like fetal alcohol spectrum disorder, that can only be diagnosed and treated when it is known that someone has been drinking. Alcohol hair testing can aid in the confirmation of someone’s drinking behaviour in such cases [1]. Currently, the consensus of the Society of Hair Testing (SoHT) [1,2] describes that concentrations of the minor metabolites EtG (ethyl glucuronide) and FAEEs (fatty acid ethyl esters) in hair are expected to be above a cut-off value (30 pg/mg and 0.5 ng/mg) for chronic heavy drinkers. Several washing procedures and solutions are currently used for hair prior to the EtG and FAEEs analysis. Most FAEEs methods include a non-polar solvent as was suggested by Politi et al. [3] to remove the greasy layer from hair, which may otherwise hinder the extraction. However, it is clear that no general consensus is reached especially not for EtG analysis. The washing protocol described for EtG analysis differ in (i) the polarity of the solvents that are used, (ii) the volume of the solvents, (iii) the amount of washes and (iv) the application of sonication. Kronstrand et al. [4],
116
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
for instance, used ultrasonication and 3 ml per wash whereas Albermann et al. [5] uses 1 ml and Morini et al. [6] did not use sonication. In our study solvents are used with a range of polarities and distinct hair swelling properties; for compounds that are protic are said to have the ability to swell hair and thereby facilitate the diffusion and thus extraction of the analytes from hair [7]. In this study the polarity and protic ability of the solvent washes were varied to investigate the influence of these characteristics on the wash and extraction efficiency of EtG and FAEEs. It is also interesting to see the effect on the two alcohol markers since their chemical and physical properties are different. EtG is polar and non-volatile whereas FAEEs are non-polar and semi-volatile. As stated before non-polar solvents can remove a greasy layer that may hinder the extraction of the analytes and protic solvents cause hair swelling and facilitate extraction. In our research is looked at these effects and the occurrence in the analysis of both markers. To our knowledge this has not yet been presented in another study whereas it is important because a difference in washing procedure can influence the reported concentration of the alcohol markers in hair. Hence, using the same cut-off values to come to a negative or positive test result for methods with different washing procedures may not be appropriate.
Table 1 Summary of washing solvents used in this research.
2. Material and methods
temperature gradient was applied of 30 ◦ C/min to 300 ◦ C, taking the total run time to 25.17 min. A chemical ionisation selected ion monitoring (SIM) program was used (88, 101, M + 1 for the FAEEs and for FAEEs-D5 93, 106 and M + 1) and the sample was introduced via HS SPME as described by Pragst et al. [8]. For EtG a shorter temperature program was used for this analysis and tandem MS was used as described by Paul et al. [9]. However, in this research no solid phase extraction (SPE) was used. Analysis for both EtG and FAEEs were performed with the helium carrier gas flow at 1 ml/min and the data was acquired and analysed using the MS workstation 6.9.2.
