Development of a Targeted GC/MS Screening Method and Validation of an HPLC/DAD Quantiﬁcation Method for Piperazines–Amphetamines Mixtures in Seized Material

Egyptian Journal of Forensic Sciences (2014) xxx, xxx–xxx

Contents lists available at ScienceDirect

Egyptian Journal of Forensic Sciences

journal homepage: www.ejfs.org

ORIGINAL ARTICLE Development of a targeted GC/MS screening method and validation of an HPLC/DAD quantiﬁcation method for piperazines–amphetamines mixtures in seized material

Yacine Boumrah a,b,1, Martine Rosset c,*, Yannick Lecompte d, Sabrina Bouanani a, Kamel Khimeche e, Abdellah Dahmani b a Drugs Laboratory, National Institute of Criminalistics and Criminology, Algiers, Algeria b Laboratoire de thermodynamique et mode´lisation mole´culaire, Faculte´ de Chimie, USTHB, BP 32 El-Alia Bab-Ezzouar, 16111 Alger, Algeria c Etat-major de la marine, bureau maitrise des risques, 2 rue royale, 75008 Paris, France d Institut de recherche criminelle de la gendarmerie nationale (IRCGN), De´partement toxicologie, 1 boulevard The´ophile Sueur, 93111 Rosny-Sous-Bois, France e Ecole Militaire Polytechnique EMP, BP 17 Bordj–el-Bahri, Alger, Algeria

Received 13 October 2013; revised 20 January 2014; accepted 23 March 2014

KEYWORDS Abstract Piperazine related drugs are sold as party pills in the form of tablets, capsules, liquids or Piperazines; powders. These party pills can contain several piperazine derivatives, or even a mixture of pipera- Amphetamines; zines and amphetamine derivatives. This paper therefore describes a screening method using a gas Designer drugs; chromatography–mass spectrometry technique allowing the separation and the identification of b-Expectation tolerance active components within these mixtures by a combined silylation and acylation derivatization interval; procedure. The studied substances namely: 1-benzylpiperazine (BZP), 1-(3,4-methylenedioxyben- Benzylpiperazine; zyl)piperazine (MDBP), 1-(3-trifluoromethylphenyl)piperazine (TFMPP), 1-(3-chlorophenyl)piper- 1-(3-Trifluoromethylphenyl) azine (mCPP), 1-(4-methoxyphenyl)piperazine (MeOPP), amphetamine, methamphetamine, piperazine ephedrine, pseudoephedrine, 3,4-methylenedioxy-N-methamphetamine (MDMA), 3,4-methylenedioxyamphetamine (MDA), 3,4-methylenedioxy-N-ethylamphetamine (MDEA) and N-methyl-1,3- benzodioxolylbutanamine (MBDB), are separated.

* Corresponding author. Tel.: +33 1 42 92 19 42. E-mail addresses: [email protected] (Y. Boumrah), [email protected] (M. Rosset). 1 Tel.: +213 0663845657. Peer review under responsibility of Forensic Medicine Authority.

Production and hosting by Elsevier

2090-536X ª 2014 Production and hosting by Elsevier B.V. on behalf of Forensic Medicine Authority. http://dx.doi.org/10.1016/j.ejfs.2014.03.002

A high performance liquid chromatography with a diode array detector quantitative method was validated following an approach using accuracy proﬁles based on b-expectation tolerance intervals for total error measurement. ª 2014 Production and hosting by Elsevier B.V. on behalf of Forensic Medicine Authority.

