Effective GC-MS procedure for detecting iso-LSD in urine after base-catalyzed conversion to LSD

¹ Division of Forensic Toxicology, Office of the Armed Forces Medical Examiner, Armed Forces Institute of Pathology, Rockville, MD 20850.
^a Address correspondence to this author at: Forensic Toxicology, AFIP Annex, 1413 Research Blvd., Rockville, MD 20850. Fax 301-319-0628; e-mail paul{at}email.afip.osd.mil.

	Abstract

Top Abstract Introduction Materials and Methods Results and Discussion References

 
A sensitive method is described to detect isolysergic acid diethylamide (iso-LSD) in urine. The compound was extracted from urine and converted to a C-8 carbanion by sodium ethoxide in ethanol. Protonation of the carbanion by water selectively produced LSD. The conversion of iso-LSD to LSD was almost quantitative (98%). The product was purified by solid-phase fractionation and acid–base separation techniques. The trimethylsilyl derivative of LSD was detected by a gas chro-matography–mass spectrometry method. The overall recovery of the procedure was approximately 69%. Quantification of iso-LSD was linear over the concentration range 50–2000 ng/L. In specimen analysis, iso-LSD was detected when the LSD concentration was below the limit of the detection (50 ng/L) of the procedure. Because iso-LSD is a byproduct of illicit preparation of LSD, presence of iso-LSD in urine is an indication of LSD use.

	Introduction

Top Abstract Introduction Materials and Methods Results and Discussion References

 
Lysergic acid diethylamide (LSD) is a potent hallucinogen, approximately 100–150 times as potent as psilocybin and 4500–9275 times as potent as mescaline (1)(2).¹ The compound is so powerful that a very small fraction of the toxic dose (estimated LD₅₀ 14 mg) may produce profound hallucinogenic effect (3). Typical street doses are in the range from 40 to 120 µg and may cost only $3–$5. The compound showed no harmful effects when tested on humans in doses of 100 µg/day for several months (4). Although the lethal dose in humans is not known and appears to be much higher than the psychoactive dose, a dose always exists that can produce death from respiratory failure. The major concern of LSD abuse is the long duration of action (up to 12 h) and fatal accidents and suicides during the state of intoxication (5).

LSD is a semisynthetic compound derived from naturally occurring ergot alkaloids. The major source of these alkaloids is a parasitic fungus, Claviceps purpuria, that grows on rye and other grains (6). These alkaloids are also found in the seeds of morning glory, Rivea corymbosa (7). In illicit preparation of LSD, these compounds are hydrolyzed to produce a common compound, lysergic acid, which is then reacted with diethylamine to produce the LSD.

Theoretically, four optical isomers are possible from the two asymmetric carbon atoms in the LSD molecule (C-5 and C-8 in Fig. 1 ). Many of the ergot alkaloids are isomeric at the C-8 position (8)(9)(10), but none of these alkaloids is isomeric at C-5. Therefore, only two compounds, lysergic acid and isolysergic acid, are formed after hydrolysis of the ergot alkaloids. In lysergic acid, the C-5 hydrogen atom and the C-8 carboxylic acid are in cis configuration, and in isolysergic acid, these two functions are in trans configuration (11)(12). If these two isomers are not separated before the illicit preparation of LSD, both LSD and iso-LSD are formed. Moreover, a strong base and long reaction time in the hydrolytic step may transform some of the lysergic acid to isolysergic acid (9).

View larger version (13K): [in this window] [in a new window]   Figure 1. Sodium ethoxide isomerization of iso-LSD to LSD.

The metabolic disposition of LSD in animals and humans has been reviewed (13). Soon after administration, LSD metabolizes extensively, with only <1% excreted in urine unchanged (13)(14)(15). The metabolism primarily involves aromatic hydroxylation of LSD to 13- and 14-hydroxy-LSD. The compounds, however, could not be conclusively identified because no reference compounds were available to compare the results. Two other urinary metabolites, 2-oxo-LSD and N-desmethyl-LSD, were also detected in animals (14)(15). The identity of N-desmethyl-LSD in human urine was confirmed by tandem mass spectrometry and negative chemical ionization mass spectrometry (16).