2.1. Solvents and reagents FAEEs-D5 and EtG-D5 were purchased from LGC Standards (Teddington, UK). N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) was purchased from Stratlab Ltd (Macherey-Nagel, Germany). Lab reagent grade dimethyl sulfoxide (DMSO), acetone, ethyl acetate, and dichloromethane (DCM); technical grade reagent sodium chloride; analytical grade acetic acid and diethyl ether; HPLC grade heptane, hexane were purchased from Fisher. HPLC grade methanol and toluene were bought from Sigma-Aldrich (Gillingham, UK). Methane of a purity of 99.9995%, Argon of a purity of 99.9999% and Helium of a purity of 99.999% were supplied by Air Liquide (Birmingham, UK). Ultra-pure water was prepared with a Purite Neptune system equipped with NCP 8 cartridges (Thame Oxon, UK). 2.2. Instrumentation and analysis Sample preparation was required before analysis. Various extraction solvents were used of which those with DMSO required an additional step to separate the DMSO from the extraction solvent heptane or hexane. These samples containing DMSO were frozen and the alkane layer was decanted to be analyzed. All samples were divided in two equal aliquots one for the determination of EtG and one for the analysis of FAEEs. The EtG samples were evaporated in a GC vial, derivatised with 10 l BSTFA in the presence of 10 l ethyl acetate at 80 ◦ C. The samples were subsequently injected directly on the gas chromatography mass spectrometry (GCMS) system. The FAEEs samples were evaporated in a 10 ml head space vial, extracted and injected onto the GCMS system by head space solid phase micro-extraction (HS SPME) with a polydimethylsiloxane/divinylbenzene (PDMS/DVB) fibre as was done by Pragst et al. [8]. Analysis was performed on a Varian CP3800 gas chromatograph equipped with a 320 series triple quadrupole mass spectrometer. The column used was a Varian Factor IV 5-MS (30 m × 0.25 mm × 0.25 m film thickness). For the FAEEs, the injector and ion source temperatures were set to 250 ◦ C and 150 ◦ C, respectively. The column temperature was initially set at 50 ◦ C for 1 min before increasing to 140 ◦ C at a rate of 20 ◦ C/min. The temperature was then increased to 220 ◦ C at 5 ◦ C/min before a final
Washing solvents
Analytes
References
De-ionised water and dichloromethane n-Heptane De-ionised water and n-heptane Dichloromethane 2× Diethyl ether and acetone
FAEEs
[10]
FAEEs EtG and FAEEs
[3,11–17] [15,18]
EtG EtG
De-ionised water, acetone and methanol Methanol
EtG
[19] [20] Used a similar protocol, but with ether [21]
EtG: FAEEs: EtG EtG
[9] [22] [23,24] [3–6,25–28]
EtG
[14,18,28–36]
EtG:
Used as part of the decontamination protocol by [8,14,15,18,21,28–37] Used as part of the decontamination protocol by [10,15,18]
Methanol and acetone Dichloromethane and methanol De-ionised water and acetone De-ionised water
FAEEs:
2.3. Obtaining hair samples from alcoholics Studies were performed on hair obtained from alcohol abusers via collaboration with a local clinic. These volunteers admitted to have been drinking more than 60 g of alcohol per day by answering the first authors questions concerning frequency of drinking and quantity per day, history of their consumption how often they washed their hair and what kind of hair care products were used. A bundle of hair of about a pencil thickness was then cut close to the scalp at the posterior vertex (back of the head). The hair of each subject was cut with a pair of scissors in 1 cm segments and mixed prior the washing step, dried and cut in mm pieces prior to the extraction. Either the 1 cm segments or the mm pieces were weighed in the washing or extraction vial to obtain 30 mg that was used as sample size in these experiments. 2.4. Comparison of washing solvents 1 ml of a washing solvent was applied to 30 mg hair; the mixture was shaken briefly after which the solvent was removed. 5 l of 200 ng/ml EtG-D5 and 20 l of a mixture of 2000 ng/ml ethyl oleateD5 and 1000 ng/ml of the other three deuterated FAEEs were then added to the washing solvent which were subsequently analysed for EtG and FAEEs. 1 ml methanol and the same amounts of deuterated standards were added to each of the washed hair samples for extraction. All experiments were performed in duplicate. The washing procedures that were investigated (from Refs. [3–6,8–37]) are listed in Table 1. One millilitre of each washing solvent was used and when more than one washing solvent was used in a washing procedure the solvents were used after each other.
