Analysis of amphetaminic drug compounds in urine by headspace-dielectric barrier discharge ionization-mass spectrometry

2.1 Reagents and materials

Amphetamine, methamphetamine, 3,4-methylenedioxyamphetamine (MDA) and 3,4-methylenedioxymethamphetamine (MDMA) were collected from Ningbo Police Department, Zhejiang Province (P.R. China) for the preparation of standard solution. The standard solutions were then used for spiking in urine. Sodium hydroxide, sodium chloride, sodium carbonate, potassium hydroxide, potassium chloride, potassium carbonate and aqueous ammonia (28–30%) were purchased from Sigma-Aldrich, Shanghai, China. Glass vials with rubber septum (5 mL) for the headspace were purchased from Thermo Fisher Scientific.

2.2 Sample preparation

Stock solutions of the drug compounds were separately prepared in water and stored in a refrigerator at −10 °C. Concentration of amphetamine and methamphetamine was 10 µg/mL and that for MDA and MDMA was 20 µg/mL. Working standard solutions were prepared just before the experiments and dilutions were made by using water. Human urine was spiked with AM, MA, MDA, MDMA and internal standard of each drug compound separately. For dilution, the urine spiked solution was diluted with fresh urine. 10 µL of 100 ng/mL (total 1 ng) of each drug compound was poured into a 5-mL headspace vial and alkalinized with 100 µL ammoniated alkali solution (4 M alkali solution + ammonia solution (28%) = 85:15) and then the vial was immediately hermetically sealed by cap. The solution was mixed well through vortex and kept for 6–7 min in order to release free base amphetaminic amine compounds from the solution phase completely. After 6–7 min, the vial was opened in front of the MS inlet where previously dielectric barrier discharge ionization (DBDI) source was running condition (Fig. 1). Experiments with urine reported in this paper were performed using human urine.

Close

Fig. 1 Schematic diagram of the experimental set up of headspace-DBDI-MS.

Export to PPT

2.3 Experimental setup

A home-made DBD ion-source was used for the ionization of the vaporized amphetaminic molecules. The ion source was placed in front of the mass spectrometer inlet in such a way that the vaporized molecules are ionized under open atmosphere before entering to the first vacuum chamber of the mass spectrometer. The detailed experimental setup is illustrated in Fig. 1. Briefly, the DBD ion-source was made of ceramic tube with 3.0 mm o.d. and 1.5 mm i.d. Copper tape was used as an outer electrode and a stainless steel wire was as an inner electrode. 15 kHz radiofrequency (RF) voltages with 3.0 kV (Vp-p) was applied to the outer electrode where the inner electrode was grounded to generate helium plasma with 250 mL min⁻¹ He flow rate. The headspace vial was opened in between of the MS inlet and the DBD ceramic tube. In this ion source, the vaporized amphetaminic molecules were not exposed to the plasma but were ionized mainly by H³O⁺ and its water clusters H³O⁺(H₂O)_n produced outside of the DBD ceramic tube (Habib et al., 2014). The temperature of effusing gas from the DBD ion source was 49 °C. The slightly higher temperature than the room temperature is due to the radiofrequency heating of the dielectric tube.

The ions generated by the DBD ion source were detected using an ion trap mass spectrometer (LTQ XL; Thermo Scientific, San Jose, CA, USA) and the raw data were processed using Xcalibur software (ver. 2.1) (Thermo Fisher Scientific). The temperature of the ion transport tube was maintained at 125 °C.