1. Introduction 2. Experimental

Synthetic drugs are among the most commonly abused drugs 2.1. Reagents and standards in the world. Some are derived from phenylethylamine, as is 3,4-methylenedioxymethamphetamine (MDMA). However, Reference compounds 1-(4-methoxyphenyl)piperazine as these drugs became scheduled, new derivatives (like pipera- (MeOPP) dihydrochloride, 1-(4-methylphenyl)piperazine (pTP) zines) have appeared on the market to bypass the law. dihydrochloride (internal standard), 1-(3-chlorophenyl)piperazine Chemically, piperazines are synthesized from the same (mCPP) hydrochloride, 1-(3-trifluoromethylphenyl)piperazine piperazine moiety and they can be classified into two classes: (TFMPP) hydrochloride, 1-(3,4-methylenedioxybenzyl)pipera- benzylpiperazines such as 1-benzylpiperazine (BZP) and its zine (MDBP) free base and 1-benzylpiperazine (BZP) were methylenedioxy analog 1-(3,4-methylenedioxybenzyl)pipera- purchased from Lancaster (Frankfurt, Germany). Analytical zine (MDBP) and phenylpiperazines such as 1-(3-trifluorometh- standards of amphetamine, methamphetamine, ephedrine, ylphenyl)piperazine (TFMPP), 1-(3-chlorophenyl)piperazine pseudoephedrine, 3,4-methylenedioxy-N-methamphet- (mCPP) and 1-(4-methoxyphenyl)piperazine (MeOPP). amine (MDMA), 3,4-methylenedioxyamphetamine (MDA), Piperazine-like compounds have spread around the world: 3,4-methylenedioxy-N-ethylamphetamine (MDEA) and according to Kovaleva et al. it was estimated that approxi- N-methyl-1,3-benzodioxolylbutanamine (MBDB) were mately 823,000 tablets of mCPP have been seized in 2008 in purchased from Promochem (Wesel, Germany) as 1 mg/mL the European Union.1 In the Netherlands, the number of methanolic solutions. Lactose was purchased from Fluka mCPP tablets seized alone or in combination with MDMA (Steinheim, Germany). increased significantly.2 A recent survey in the United King- Heptafluorobutyric anhydride (HFBA) and the silylation dom, found that piperazines are among the most common reagent SilPrep were purchased from Alltech (Templemars, active drugs in tablets purchased from internet supplier sites.3 France). These drugs have also been seized in New Zealand,4,5 Japan,6 Methanol and acetonitrile (HPLC grade), n-hexane, potas- 7 Brazil and many other countries such as Bulgaria, Sweden sium dihydrogenphosphate (KH2PO4) and 85% ortho-phos- and South Africa.8 phoric acid (analytical grade) were purchased from VWR Commonly, piperazines are sold as party pills in the form of Prolabo (Fontenay-sous-Bois, France). tablets, capsules, liquids or powders on the black drug-market and in so-called headshops or over the internet under the 2.2. Calibrators and specimen preparation for GC/MS names of ‘‘Rapture,’’ ‘‘Frenzy,’’ ‘‘Bliss,’’ ‘‘Charge,’’ ‘‘Herbal qualitative analysis ecstasy,’’ ‘‘A2,’’ ‘‘Legal X’’ and ‘‘Legal E’’.9,10 They are also 9–17 found in tablets sold as ecstasy or amphetamine. Stock solutions of BZP, MBDP, mCPP, TFMPP and MeOPP Generally, the party pills contain piperazine blends. Besides 18,19 at a concentration of 2 mg/mL were prepared in methanol. the most prevalent mixture of BZP with TFMPP, there are Working solutions of amphetamine–piperazine mixtures were also combinations of up to four different piperazines which are obtained by dilution in methanol. sold to consumers as ‘‘party pills’’ or ‘‘P.E.P. pills’’ called 20–24 All these solutions were stored protected from light at X4. Furthermore, a large number of seized tablets con- +5 C. taining mixtures of piperazine and amphetamine type stimu- 2,11–13,25–35 GC/MS analyses of the piperazine derivative mixture were lants (ATS) have been reported. performed first without derivatization at a concentration of Several methods related to the identification of piperazine 2 mg/mL, and then carried out after silylation and acylation like-compounds in both seized and biological samples have 12,36–42 at a concentration of 10 mg/L. The silylation was performed been reported but an identification method for amphet- by adding 125 lL of SilPrep reagent to the evaporation res- amine-piperazine mixtures, as encountered in a large number idue of 125 lL of each 10 mg/L working solution. The acyla- of seizures, has not been described yet. The aim of this work tion was performed by adding 25 lL of HFBA reagent to is therefore to develop a qualitative method using a gas chro- the evaporation residue of 300 lL of each 10 mg/L working matography-mass spectrometry technique which allows the solution, followed by 300 lL of hexane and 1 mL of 0.02 M separation and the identification of a mixture composed of potassium dihydrogen phosphate neutralizing solution. The piperazines and ATS. For the quantitative method, we have mixture was centrifuged for 1 min at 302 g and the organic chosen to extend to piperazines an HPLC/DAD method, pre- layer analyzed by GCMS. viously implemented for the quantification of amphetamine 43 The combined silylation–acylation derivatization procedure type stimulants. The method was validated following an was performed by adding 100 lL of the upper layers of each approach using accuracy profiles based on b-expectation acylated solution to its silylated counterpart. tolerance intervals for total error measurement.