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119 Table 2 Summary of extraction solvents used for FAEEs and EtG in this research. Ratios are expressed in volume. Extraction solvent
Analytes
References
Methanol Acetone n-Heptane: DMSO (4:1) n-Hexane: DMSO (4:1) de-ionised water Methanol: water (1:1) Methanol: DCM (1:1))
EtG FAEE FAEE FAEE EtG EtG
[32] – [5,12–18,27] [8] [6,9,14,15,19,23–29,31–36,38] [32] –
These washing solvents were selected based on the variation in polarity and have been used by several laboratories (see Table 1). To assess the washing efficiency of the solvents, analyte concentrations in both the extract and corresponding washing solvent were set to 100% thereby normalizing the results. The amount of analyte found in the washing step was then set out in a percentage of this total. This was described as the washing-out effect. 2.5. Comparison of extraction solvents For these studies 30 mg hair was washed with 1 ml DCM and subsequently washed with 1 ml methanol. Deuterated standards were added to the 1 ml extraction solvent as previously described. For extraction each of the samples were treated with one of the following solvents: acetone, water, methanol, methanol + water (1:1 v/v), methanol + DCM (1:1 v/v), heptane + DMSO (4:1 v/v) or hexane + DMSO (4:1 v/v). The extraction was facilitated by 30 min or 4 h of sonication. All experiments were performed in duplicate. Note that these extraction solvents have been previously discussed in the literature (see Table 2 where Refs. [5,6,8,9,12–19,23–29,31–36,38] were used) and provide a range of properties from protic polar to non-polar. 2.6. Transesterification For the extraction of the FAEEs in methanol by sonication it is known that transesterification can take place [8]. To investigate the extent of transesterfication, 30 and 500 ng/ml FAEEs in methanol were sonicated for 15, 30, 60, 120, 240 min up to 14 h of sonication. The methyl esters were detected using GCMS. The same conditions were used as for the ethyl esters, but other m/z values were selected based on the full scan (74, 87, M+). With this the ratio of methyl to ethyl esters was calculated. For this experiment two different sonication baths were used: a Fisherbrand FB11022 and an Ultrawave U500 of which the first generated more heat during sonication. This was done to investigate whether this additional heat would affect the amount of transesterification. 3. Results and discussion 3.1. Comparison of washing solvents Most solvents removed roughly half of the EtG during the washing step and half during the 30 min of sonication with methanol. This is in contrast with what Morini et al. [6] found in which the loss of EtG in the washes was considered to be negligible. This may be related to the hair samples that were used in their work. For our study, hair was analysed from people who claimed to regularly wash their hair (i.e. from once a day up to once a week) and more importantly also claimed to consume alcohol nearly constantly. High concentrations of EtG may be expected on the surface of hair if the main route of incorporation is via sweat [7,39]. This incorporation mechanism and possibly the difference in the concentration of EtG in sweat due to the high alcohol consumption of the subjects
117
Table 3 The washing efficiency of various washing solvents for EtG analysis from hair. Washing-out effect
Solvent(s)
High Medium
DCM, n-heptane H2 O, H2 O + DCM, H2 O + acetone, H2 O + acetone + methanol, methanol, methanol + DCM, methanol + acetone, diethyl-ether + acetone Methanol, H2 O + n-heptane
Low
used may be the reason why more EtG was washed-out in these experiments compared to what was observed by Morini et al. [6]. The effect of the washing step on EtG from hair was divided in three groups: high, medium and low. When at least twice as much analyte was found in the washing solvent compared to the extraction solution, the washing solvent was categorised as having a ‘high washing-out effect’ for EtG. For amounts that are washed away between 30% and 60% of the total analysed in the washing and extraction solvents, the washing procedure was categorised as ‘medium’, and for amounts less than 25%—‘low washing-out effects’. The results showed that most washes were comparable and had a medium washing-out effect for EtG (see Table 3). It can be observed from Table 3 that of all the solvents tested DCM and heptane showed superior washing-out effect and a lower than average was demonstrated by pure methanol as well as a combination of water and n-heptane. An explanation for this is that DCM and n-heptane may wash away more EtG than solvents that are comparable in polarity to EtG, due to (i) better hair swelling property which facilitates the extraction of the analytes or (ii) a more efficient grease removal. It may be possible that grease has encapsulated EtG, and therefore a non-polar solvent needs to be used first to liberate it. When water is used prior to a non-polar solvent, the water may form a hydrophilic layer which may hinder the hydrophobic