3.1 Analysis of amphetaminic drug compounds in water

The amphetaminic compounds belong to the group of secondary aliphatic amines and they are either volatile and/or semi-volatile in nature, therefore these convert into hydrochloride form for preservation. The hydrochloride form compounds are water soluble and the free base amines can be attained in gas-phase upon addition of alkali solution because of their considerable vapor pressures at 25 °C: AM, 3.1E-1; MA, 5.4E-3; MDA, 1.0E-3; MDMA, 1.6E-3 mm of Hg. The amphetaminics are amine compounds, thus they show significant ionization efficiency through the formation of their protonated molecular ions, e.g., (M + H)⁺. Proton affinity (PA) of this group of compounds is quite high, for example, PA for MA is 965 kJ/mol and follows the increasing order: AM < MDA < MA < MDMA (Matsumura et al., 2003). The amphetaminic compounds are known as illicit compounds and have been used widely as abuse drugs for a long time, therefore numerous attempts were taken to analyze these compounds in human body fluids, for example, in urine, blood, saliva, etc. in order to doping test, clinical toxicology, forensic analysis, etc. (Tsuchihashi et al., 1991; Katagi et al., 1996; Tatsuno et al., 1996; AI-Dirbashi et al., 1997; Bogusz et al., 1997; Cheung et al., 1997; Marquet et al., 1997; Sato and Mitsui, 1997; Kraemer and Maurer, 1998; Maurer, 1998; Moeller et al., 1998; Ortuno et al., 1999; Talwar et al., 1999; Lee et al., 2000; Nishida et al., 2003; Raikos et al., 2003; Gentili et al., 2004; Meng et al., 2006; Concheiro et al., 2007; Pavlova and Petrovska-Jovanović, 2007; Miranda et al., 2007; Johamsen and Jornil, 2009; Boles and Wells, 2016). Most of the previous studies showed that analysis of the amphetaminic compounds has been performed through derivatization of the amine compounds by headspace-SPME-coupled HPLC/GC/LC-MS methods (Tatsuno et al., 1996; AI-Dirbashi et al., 1997; Marquet et al., 1997; Sato and Mitsui, 1997; Ortuno et al., 1999; Talwar et al., 1999; Lee et al., 2000; Nishida et al., 2003; Raikos et al., 2003; Gentili et al., 2004; Miranda et al., 2007; Johamsen and Jornil, 2009). SPME method has proven its acceptability to analyze a wide range of compounds at ng-level where the target analytes have showed difficult to vaporize from complex matrices and/or poor ionization efficiency. In the present case, headspace method has proven as an efficient desorption method of the target amphetaminic compounds from water or biological sample i.e., human urine upon addition of alkali solution. A home-made DBD ion source was used for direct ionization of the vaporized amphetaminic molecules without any derivatization process and/or adduct formation. The home-made DBD was found to be a suitable ion source as it is a soft atmospheric pressure ionization technique having a capability of ionizing of a wide range of compounds and helium DBD plasma is a field free source of reagents for atmospheric pressure ionization (Habib et al., 2014).

Fig. 2a–d shows the individual mass spectrum of AM, MA, MDA and MDMA. The amphetaminic compounds were vaporized from their standard solution (each amphetaminic compound 1 ng in water) upon addition of an equal volume of ammoniated K₂CO₃ solution and then ionized by the homemade DBD ion source before introducing into the MS. As shown from Fig. 2a–d, the four amphetaminic molecules form protonated molecular ion, (M + H)⁺. MA shows the highest intensity (1.99E5) (M + H)⁺ peak at m/z = 150 which is slightly higher than that for AM (1.38E5) (m/z = 136). This is reasonable because of its higher proton affinity (PA) of MA (965 kJ/mol) than that of AM although the vapor pressure of AM (3.07E-1 mm of Hg) is quite high compared to that of MA (5.40E-3 mm of Hg) (Matsumura et al., 2003), on the other hand, the protonated molecular ion, (M + H)⁺, of MDA (m/z = 180) and MDMA (m/z = 194) are appeared with relatively low intensities (MDA: 4.48E3; MDMA: 1.57E4). Although PA of MDMA is higher than that of MA (965 kJ/mol) and PA for MDA is higher than that of AM but the vapor pressures of these two methylenedioxy derivatives are relatively lower (MDA: 1.0E-3 mm of Hg; MDMA: 1.6E-3 mm of Hg) than those of AM and MA, this could be the reason of lower intensities of their protonated peaks in the mass spectra.

Close

Fig. 2 Mass spectra of (a) AM, (b) MA, (c) MDA and (d) MDMA by using headspace-DBDI-MS. Amount of each drug compound was 1 ng/mL in water as standard and treated by equal volume of ammoniated K₂CO₃ solution [85% K₂CO₃ (4 M) + 15% NH₃ (28%)].