2.3. Calibrators and validation standards preparation for the The b risk for the confidence interval has been set to 5%, validation of the HPLC/DAD quantification method and the ±k acceptance limits to 10%. Accuracy profiles and validation parameters have been computed with e.noval soft- Stock solutions of amphetamines, piperazines, internal stan- ware V2.0 (Arlenda, Lie` ge, Belgium). dard and lactose at a concentration of 1 mg/mL were prepared The pre-validation step of the software, allows choosing the in methanol/water 50:50 (v/v). mathematical model that is closest to the response function. Subsequent working solutions were prepared by proper dilu- This protocol requires the preparation of two kinds of sam- tions of the stock solutions with methanol/water 50:50 (v/v). The ples in two independent ways: calibration standards and vali- calibration standard solutions were prepared from a different dation standards, as described above. stock solution batch than that of the validation standards. The experimental plan consisted of three series (reproduc- Each operator prepared his own piperazine derivative mix ibility conditions) of four repetitions (repeatability conditions) calibration standards, by dilution of the stock solutions with of analyses. methanol/water 50:50 (v/v) and 1 mg/mL internal standard (pTP) to obtain eight calibration standards of 2, 10, 20, 30, 2.5. GC/MS qualitative analysis 40, 50, 80 and 100 mg/L (m = 8). This operation was repeated on three different days, by three different operators. GC/MS analyses were performed using an HP 5890 series II The validation standards, were prepared once, and analyzed gas chromatograph (Hewlett–Packard, Agilent, Massy, by each operator, with the calibration standards he had pre- France), equipped with a 5971A mass selective detector (Hew- pared. For routine work, tablets from the seized material were lett–Packard, Agilent, Massy, France). A 5% phenyl-meth- crushed in a mortar and a pestle and 50 mg of the resulting ylsiloxane (HP-5MS) capillary column (30 m · 0.25 mm I.D., powder, were weighed on a precision scale (e =10À4 g) and 0.25 lm film thickness), supplied by Hewlett–Packard (Agi- subsequently dissolved in 50 mL of the methanol/water 50:50 lent, Massy, France), was used. (v/v) mixture. The solution was mixed on vortex, sonicated Chromatographic conditions were as follows: initial col- for 15 min and centrifuged for 1 min at 302 g. One hundred umn temperature was 90 C for 0.5 min, increased by 20 C/ (100) lL of the prepared solution was added to 100 lL of inter- min to 200 C, then to 280 Cat15C/min and finally to nal standard solution (1 mg/mL), and then filled up to 1 mL 320 Cat20C/min for 3.67 min. The temperatures of the volume by adding methanol/water 50:50 (v/v) mixture solvent. injection port and detector were set at 280 and 320 C, respec- Lactose is a substance known to be present in tablets (filling tively. Injection of 1 lL of the sample solution was performed agent). Lactose has been added to the validation standards in automatically in splitless mode, and helium (carrier gas) at a order to mimic the matrix effect that can potentially be seen flow rate of 1.0 mL/min. The mass spectrometer was operated in the analysis of tablets. This is why the stock solutions were with a filament current of 300 lA and electron energy of 70 eV mixed with the lactose solution and 1 mg/mL internal standard in the electron ionization (EI) mode. Positive ions were ana- (pTP) to obtain eleven validation standards of 2, 10, 20, 30, 40, lyzed. Acquisition was carried out in the scan mode, and the 50, 60, 70, 80, 90 and 100 mg/L (m = 11). entire mass range from 38 to 650 m/z was collected. Four replications of each validation standard were performed (n = 4). This operation was repeated on three different 2.6. HPLC/DAD quantitative analysis days (p = 3), by three different operators. Quantitative analysis was performed by high-performance 2.4. Validation software and principles of the quantification liquid chromatography, using a liquid chromatograph method equipped with a spectrophotometric diode-array detector (HPLC-DAD). The study was conducted on a Hewlett–Packard In accordance to ISO/CEI 17025:2005 and the guidelines of 1050 series HPLC. Chromatographic separation was the French standard NF V 03–110,44 the present method was carried out on a Thermo Hypersil C18 analytical column fully validated using the total error approach.45–47 (3 lm · 125 mm · 3 mm i.d.), protected by a Thermo BDS

Figure 1 GCMS Total Ion Chromatograms of piperazine derivative mixture: (A) without derivatization (1: BZP, 2: TFMPP, 3: mCPP, 4: MeOPP and 5: MDBP); (B) after silylation (1: BZP-TMS, 2: TFMPP-TMS, 3: mCPP-TMS, 4: MeOPP-TMS and 5: MDBP-TMS); (C) after acylation (1: BZP-HFB, 2: TFMPP-HFB, 3: mCPP-HFB, 4: MeOPP-HFB and 5: MDBP-HFB).

Table 1 Retention times (RT), mass spectra (m/z) and the limits of detection (LODs) for ﬁve piperazine derivatives. Substance RT (min) m/z LOD (lg/mL) Without derivatization BZP 4.454 91;56 ;134 ;176 100.00 TFMPP 4.647 188;56;95;172;145;230 93.50 mCPP 6.045 154;56;75;11 ;138;196 71.50 MeOPP 6.143 150;56;92;120;135;192 73.00 MDBP 6.866 135;56;85;178;164;220 116.00 After silylation BZP-TMS 5.733 102;59;116;157;233;248 0.70 TFMPP-TMS 5.905 302;59;73;101;128;173 0.75 mCPP-TMS 7.283 128;59;73;101;;226;268 0.80 MeOPP-TMS 7.370 264;59;73;101;135;162 0.80 MDBP-TMS 8.039 135;59;73;102;;157;292 1.00 After acylation BZP-HFB 6.147 91;56;146;175;281;372 0.50 TFMPP-HFB 6.156 200;56;69;145;173;426 0.60 mCPP-HFB 7.446 392;56;69;139;166;195 0.60 MeOPP-HFB 7.525 388;56;69;135;191 0.80 MDBP-HFB 8.364 135;56;77;281;416 0.90

Figure 2 GCMS total ion chromatograms of amphetamine derivative mixture: (A) after silylation (1: Ephedrine-TMS, 2: Ephedrine- 2TMS, 3: Amphetamine-TMS, 4: Methamphetamine-TMS, 5: Ephedrine-2TMS and 6: MBDB-TMS); (B) after acylation (Amphetamine- HFB, 2: Methamphetamine-HFB, 3: Ephedrine-HFB, 4: Pseudoephedrine-HFB, 5: MDA-HFB, 6: MDMA-HFB, 7: MDEA-HFB and 8: MBDB-HFB).