solvent from dissolving the greasy layer on the hair shaft and is thereby unable to liberate the encapsulated EtG. This may explain why n-heptane alone washes better than water plus n-heptane. This also suggests that the hair sample should be dry before another wash solution is applied when the solution is immiscible with the previous wash solution. Water alone was as effective as most other solvents as a wash solvent but lower than DCM. Water is protic and thus has hair swelling properties. It is also an excellent solvent for EtG. However, as suggested earlier, the greasy layer may need to be removed first to liberate EtG. Normal hair hygiene including shampoo that removes this greasy layer, for instance, is thought to explain the significant decrease of EtG in the more distal segments of hair from chronic drinkers [40]. Another observation was that the results for pure methanol were variable; two samples showed that methanol had a ‘medium’ washing-out effect and two others samples showed that methanol had a ‘low’ washing-out effect. This may indicate that the washing-out effect may be difficult to determine. The amount that was removed during the washing procedures was relatively low for the FAEEs compared to what was observed for EtG. From the ratio between the response for the washes and extraction amount it was established that approximately 10% was removed, whereas for EtG this was about 50%. This observation is in agreement with the published hypothesis that normal hair hygiene may remove EtG and has a relatively low impact on the concentration of the FAEEs in hair [17]. In addition, the use of methanol + acetone or methanol + DCM removed more than 20% of the total FAEEs measured in the washing and extraction solutions. The use of these protocols were considered for the washing step as they showed a comparatively high washing-out effect (see Table 4). The efficiency of the washing solvents compared to the FAEEs extraction was divided in three groups: ‘high’, ‘medium’ and ‘low’
118
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
Table 4 The washing efficiency of various washing solvents for FAEEs analysis from hair. Efficiency
Washing solvents
High Medium
Methanol + acetone, DCM + methanol H2 O + DCM, DCM 2×, H2 O + acetone + methanol, methanol, H2 O + acetone, H2 O n-Heptane, H2 O + n-heptane, diethyl ether + acetone
Low
washing-out effect. The decision for the efficiency was based on the ratio between quantity detected in the washing solution(s) and the quantity extracted from hair after this washing procedure. The values in the range of: the average (11%) ± the standard deviation of 4% obtained from all solutions tested was used as ‘medium’ washingout effect. ‘Low’ and ‘high’ washing-out effects were below 8% and above 16% of the total that was removed from hair during the washing and extraction procedure. A wash with a solvent that swells the hair fibre and is nonmiscible with heptane (e.g. water) prior to the heptane wash reduced the washing-out effect of the heptane wash (see Table 4). Water may form a layer and thereby hinders the non-polar heptane wash to remove the FAEEs. Since this was not observed for the combination of water and DCM, some other factors such as drying time between the washes may be crucial. The use of methanol and DCM as wash steps for both EtG and FAEEs seems to be a good compromise. This procedure had a medium washing-out effect (approximately 50%) for EtG and a high washing-out effect for FAEEs (lower than 50%). 3.2. Comparison of extraction solvents The effect of extraction time was investigated. Two different extraction times were used: 30 min and 4 h. Results show that an extended period of 4 h did not result in an increase in the signal for EtG nor in a higher response (i.e. ratio area IS/area DIS). SPE was not performed on samples prior to final analysis using GCMS to avoid the introduction of other potential variables. Using methanol, EtG was detected after 30 min of extraction by sonication, but was not detected after 4 h. This was due to the extraction of other matrix constituents, thereby increasing the background in the chromatogram and degradation of this analyte may have contributed to this too. Therefore in this work 30 min was used as a guideline for the extraction time. From the extraction results for EtG it is clear that water is the best solvent (see Fig. 1). The second-best extraction solvents for EtG are those containing methanol. This is supported by the work of Jurado et al. [32] in which a lower recovery of EtG for methanol was found than water. For the alkane/DMSO mixtures EtG was not detected in the heptane or hexane probably due to the higher solubility of EtG in DMSO. EtG was not measured in DMSO because the evaporation process after extraction, necessary for subsequent
Fig. 2. FAEEs extraction with different solvents of which the heptane and hexane layers were measured but the extraction was performed in the presence of DMSO. The values were normalised to highest concentration for the fatty acid detected.