For the past 10 years, we have been testing LSD in urine to curb LSD abuse in the Department of Defense. In routine analysis, a RIA screening method is used to identify the presumptive positive specimens in a large number of test samples. Presence of LSD in urine is then confirmed by a gas chromatographic and electron impact mass spectrometric technique (17). In confirmation, we observed that several specimens contained iso-LSD. The compound is also present in some urine specimens, when the LSD concentration is below the limit of detection of the procedure (50 ng/L). Although iso-LSD can be detected by comparing retention time and two ion ratios with that of a reference calibrator, procedures to quantify iso-LSD in urine failed because calibration solutions with definite iso-LSD concentrations cannot be prepared. Some of the iso-LSD was changed to LSD during standard laboratory storage conditions.

Iso-LSD is not psychoactive (18), but like LSD, the compound is classified as a schedule I drug under the Controlled Substances Act of 1970. Because LSD is prepared from ergot alkaloids with isomeric configuration at C-8 position, both LSD and iso-LSD are present in most illicit preparations. To control the use of LSD and iso-LSD, establishing a procedure to detect and quantify these compounds in urine is prudent. In this report, we describe a method that can be used routinely to detect the iso-LSD in large numbers of urine specimens.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results and Discussion References

 
chemicals, reagents, and supplies
LSD, iso-LSD, and lysergic acid methylpropylamide (LAMPA) were purchased from Research Triangle Institute. Solid-phase extraction (SPE) columns containing silica-based propylamine (500 mg) were purchased from Supelco or United Chemical Technologies. Bis(trimethylsilyl)trifluoroacetamide (BSTFA) and Teflon-lined screw-capped Reacti-vials were purchased from Pierce. All solvents and reagents were analytical- or HPLC-grade quality and purchased from Sigma Chemical Co. or Fisher Scientific.

instrument
A Hewlett-Packard (HP) gas chromatography–mass spectrometry (GC-MS) system consisted of an HP 5890 series II Plus GC and 5972 quadrupole mass selective detector (MSD), and the Vectra XM2 4/100i computer workstation was used.

instrument conditions
The flow rate of helium through a DB-5MS capillary column [5:95 phenyl:methyl siloxane, 15 m x 0.25 mm (i.d.)] (J&W Scientific) was 1.4 mL/min. The instrument was operated in splitless and temperature program modes. Oven temperature was increased from 160 to 300 °C at 30 °C/min. The oven was held at 160 °C and 300 °C for 1 and 3 min, respectively. Injection and detector temperatures were 285 °C and 300 °C, respectively. Glass wool was not used in the GC liner. Active sites in the glass wool decreased the instrument sensitivity. The MSD was operated in the electron ionization mode. The electron multiplier voltage of the detector was 800–1000 V above the autotune value. The selected ion monitoring window and dwell time of the ions were 0.2 amu and 50 ms, respectively.

preparation of sodium ethoxide in ethanol (0.5 mol/l)
The reagent was prepared by dissolving 575 mg of sodium (3–8 mm spherical) in portions in 50 mL of anhydrous ethanol. The reaction was conducted in a round-bottom flask protected with a drying tube and away from any flame and spark. When the evolution of hydrogen gas subsided (~30 min), the solution was transferred into a screw-capped bottle. The bottle was tightly closed and stored in a refrigerator. The reagent was found suitable for isomerization of iso-LSD to LSD for at least 6 months.