Figure 3 GCMS total ion chromatograms of a mixture containing piperazine and amphetamine derivatives: (A) after silylation (1: Ephedrine-TMS, 2: Ephedrine-2TMS, 3: Amphetamine-TMS, 4: BZP-TMS, 5: TFMPP-TMS, 6: Methamphetamine-TMS, 7: Ephedrine- 2TMS, 8: MBDB-TMS, 9: mCPP-TMS, 10: MeOPP-TMS and 11: MDBP-TMS); (B) after acylation (1: Amphetamine-HFB, 2: Methamphetamine-HFB, 3: Ephedrine-HFB, 4: Pseudoephedrine-HFB, 5: MDA-HFB, 6: BZP-HFB, 7: TFMPP-HFB, 8: MDMA-HFB, 9: MDEA-HFB, 10: MBDB-HFB, 11: mCPP-HFB, 12: MeOPP-HFB and 13: MDBP-HFB).

C18 guard column, in reversed-phase mode, with a mobile The gradient applied was increased from 20% to 40% B phase gradient. The components of the mobile phase were: within 12 min, starting from 0 min after injection, then KH2PO4 buffer, 20 mM, adjusted to pH 3.2 with 85% ortho- increased to 70% B in 2 min, and kept at 70% B for 4 min. phosphoric acid (A) and acetonitrile (B). The ﬂow rate of the Finally, the gradient was raised to 97% B from 18 to 19 min. mobile phase was 0.4 mL/min. The total analysis time was The column temperature during the analysis run was main- 19 min. tained at 30 C. The volume of injection was 10 lL. The HP

Figure 4 GCMS total ion chromatogram of a mixture containing piperazines and amphetamines after a combined silylation and acylation: (1) Amphetamine-HFB, (2) Ephedrine-TMS, (3) Methamphetamine-HFB, (4) Ephedrine-HFB, (5) Pseudoephedrine-HFB, (6) Ephedrine-2TMS, (7) Ephedrine-2TMS, (8) Pseudoephedrine-HFB-TMS, (9) Amphetamine-TMS, (10) MDA-HFB, (11) BZP-TMS, (12) TFMPP-TMS, (13) Methamphetamine-TMS, (14) BZP-HFB, (15) MDMA-HFB, (16) MDEA-HFB, (17) MBDB-TMS, (18) MBDB- HFB, (19) mCPP-TMS, (20) MeOPP-TMS, (21) mCPP-HFB, (22) MeOPP-HFB and (23) MDBP-HFB.

Figure 5 HPLC-UV/DAD chromatograms of piperazine derivative mixture: (A) Piperazine derivative mixture with internal standard P- TP, detection at 210.4 nm [1: BZP; 2: MDBP; 3: MeOPP; 4: P-TP; 5: mCPP and 6: TFMPP]; (B) Piperazines mixed with amphetamines, detection at 230.4 nm [1: BZP; 2: MDBP; 3: MeOPP; 4: MDMA; 5: mCPP and 6: TFMPP].

Table 2 Integration criteria of the piperazine derivatives. BZP BZP MDBP MeOPP mCPP TFMPP Dosing range (mg/L) 2–30 30–100 2–100 10–100 10–100 10–100 Wave length (nm) 210.4 254.4 286.4 230.4 210.4 210.2 Criteria of integration Height Area Area Area Area Area

UV-DAD detector operates with a time response of 1 s, peak width taken >0.05 min and spectra of substances were recorded from 210 to 400 nm.

3. Results

3.1. GC/MS qualitative analysis

In the applied GC/MS conditions, a good chromatographic separation for the silylated piperazines was obtained as shown in Fig. 1B. This was not the case for the acylated ones, where a total co-elution between BZP and TFMPP was observed (Fig. 1C). The derivatizing agents, acting on different sites (a mobile hydrogen for the silylation reaction, primary amine moiety for acylation) generate chemical structures that can easily be differentiated by mass spectrometry. Mass spectra are very specific, and allow unambiguous identification on the basis of major ions with m/z values given in Table 1. The limits of detection (LODs) for the GC/MS method were determined. A concentration causing a detector signal three times greater than the noise (S/N = 3) was accepted as the lower LOD. The LLODs for these piperazines were in the concentration range of 0.5–1.0 mg/L after acylation or silylation and 71.5–100 mg/L without derivatization. The acylated amphetamines were well separated and identified (Fig. 2B), whereas, this was not the case with the silylated ones (Fig. 2A). The mixture containing the studied piperazine and amphetamine derivatives is then well identified after HFBA-derivatization and the target components are well sep-

0.915 –44.03 – – – – 167.50 – 287.90 – – 17.39 – 0.878 – –arated – 175.60 except 68.88 – – those – of – BZP and TFMPP as shown in Fig. 3B. À À For this reason, a combined silylation–acylation derivatization protocol was carried out and the results show a better identiﬁ- cation and a total separation of all the target components as shown in Fig. 4.