derivatisation, requires too much time. The EtG level in DMSO could possibly be tested when another method would be deployed such as SPE to redissolve EtG in another solvent which is beyond the scope of this article. In conclusion, in the current sample preparation process the second-best extraction solvent for EtG are those containing methanol. FAEEs were normalised based on the highest peak area found for ethyl myristate (E14:0), ethyl palmitate (E16:0), ethyl oleate (E18:1) and ethyl stearate (E18:0). This makes it easier to interpret all of the FAEEs in one figure since for these hair samples the ethyl oleate concentration in hair was much higher than the concentration of the other FAEEs. That ethyl oleate has a high concentration in hair compared to the other FAEEs is in itself interesting since ethyl oleate may be incorporated in hair to a lesser extent than the three other FAEEs who have a more linear structure and therefore can cross the membrane more easily. As was suggested by Potsch et al. [41] when discussing the general factors that influence the incorporation of an analyte in hair. This may be countered by the phenomenon described by Auwarter et al. [12] in which higher proportion of ethyl oleate were found in the sebum that was deposited on hair compared to the three other FAEEs. From the normalised chart (see Fig. 2) it is clear that hardly any ethyl myristate was detected for the methanol/water mixture. Ethyl myristate is thought to be low for the extraction with water and methanol due to the increased evaporation time after the extraction. Note the evaporation is performed after the extraction to perform HS SPME GCMS and FAEEs are semi-volatiles of which ethyl myristate has the lowest boiling point of the four that are analysed. That the low signal for ethyl myristate is caused by evaporation is supported by the undetected deuterated form of ethyl myristate. This phenomenon was also observed for the extraction with water. Moreover, water had the lowest extraction efficiency. This may be explained by the low solubility of the FAEEs in water and the relatively long evaporation step after the extraction. The first is supported by Pragst [10] who states that the extraction by aqueous buffer is inefficient for these highly lipophilic substances. Methanol (with and without DCM) is an interesting candidate for the extraction of FAEEs from hair (see Fig. 2). Methanol has been used in the past for both EtG and FAEEs extraction from hair. In this research methanol was established to be a reasonable candidate for EtG since it is the second-best extraction solvent of those that are tested. However, since water has a superior extraction efficiency this would be preferred for tests that solely analyse EtG. 3.3. Transesterification during the extraction with methanol
Fig. 1. EtG extraction with different solvents of which the heptane and hexane layers were measured but the extraction was performed in the presence of DMSO. The values are a percentage of the EtG that was extracted with water.
Methanol has the perceived disadvantage that the ethyl esters may transform into methyl esters, as was described by Pragst et al. [8]. Therefore, in our study transesterfication was tested. No difference was observed between the two sonication baths. At the
L.C.A.M. Bossers et al. / J. Chromatogr. B 953–954 (2014) 115–119