Export to PPT

As shown from Fig. 3, in the simultaneous analysis of the four amphetaminic compounds in water upon addition of ammoniated K₂CO₃ solution (K₂CO₃ (4M): NH₃ (28%) = 85:15) (amount of each compound 1 ng/mL), both the AM and MA have showed dominant protonated molecular ion peak, (M + H)⁺ (relative abundance (RA): 1.58E5), however MDA and MDMA have also showed the (M + H)⁺ peaks but with low intensities (RA for MDA, 3.7E3 and RA for MDMA 6.13E3). The detection sensitivity, alternatively limit of detection (LOD), of the amphetaminic compounds depends on the type of the alkali solution used. Better LOD for the four amphetaminic compounds were found by using carbonate alkali solution i.e., K₂CO₃ and the range of LOD was from 0.60 to 3.00 ng/mL, on the other hand, that was from 2.0 to 5.0 ng/mL by non-carbonate alkali solution (e.g., KOH) (Table 1). The detection sensitivity has also been enhanced by using ammonia solution with the both alkali solution (85% (4 M alkali solution) + 15% (28% NH₃ solution). The enhancement effect by carbonate alkali solution may be due to formation of CO₂ from the reaction of acidic hydrochloride amphetaminic molecules with the alkali and the in-situ formed CO₂ acts as a carrier gas to vaporize the free base amine molecules effectively from the solution phase. The use of ammonia solution also enhances the rate of vaporization of the in-situ formed amphetaminic molecules acting as a carrier gas. The LOD of AM and MA were in the range of 0.10 to 0.80 ng/mL by using ammoniated carbonate alkali solution while the range were 1.00–3.50 ng/mL for ammoniated non-carbonate alkali solution (Table 1). It is surprising to observe poor LOD values by the ammoniated non-carbonate alkali solution (Table 1) although NH₃ act as a carrier gas for the free base amphetaminic molecules. Beside the role of CO₂ as a carrier gas, the in-situ formation of CO₂ plays a crucial role for better LOD by using carbonate-alkali solution. This is because in-situ formation of CO₂ might help to vaporize the free base amphetaminic molecules as in-situ condition and then they come out to gas phase by the CO₂ and/or NH₃. Buffering properties of the carbonate alkali solution may also play a role for better LOD values. Little and/or no salting-out effect on LOD upon addition NaCl was observed. Addition of highly soluble salt like NaCl is responsible to salting-out of volatile and/or semi-volatile compounds in water but the presence of 4 M alkali solution in the amphetaminic solution may hinder the salting-out effect by NaCl for the amine molecules.

Close

Fig. 3 Mass spectra of the mixture of AM, MA, MDA and MDMAby using headspace-DBDI-MS. Amount of each drug compound was 1 ng/mL in water as standard and treated by equal volume of ammoniated K₂CO₃ solution [85% K₂CO₃ (4 M) + 15% NH₃ (28%)].

Export to PPT

3.2 Analysis of amphetaminic drug compounds in urine

Among the body fluids, urine is considered as easily achievable with large volume as non-invasive fluid for doping testing. Urine contains high levels of urea and creatinine, so direct analysis by mass spectrometry becomes difficult because of high proton affinities of these two compounds, e.g. PA of urea is 868.4 kJ/mol which give higher intensities molecular ion peaks, (M + H)⁺, in positive mode ionization (Zheng and Cooks, 2002). However, we found better LODs for the four amphetaminic compounds in urine in the range from 0.04 to 0.40 ng/mL (Table 2) upon addition of ammoniated carbonate alkali solution compared to their standard solution (0.10–0.80 ng/mL) (Table 1). AM and MA showed better LODs compared to MDA and MDMA in urine. LODs for AM and MA were 0.05 and 0.04 ng/mL while for MDA and MDMA were 0.40 and 0.20 ng/mL respectively. Tsuchihashi et al. (1991) reported high LOD values for AM (9.0 ng/mL) and MA (7.0 ng/mL) in urine by headspace chemical ionization coupled GC-MS (Tsuchihashi et al., 1991) while Lee et al. (2000) found 0.06 ng/mL for AM and 0.04 ng/mL for MA in serum by using headspace derivatization HPME-GC-MS method (Lee et al., 2000). Cheng et al. (2017) reported simultaneous detection of AM, MA, MDA and MDMA from 200 ng dried samples deposited on a glass surface by using desorption flame-induced atmospheric pressure chemical ionization mass spectrometry.

In the present experiment, human urine was spiked with the standard four amphetaminic compounds separately and mixed with equal volume of the alkali solution to vaporize the free base amphetaminic molecules. Fig. 4a–d show the mass spectra of AM, MA, MDA and MDMA which were extracted from the urine spiked solution upon addition of equal volume of ammoniated K₂CO₃ solution (amount of each amphetaminic compound was 1 ng/mL). As shown from Fig. 4a–b, AM and MA have showed the protonated molecular ion, (M + H)⁺, as a base peak and a peak at m/z = 86 is appeared from urine. Hušková et al. (2004) and Huang et al. (2016) reported the peak at m/z = 86, (creatinine-CO + H)⁺, originates from fragmentation of creatinine through neutral loss of CO. The strong ion signal at m/z = 86 may suppress the other ions which may also originate from urine e.g., protonated urea and creatinine and their protonated dimers, etc. Relative abundances (RA) of the (M + H)⁺ peaks for AM and MA are 2.25E5 and 3.78E5 respectively, which are 1.63 and 1.90 times higher compared to those for the standard solution (Fig. 2a and b). In the case of MDA and MDMA (Fig. 4c and d), we found about 1.38 times higher relative abundance for the (M + H)⁺ peaks in the urine compared to those in water (Fig. 2c and d). Therefore, it may conclude that in-situ reactions upon addition of alkali solution in urine are responsible for the enhanced detection sensitivity of the amphetaminic compounds. In in-situ reactions, free base amphetaminic molecules and CO₂ form from the reactions between hydrochloride amphetaminic compounds and carbonated alkali molecules. Moreover, amine type compounds of urine (e.g., urea, creatinine, etc.) also generate upon addition of alkali solution. Reactions among the above-mentioned chemical components in in-situ favors releasing of the as-formed semi-volatile free base amphetaminic molecules effectively. This could be the reason for the high sensitivity of the amphetaminic drug compounds by carbonated alkali headspace-DBDI-MS method.