3.2. HPLC/DAD validated method for quantitative analysis

Typical HPLC/DAD chromatograms of the piperazine derivatives and of the piperazines mixed with amphetamine derivatives are shown in Fig. 5. The achieved separation was satisfactory in both analyses with a small co-elution between BZP and MDBP. The ﬁve piperazine derivatives are well separated from the amphetamines, especially from MDMA which is the most encountered substance in an admixture with the piperazine derivatives. 29.83 – – 51.51 184.90 – 0.973 – – – – – Because of the BZP-MDBP co-elution, the quantiﬁcation 1.000 – – 0.999 1.000 – 0.999 – – 0.999 1.000 – 0.999 – – 1.000 1.000 – À BZP BZPÀ MDBP MeOPP mCPP TFMPP BZP BZP MDBP MeOPP mCPP TFMPP BZP BZP MDBP MeOPP mCPP TFMPP of BZP has to be carried out for lower concentrations by considering the peak height at the absorption maximum = 1) Day1 Day2 Day3

n (210.4 nm), then by considering the area for upper concentrations, at 254.4 nm (Table 2). However, these two substances =8, ). are rarely simultaneously present in the same tablet. m m ). k ). n

=3, 3.3. Method validation k

A weighted (1/X2) quadratic regression model was selected for the quantitative determination of BZP from 2 to 30 mg/L and Validation results (response function) for the HPLC/DAD quantitative determination of piperazine derivatives. a linear regression through 0 ﬁtted with the level 100 model, from 30 to 100 mg/L. A linear regression through 0 ﬁtted with the level 100 was selected for both MBDP and TFMPP from 2 2 Table 3 Response function ( Number of replicates ( Number of amount levels ( Number of series (Day) ( SlopeIntercept r 109.60 34.16 47.37 139.50 391.00 157.40 105.00 34.16 47.70 140.00 394.10 157.60 94.54 36.50 47.56 137.80 394.10 157.10 Quadratic term to 100 mg/L. For mCPP and MeOPP, a linear regression

Table 4 Validation results (trueness, precision, and accuracy) for the HPLC/DAD quantitative determination of piperazine derivatives. BZP BZP MDBP MeOPP mCPP TFMPP BZP BZP MDBP MeOPP mCPP TFMPP Trueness (k = 3, m = 11, n = 4) Relative bias (%) Recovery (%) 2 À0.6 – À6.3 – – 6.6 99.4 – 93.7 – – 106.6 10 À1.0 – À3.0 1.5 1.2 2.3 99.0 – 97.0 101.5 101.2 102.3

20 0.2 – À0.4 1.4 1.3 1.7 100.2 – 99.6 101.4 101.3 101.7 30 À0.1 À1.1 4.4 2.9 2.0 3.0 99.9 98.9 104.4 102.9 102.0 103.0 40 – À0.2 2.7 2.9 2.2 2.9 – 99.8 102.7 102.9 102.2 102.9 50 – 1.5 1.0 1.2 0.9 13 – 98.5 101.0 101.2 100.9 101.3 60 – À0.9 2.0 1.7 1.8 2.1 – 99.0 102.0 101.7 101.8 102.1 70 – À1.0 1.5 2.1 1.4 2.0 – 99.0 101.5 102.1 101.4 102.0 80 – 1.3 1.4 1.2 1.1 1.6 – 98.8 101.4 101.2 101.1 101.6 90 – À1.6 1.5 1.1 0.8 1.6 – 98.4 101.5 101.1 100.8 101.6 100 – À2.3 1.6 2.4 0.5 1.6 – 97.7 101.6 102.4 100.5 101.6 Precision (k = 3, m = 11, n = 4) Repeatability (RSD,%) Intermediate precision (RSD,%) 2 5.2 – 4.7 – – 4.3 7.2 – 5.9 – – 4.3 10 1.6 – 2.0 2.1 1.5 1.6 3.0 – 3.8 2.2 2.2 1.6 20 3.7 – 1.9 1.9 1.3 1.6 3.7 – 3.5 1.9 1.3 1.6 30 1.7 1.3 2.1 1.5 1.6 17 3.4 1.5 2.0 1.9 1.9 1.7 40 – 0.5 0.6 0.6 0.4 0.4 – 1.3 2.0 1.0 1.2 1.1 50 – 0.9 0.7 1.0 0.7 0.5 – 0.9 2.0 1.6 1.3 1.2 60 – 0.8 0.9 1.0 0.7 0.7 – 1.0 1.3 1.0 1.0 1.1 70 – 0.7 0.7 1.6 0.6 0.7 – 1.0 1.4 1.6 0.8 1.2 80 – 0.7 0.4 0.5 0.6 0.7 – 0.7 1.5 1.2 1.3 1.3 90 – 0.8 0.4 0.6 0.6 0.6 – 1.2 1.8 1.4 1.4 1.8 100 – 0.8 0.6 0.6 0.7 0.8 – 0.8 1.4 1.2 1.0 1.5 Accuracy (k =3,m = 11, n =4) b-expectation tolerance limits (ng) BZP BZP MDBP MeOPP mCPP TFMPP 2 [01.7, 02.3] – [01.6, 02.1] – – [01.9, 02.3] 10 [09.1, 10.6] – [08.7, 10.6] [09.8, 10.7] [09.6, 10.6] [09.8, 10.4] 20 [18.6, 21.4] – [18.2, 21.7] [19.7, 21.2] [19.9, 20.8] [19.6, 20.8] 30 [27.2, 32.7] [28.8, 30.5] [29.4, 31.7] [30.0, 32.3] [29.6, 32.0] [29.7, 31.6] 40 – [38.4, 41.4] [38.4, 43.6] [40.5, 42.5] [39.7, 42.7] [39.5, 42.2] 50 – [48.4, 50.1] [47.4, 53.3] [49.1, 53.1] [49.1, 52.5] [48.6, 52.1] 60 – [58.1, 60.8] [59.1, 62.8] [60.5, 62.7] [60.2, 62.9] [59.2, 62.5] 70 – [67.5, 71.1] [68.3, 73.4] [70.1, 74.2] [70.4, 72.8] [68.7, 73.0] 80 – [77.9, 80.1] [77.2, 84.7] [79.3, 84.4] [78.5, 84.3] [77.9, 83.5] 90 – [85.3, 91.7] [85.9, 96.3] [88.4, 95.6] [87.8, 95.0] [86.1, 95.4] 100 – [96.1, 99.3] [97.5, 105.2] [99.4, 105.4] [99.0, 103.7] [96.8, 104.9] Number of replicates (n). Number of amount levels (m). Number of series (Day) (k). Relative standard deviation (RSD).