lower concentration (30 ng/ml), more in the range of values that can be detected in real hair samples, no methyl esters were detected. With the higher concentration sample (500 ng/ml), only two markers were observed to convert to methyl esters: ethyl stearate and ethyl palmitate. These markers showed 2% conversion from ethyl to methyl at sonication times ranging from 15 min to several hours. Sonication for 14 h gave up to 10% conversion and should therefore be avoided. 3.4. Practical implications The previous sections showed that with an extraction of 30 min (i) the intensity of the signal is higher than for several hours and (ii) only 2% transesterification occurs for the concentration of 500 ng/ml FAEEs. Under these conditions linear calibration curves (r2 > 0.99) were obtained with the sonication in the presence of methanol, indicating that due to the added deuterated internal standards FAEEs can be extracted in methanol by sonication and quantification can be used thereafter. Hence, an extraction with methanol by sonication is possible when the extraction time is 30 min. Some practical implications however, do arise from the washing and extraction experiments. Methanol seemed to be a good common ground for which both EtG and FAEEs were extracted, although to a lesser extent than the extraction solvents currently used in the industry for the single methods (water and a DMSO/heptane mixture, respectively). It may, however, be necessary to establish a new cut-off value, due to the difference in extraction efficiency for in particular EtG compared to the standard water extraction method. Possibly the percentage that is extracted with for instance MeOH may depend on the concentration of the alcohol metabolite present. Hence receiver operating characteristic curves need to be processed and analysed for this extraction method to determine the new cut-off values. It may also be required for autopsy cases to use an additional washing step prior to the washing procedure discussed in this article, especially when the hair is heavily contaminated with blood for instance. Yegles et al. [15] suggest a first wash with de-ionised water for this purpose. 4. Conclusion
solvents for hair alcohol testing may lead to discrepancies in the final reported concentrations of EtG and FAEE. Acknowledgements The authors thank all volunteers for their participation in the study and social workers from Huggard, Cardiff and Drug Aid Cymru for their support. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]
This work shows that a difference exists between the evaluated wash and extraction solutions. It was demonstrated that of the ones tested for EtG and FAEEs a very efficient washing protocol consists of a wash with DCM followed by methanol. An interesting finding was that the most efficient washing solution was neither the solution in which the analyte dissolves the best nor the solution that was known to swell the hair fibre. Methanol, with and without DCM, is an interesting possibility for the extraction of FAEEs and EtG. In particular the FAEEs analysis would benefit from extraction with methanol compared to the currently used DMSO/heptane mixture. EtG, on the other hand, shows a lower extraction for methanol than for the standard solution currently used (water). For the extraction with methanol 30 min of sonication is preferred over a longer extraction time. First, 14 h sonication results in the formation of a higher percentage of methyl esters. Second, 4 h sonication with methanol did not increase the signal for EtG compared to 30 min of sonication. In summary, 30 min of sonication with methanol is suitable for the extraction of FAEEs and for the combined extraction of EtG and FAEEs. This research may indicate that the lack of a common approach amongst commercial hair testing laboratories regarding washing and extraction
119
[26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]