Close

Fig. 4 Mass spectra of (a) AM, (b) MA, (c) MDA and (d) MDMA by using headspace-DBDI-MS. Amount of each drug compound was 1 ng/mL in human urine and treated by equal volume of ammoniated K₂CO₃ solution [85% K₂CO₃ (4M) + 15% NH₃ (28%)].

Export to PPT

Fig. 5 shows mass spectrum of AM, MA, MDA and MDMA in simultaneous analysis in urine, which treated with an equal volume of ammoniated K₂CO₃ (K₂CO₃ (4 M):NH₃ (28%) = 85:15) (amount of each amphetaminic compounds 1 ng/mL). The protonated molecular ion peaks, (M + H)⁺, for AM and MA are dominated in the mass spectrum with RA 3.26E5 while the RA for MDA and MDMA are 4.68E3 and 8.49E3 respectively. In case of standard solution, the (M + H)⁺ ion peaks for AM and MA were also predominant compared to those for MDA and MDMA (Fig. 3). The relative abundance (RA) of the (M + H)⁺ peak for AM and MA in urine 2 times higher than that in standard, while the RA for MDA and MDMA in urine 1.38 times higher than that in standard. As mentioned in the previous section, the detection sensitivity depends on the type of alkali solutions used. Ammoniated carbonate alkali solution provides maximum detection sensitivity alternatively better LOD. Possible explanations for the highest detection sensitivity of the combination of ammoniated carbonate alkali solution of analysis of the amphetaminic drug compounds have also discussed in the preceding section.

Close

Fig. 5 Mass spectra of the mixture of AM, MA, MDA and MDMAby using headspace-DBDI-MS. Amount of each drug compound was 1 ng/mL in human urine and treated by equal volume of ammoniated K₂CO₃ solution [85% K₂CO₃ (4M) + 15% NH₃ (28%)].

Export to PPT

3.3 Method validation

In order to find out the quantitative abilities of the present method and its practical application in the real world, several factors, for instance, limit of detection (LOD), precision, linear range, correlation co-efficient of linearity (R²), and analyte recovery rate were taken into account. Peak intensities were used for the evaluation of quantitative parameters. At LOD level, each of the analyte showed a signal to noise ratio at least 3(S/N = 3). Surprisingly, at least one order of magnitude lower LODs were realized for urine samples compared to the corresponding standard solutions (Tables 1 and 2). For this better sensitivity, one of the possible reasons could be in-situ reactions between the amphetaminic hydrochloride compounds and/or urine components with alkali which favors to vaporize the as-formed free base amine molecules from the solution phase. The in-situ formation of CO₂ and volatile compounds from urine upon addition of carbonated alkali solution could be another reason for better LODs in urine. The precision of the present method is expressed in terms of percentage of relative standard deviation (% RSD). The % RSD were calculated at the limit of quantitation (LOQ) of the analytes, where LOQ concentration was selected as three times of LOD level for each amphetaminic compound. The three replicate analyses showed 14% RSD values for standard amphetaminic compounds solutions whereas for urine spiked these were up to 7.89%. The in-situ reactions in urine and formation of CO₂ and/or volatile components from urine could be the lower % RSD values for urine samples compared to the corresponding standard samples. Linearity of the present method was evaluated in a concentration range of 0.04–20 ng/mL for AM and MA and 0.20–30 ng/mL for MDA and MDMA. A series of solutions were prepared for each amphetaminic compound using an internal standard to make calibration curve. A fix concentration (5 ng/mL) of internal standard was spiked in each solution and the target compound’s concentration was varied. The linear calibration curves were generated by plotting the response factor against the concentration of target amphetaminic compounds where the response factor was calculated as the ratio of intensities of target compound and internal standard at each concentration level. The 9-points calibration curves showed a good linearity for all the amphetaminic compounds with correlation co-efficients, R² = 0.999 (Fig. S1, Supporting Information). Recovery of the present method was determined by spiking a fix concentration of standard solutions (10 ng/mL) and internal standards (5 ng/mL) to the urine and calculated the amount from standard calibration curves. The percentage of recoveries were quite high (from 94% to 99%) for the studied compounds from urine (Table 2). The high recovery rates are reasonable because of in-situ reactions between urine and alkali solutions which favors the vaporization of the free base amine molecules, and generation of CO₂ and other volatile compounds from urine and/or NH₃ act as carrier gas for the free base amphetaminic compounds.