Table 5 Validation results (linearity, limits of quantification) for the quantitative determination of piperazine derivatives. BZP BZP MDBP MeOPP mCPP TFMPP Linearity (k = 3,m = 11, n = 4) Range (ng) 8.545–30 30–100 17.54–100 10–100 10–100 5.95–100 Slope 1.0010 0.9746 1.0170 1.0160 1.0060 1.0140 Intercept 0.04276 0.74460 0.04606 0.09157 0.3459 0.2102 r2 0.9970 0.9991 0.9992 0.9992 0.9994 0.9994 Limits of quantification (k = 3,m = 11, n = 4) Lower limit of quantification (lg/mL) 8.545 30 17.54 10.10 10 5.95 Upper limit of quantification (lg/mL) 30 100 100 100 100 100 Number of replicates (n). Number of amount levels (m). Number of series (Day) (k).

4. Discussion expectation tolerance limits builds up the accuracy profile of the quantification method. The use of accuracy profiles for the validation based on total On a practical point of view, the validation decision is error concept is compliant with the fundamental demands of based on the graphical comparison of the accuracy profile the ISO/CEI 17025 standard, for example validation and obtained previously with the acceptance limits ±k that are determination of measurement uncertainty. This method is set prior to the beginning of the experiment. The acceptance based on the determination of confidence interval of measure- limits correspond to the maximal uncertainty of measurement ments, with a risk b. This interval, also called b expectation that is accepted for the results obtained by the analytical interval, is computed from accuracy and intermediate preci- method. These limits have to be set according to the character- sion. For each concentration level included in the experiment, istic constraints of the domain or activity in which the dosing the standard deviation of intermediate precision is calculated will be implemented. For example, in forensic toxicology the by adding the repeatability variance to the inter-series variance limit is ±k = 40%. For the analysis of illicit substances in (S2 = S2 + S2 ). The graphical representation of these b seized material described in this case, we have chosen ±k of IF R IS 10%. The overall performance of the method described in this

Figure 6 Accuracy profiles of piperazine derivatives (amount in mg/L): (1) Accuracy profile of BZP from 02 to 30 mg/L using a weighted (1/X2) quadratic regression model; (2) Accuracy profile of BZP from 30 to 100 mg/L using a linear regression through 0 fitted with the level 100 only model; (3) Accuracy profile of MDBP from 02 to 100 mg/L using a linear regression through 0 fitted with the level 100 only model; (4) Accuracy profile of MeOPP from 10 to 100 mg/L using a linear regression model; (5) Accuracy profile of mCPP from 10 to 100 mg/L using a linear regression model; (6) Accuracy profile of TFMPP from 02 to 100 mg/L using a linear regression through 0 fitted with the level 100 only model. The plain line is the relative bias, the dashed lines are the b-expectation tolerance limits and the dotted lines represent the acceptance limits (10%). The dots represent the relative back-calculated concentrations and are plotted with respect to their targeted concentration.