P. Kintz, Forensic Sci. Int. 218 (2012) 2. G.A.A. Cooper, R. Kronstrand, P. Kintz, Forensic Sci. Int. 218 (2012) 20. L. Politi, F. Leone, L. Morini, A. Polettini, Anal. Biochem. 368 (2007) 1. R. Kronstrand, L. Brinkhagen, F.H. Nyström, Forensic Sci. Int. 215 (2012) 51. M.E. Albermann, F. Musshoff, B. Madea, Anal. Bioanal. Chem. 396 (2010) 2441. L. Morini, L. Politi, A. Groppi, C. Stramesi, A. Polettini, J. Mass Spectrom. 41 (2006) 34. P. Kintz, Analytical and Practical Aspects of Drug Testing in Hair, CRC/Taylor & Francis, Boca Raton, FL, 2007. F. Pragst, V. Auwaerter, F. Sporkert, K. Spiegel, Forensic Sci. Int. 121 (2001) 76. R. Paul, R. Kingston, L. Tsanaclis, A. Berry, A. Guwy, Forensic Sci. Int. 176 (2008) 82. F. Pragst, K. Spiegel, F. Sporkert, M. Bohnenkamp, Forensic Sci. Int. 107 (2000) 201. M.E. Albermann, F. Musshoff, B. Madea, Anal. Bioanal. Chem. 400 (2011) 175. V. Auwarter, F. Sporkert, S. Hartwig, F. Pragst, H. Vater, A. Diefenbacher, Clin. Chem. 47 (2001) 2114. V. Kulaga, J. Gareri, N. Fulga, G. Koren, Ther. Drug Monit. 32 (2010) 294. J. Gareri, B. Appenzeller, P. Walasek, G. Koren, Anal. Bioanal. Chem. 400 (2011) 183. M. Yegles, A. Labarthe, V. Auwarter, S. Hartwig, H. Vater, R. Wennig, F. Pragst, Forensic Sci. Int. 145 (2004) 167. S. Süße, C.M. Selavka, T. Mieczkowski, F. Pragst, Forensic Sci. Int. 196 (2010) 111. S. Hartwig, V. Auwärter, F. Pragst, Forensic Sci. Int. 131 (2003) 90. F. Pragst, M. Yegles, Ther. Drug Monit. 30 (2008) 255. P. Kintz, M. Villain, E. Vallet, M. Etter, G. Salquebre, V. Cirimele, Forensic Sci. Int. 176 (2008) 87. G. Skopp, G. Schmitt, L. Potsch, P. Dronner, R. Aderjan, R. Mattern, Alcohol Alcohol. 35 (2000) 283. M. Klys, K. Wozniak, S. Rojek, E. Rzepecka-Wozniak, P. Kowalski, Forensic Sci. Int. 179 (2008) e1. L. Politi, F. Mari, S. Furlanetto, E. Del Bravo, E. Bertol, J. Pharm. Biomed. Anal. 54 (2011) 1192. A. Alt, I. Janda, S. Seidl, F-M. Wurst, Alcohol Alcohol. 35 (2000) 313. I. Janda, W. Weinmann, T. Kuehnle, M. Lahode, A. Alt, Forensic Sci. Int. 128 (2002) 59. L. Morini, E. Marchei, F. Vagnarelli, O. Garcia Algar, A. Groppi, L. Mastrobattista, S. Pichini, Forensic Sci. Int. 196 (2010) 74. P. Bendroth, R. Kronstrand, A. Helander, J. Greby, N. Stephanson, P. Krantz, Forensic Sci. Int. 176 (2008) 76. F. Pragst, M. Rothe, B. Moench, M. Hastedt, S. Herre, D. Simmert, Forensic Sci. Int. 196 (2010) 101. R. Agius, P. Kintz, Drug Test. Anal. 2 (2010) 367. H. Kharbouche, F. Sporkert, S. Troxler, M. Augsburger, P. Mangin, C. Staub, J. Chromatogr. B 877 (2009) 2337. I. Tarcomnicu, A.L.N. van Nuijs, K. Aerts, M. De Doncker, A. Covaci, H. Neels, Forensic Sci. Int. 196 (2010) 121. B.M.R. Appenzeller, R. Agirman, P. Neuberg, M. Yegles, R. Wennig, Forensic Sci. Int. 173 (2007) 87. C. Jurado, T. Soriano, M.P. Gimenez, M. Menedez, Forensic Sci. Int. 145 (2004) 161. I. Kerekes, M. Yegles, U. Grimm, R. Wennig, Alcohol Alcohol. 44 (2008) 62. R. Agius, T. Nadulski, H.-G. Kahl, B. Dufaux, Forensic Sci. Int. 218 (2012) 10. L.M. Ferreira, T. Binz, M. Yegles, Forensic Sci. Int. 218 (2012) 123. F. Sporkert, H. Kharbouche, M.P. Augsburger, C. Klemm, M.R. Baumgartner, Forensic Sci. Int. 218 (2012) 97. I. Álvarez, A.J. Bermejo, M.J. Tabernero, P. Fernandez, P. Cabarcos, P. Lopez, Anal. Bioanal. Chem. 393 (2008) 1345. R. Lees, R. Kingston, T.M. Williams, G. Henderson, A. Lingford-Hughes, M. Hickman, Alcohol Alcohol. 47 (2012) 267. H. Kharbouche, N. Steiner, M. Morelato, C. Staub, B. Boutrel, P. Mangin, F. Sporkert, M. Augsburger, Alcohol 44 (2010) 507. L. Tsanaclis, R. Kingston, J. Wicks, Ann. Toxicol. Anal. 21 (2009) 67. L. Potsch, G. Skopp, M.R. Moeller, Forensic Sci. Int. 84 (1997) 25.