Please cite this article in press as: Boumrah Y et al. Development of a targeted GC/MS screening method and validation of an HPLC/DAD quantification method for piperazines–amphetamines mixtures in seized material, Egypt J Forensic Sci (2014), http://dx.doi.org/10.1016/j.ejfs.2014.03.002 Development and validation of methods for piperazine–amphetamine mixture analysis 9 paper is much better than expected, as can be seen in Fig. 6 for spectrophotometry detector and gas chromatography-mass MeOPP and mCPP. Moreover, during the pre-validation step, spectrometry. Talanta 2009;77(4):1245–72. the best mathematical model of response function which is 7. Lanaro R, Costa JL, Zanolli-Filho LA, Cazenave SOS. Identifi- chosen is the one which gives the best accuracy index, or the cacß aõquı´mica da clorofenilpiperazina (CPP) emcomprimidosa- preendidos. Quim Nova 2010;33:725–9. best analytical performance in terms of LOQ and uncertainty 8. Cohen BMZ, Butler R. BZP-party pills: a review of research on measurement. benzylpiperazine as recreational drug. Int J Drug Policy 2011;22:95–101. 5. Conclusions 9. Gee P, Richardson S, Woltersdorf W, Moore G. Toxic effects of BZP-based herbal party pills in humans: a prospective study in Christchurch, New Zealand. NZ Med J 2005;118:1784–94. The GC/MS method developed in this work has allowed the 10. Drug Enforcement Administration Office of Diversion Control separation and the identification of a large number of pipera- Drug & Chemical Evaluation Section. 1-3-(Trifluoro-methyl)- zine and amphetamine type substances. The HPLC/DAD quan- phenyl piperazine (Street Names: ‘‘TFMPP’’ or ‘‘Molly’’. Often titative method is fully validated following the concept of total found in combination with BZP: ‘‘A2’’, ‘‘Legal E’’ or ‘‘Legal X’’). error, by means of accuracy profiles. This graphical approach Available at: ; 2013 [accessed 01.09.13]. reproducibility, and eventually to easily set LOQ and dosing 11. Yarosh HL, Katz EB, Coop A, Fantegrossi WE. MDMA-like range. The quantitative method has shown to be linear within behavioral effects of N-substituted piperazines in the mouse. Pharmacol Biochem Behav 2007;88:18–27. the adopted ranges for all the studied piperazines. Furthermore, 12. De Boer D, Bosman IJ, Hidve´gi E, Manzoni C, Benko AA, dos both qualitative and quantitative methods can be useful to lab- Reys LAL, et al. Piperazine-like compounds: a new group of oratories for performing routine analysis of seized material con- designer drugs-of abuse on the European market. Forensic Sci Int taining piperazine and amphetamine type substances. 2001;121:47–56. 13. European monitoring centre for drugs and drug addiction. Annual Funding report on the state of drugs problem in Europe. Available at: ; 2012 accessed 10.01.13. None. 14. U.S. Department of Justice-Drug Enforcement Administration. Microgram Bull 2007;40(9):91. Conflict of interest 15. U.S. Department of Justice-Drug Enforcement Administration. Microgram Bull 2009;42(4):37. 16. U.S. Department of Justice-Drug Enforcement Administration. None declared. Microgram Bull 2008;41(5):48. 17. U.S. Department of Justice-Drug Enforcement Administration. Ethical approval Microgram Bull 2008;41(2):18. 18. U.S. Department of Justice-Drug Enforcement Administration. Microgram Bull 2008;41(1):1–4. Necessary ethical approval was obtained from the institute 19. United States Department of Justice/Drug enforcement adminis- ethics committee. tration, system to retrieve information from drug evidence II (STRIDE), 2003. Acknowledgements 20. Cuddy L. Common drugs of abuse–Part II. J Pract Nurse 2004;54(1):25–31. The author would like to thank Aziz Khiari and Salah Haroun 21. Maurer HH, Kraemer T, Springer D, Staack RF. Chemistry, for their helpful comments on an early draft of this pharmacology, toxicology and hepatic metabolism of designer manuscript. drugs of the amphetamine (ecstasy), piperazine, and pyrrolidin- ophenone types. Drug Monit 2004;26:127–31. 22. Wikstrom M, Holmgren P, Ahlner J. A2 (N-benzylpiperazine) a References new drug of abuse in Sweden. J Anal Toxicol 2004;28:67–70. 23. U.S. Department of Justice-Drug Enforcement Administration. 1. Kovaleva J, Devuyst E, De Paepe P, Verstraete A. Acute Microgram Bull 2003;36(6):118–9. chlorophenylpiperazine overdose: a case report and review of the 24. Arbo MD, Bastos ML, Carmo HF. Piperazine compounds as literature. Ther Drug Monit 2008;30:394. drugs of abuse. Drug Alcohol Depend 2012;122:174–85. 2. Bossong MG, Brunt TM, Van Dijk JP, Rigter SM, Hoek J, 25. United Nations of Drugs and Crime. World Drug Report. Goldschmidt HMJ, et al. MCPP: an undesired addition to the Available at: ; 2013 accessed 3. Davies S, Wood DM, Smith G, Button J, Ramsey J, Archer R, 09.08.13. et al. Purchasing ‘legal highs’ on the Internet––is there consistency 26. EMCDDA–Europol New drugs in Europe, 2012. EMCDDA– in what you get? QJMed2010;103:489. Europol 2012 annual report on the implementation of council 4. Gee P, Richardson S, Woltersdorf W, Moore G. Toxic effects of decision 2005/387/JHA. Available at: ; 2012 Christchurch. NZ Med J 2005;118:1784. accessed 11.12.12. 5. Sheridan J, Butler R, Wilkins C, Russel B. Legal piperazine- 27. U.S. Department of Justice-Drug Enforcement Administration. containing party pills–a new trend in substance misuse. Drug Microgram Bull 2009;42(6):53. Alcohol Rev 2007;26:335. 28. U.S. Department of Justice-Drug Enforcement Administration. 6. Takahashi M, Nagashima M, Suzuki J, Seto T, Yasuda I, Yoshida Microgram Bull 2009;42(7):64. T. Creation and application of psychoactive designer drugs data 29. U.S. Department of Justice-Drug Enforcement Administration. library using liquid chromatography with photodiode array Microgram Bull 2009;42(8):68–70.

30. U.S. Department of Justice-Drug Enforcement Administration. 42. Moreno IED, da Fonseca BM, Magalha˜es AR, Geraldes VS, Microgram Bull 2009;42(5):45–9. Queiroz JA, Barroso M, et al. Rapid determination of piperazine- 31. U.S. Department of Justice-Drug Enforcement Administration. type stimulants in human urine by microextraction in packed Microgram Bull 2009;42(2):16–7. sorbent after method optimization using a multivariate approach. 32. U.S. Department of Justice-Drug Enforcement Administration. J Chromatogr A 2012;1222:116–20. Microgram Bull 2007;40(4):39. 43. Sadeghipour F, Giroud C, Rivier L, Veuthey JL. Rapid determi- 33. Perrin M, Lecompte Y, Roussel O, Le Boisselier R, Bourgine J, nation of amphetamines by high-performance liquid chromatog- Coquerel A. Nouvelles drogues in Traite´de toxicologie me´dico- raphy with UV detection. J Chromatogr A 1997;761:71–8. judiciaire. Paris, 2012, (ISBN 978-2-994-71561-7) p. 517–585. 44. Norme Française: NF V 03–110. Analysis of agri-foodstuffs –– 34. Sheridan J, Butler R, Wilkins C, Russell B. Legal piperazine- Protocol of characterization for the validation of a quantitative containing party pills – a new trend in substance misuse. Drug method of analysis by construction of an accuracy profile, 2010. Alcohol Rev 2007;26:335–43. 45. Viswanathan CT, Bansal S, Booth B, DeStefano AJ, Rose MJ. 35. Bishop SC, McCord BR, Gratz SR, Loeliger JR, Witkowski MR. Workshop/conference report –– quantitative bioanalytical meth- Simultaneous separation of different types of amphetamine and ods validation and implementation: best practices for chromato- piperazine designer drugs by capillary electrophoresis with a chiral graphic and ligand binding assays. AAPS J. 2007;9:30–42. selector. J Forensic Sci 2005;50:326–35. 46. Hubert P, Nguyen-Huu JJ, Boulanger B, Chapuzet E, Cohen N, 36. Staack RF, Fritschi G, Maurer HH. Studies on the metabolism Compagnon PA, et al. Harmonization of strategies for the and toxicological detection of the new designer drug N-benzylpi- validation of quantitative analytical procedures. A SFSTP pro- perazine in urine using gas chromatography-mass spectrometry. J. posal–part III. J Pharm Biomed Anal 2007;45(1):82–96. Chromatogr B 2002;773(1):35–46. 47. Hubert P, Nguyen-Huu JJ, Boulanger B, Chapuzet E, Chiap P, 37. Peters FT, Schaefer S, Staack RF, Kraemer T, Maurer HH. Cohen N, et al. Harmonization of strategies for the validation of Screening for and validated quantification of amphetamines and quantitative analytical procedures. A SFSTP proposal–part II. J of amphetamine and piperazine-derived designer drugs in human Pharm Biomed Anal 2007;45(1):70–81. blood plasma by gas chromatography/mass spectrometry. J Mass 48. Hubert P, Nguyen-Huu JJ, Boulanger B, Chapuzet E, Chiap P, Spectrom 2003;38:659–76. Cohen N, et al. Validation des proce´dures analytiques quantita- 38. Barroso M, Costa S, Dias M, Vieira DN, Queiroz JA, Lopez- tives. Harmonisation des de´marches. STP Pharma Pratiques Rivadulla M. Analysis of phenylpiperazine-like stimulants in 2003;13:101–38. human hair as trimethylsilyl derivatives by gas chromatography– 49. Food and Drug Administration: US Department of Health and mass spectrometry. J. Chromatogr A 2010;1217:6274–80. Human Services. Guidance for Industry, Bioanalytical Method 39. Navaneeswari R, Raveendra RP. Analytical method for pipera- Validation, FDA. Available at: ; 2001 accessed 19.03.13. derivatization and HPLC-UV. J Chem Pharm Res 2012;4:2854–9. 50. Shah VP, Midha KK, Dighe S, McGilneray IJ, Skelly JP, Yakobi 40. Tsutsumi H, Katagi M, Miki A, Shima N, Kamata T, Nishikaw, A, et al. Analytical methods validation: bioavailability, bioequiv- et al. Development of simultaneous gas chromatography mass alence, and pharmacokinetic studies. J Pharm Sci 1992;81:309–12. spectrometric and liquid chromatography–electrospray ionization 51. Hubert Ph, Chiap P, Crommen J, Boulanger B, Chapuzet E, mass spectrometric determination method for the new designer Mercier N, et al. The SFSTP guide on the validation of drugs Nbenzylpiperazine (BZP), 1-(3-trifluoromethylphenyl)piper- chromatographic methods for drug bioanalysis: from the Wash- azine (TFMPP) and their main metabolites in urine. J Chromatogr ington Conference to the laboratory. Anal Chim Acta B 2005;819:315–22. 1999;391:135–48. 41. Vorce SP, Holler JM, Levine B, Past MR. Detection of 1- 52. ISO, 5725–1. Application of statistics. Accuracy (trueness and benzylpiperazine and 1-(3-trifluoromethylphenyl)-piperazine in precision) of measurement methods and results––part 1: general urine analysis specimens using GC–MS and LC–ESI-MS. J Anal principles and definitions. International Organization for Stan- Toxicol 2008;32:444–50. dardization, Geneva 1994.