initial extraction of lsd and iso-lsd from urine
Internal calibrator, LAMPA (0.1 mL, 80 µg/L), was added to 10-mL urine specimens and standard urine solutions containing known amounts of iso-LSD. The solutions were made basic with 0.5 mL of NH₄OH (14.8 mol/L) and saturated with approximately 5 g of NaCl. The drugs were then extracted with 5 mL of 1-chlorobutane by shaking the mixtures for 15 min on an orbital shaker at 60 rpm. The mixtures were then centrifuged. The clear supernatants were transferred to other tubes and evaporated to complete dryness.

isomerization of iso-lsd to lsd
The dried extract containing iso-LSD was dissolved in 100 µL of 0.5 mol/L ethanolic sodium ethoxide. The tubes were capped and heated in a metal block at 50 °C for 10 min. The reaction was terminated by adding 2 mL of distilled water to the uncapped tubes. Approximately 1 g of NaCl and 3 mL of 1-chlorobutane were added to the solutions. The mixtures were vortex-mixed for 1 min and centrifuged. The clear supernatants were transferred to other tubes and evaporated to dryness.

solid-phase purification
The isomerized compound and the internal calibrator were dissolved in 2 mL of isooctane:CH₂Cl₂:triethylamine (TEA) (50:50:0.1). The solutions were poured into SPE tubes prewashed with 3 mL each of MeOH:TEA (100:0.1), CH₂Cl₂:TEA (100:0.1), and isooctane:CH₂Cl₂:TEA (50:50:0.1). The extracts were allowed to pass through the sorbent by its gravity flow. The columns were washed with 3 mL of CH₂Cl₂:TEA (100:0.1) to remove low polar impurities and then dried under a vacuum (300 mmHg) for approximately 10 s. The drugs were eluted from the columns with 3 mL of MeOH:CH₂Cl₂:TEA (0.2:10:0.01). The solutions were evaporated to dryness.

acid–base purification
After the solid-phase purification, the extracts were dissolved in 3 mL of 1-chlorobutane. The alkaloids were then extracted with 3 mL of phosphate buffer (0.1 mol/L, pH 4.5) by vortex-mixing the solutions for 1 min. The solvents were discarded. An additional 3 mL of 1-chlorobutane was used to wash the aqueous phases and then discarded. This process removed most of the acidic and neutral impurities from the acidic solutions. The solutions were made basic with 0.1 mL of NH₄OH (14.8 mol/L) and then saturated with approximately 1 g of NaCl. The basic drugs were extracted with 3 mL of 1-chlorobutane by vortex-mixing the mixtures for 1 min and centrifugation. The clear organic solutions were transferred into 5-mL Reacti-vials and evaporated to dryness. The extracts were dissolved in 0.1 mL of ethanol containing 1 mL/L TEA and the solutions were again evaporated to dryness. The TEA efficiently dissolved the LSD bound to the silicic acid on the glass tube.

derivatization
The purified extracts were dissolved in 20 µL of BSTFA. The vials were capped and vortex-mixed for 1 min. The solutions were heated in a metal block at 70 °C for 15 min. Approximately 3 µL was used for GC-MS analysis. The product was sensitive to moisture but was stable for at least 5 days when protected from moisture and stored at 2–6 °C.

	Results and Discussion

Top Abstract Introduction Materials and Methods Results and Discussion References

 
In our routine analysis for LSD, a RIA method is initially used to screen a large number of specimens. Positive specimens are tested again by a GC-MS confirmatory procedure (17). During GC-MS analysis, we observed that several specimens contained iso-LSD. The identification was based on monitoring three ions (m/z 395, 293, and 253) and comparing the retention times (±2%) and ion ratios (±20%) with the reference compound. Often, the identifications were not conclusive because interfering compounds appeared at the retention time of iso-LSD.

Freshly prepared solutions of iso-LSD (2000 ng/L) in methanol or acetonitrile were evaporated to dryness. The compound was derivatized with BSTFA at 70 °C for 15 min and analyzed by the GC-MS method. On the basis of the total ion chromatogram, approximately 5–7% LSD was detected in the iso-LSD samples. Although heating was required to ensure complete derivatization, a substantial amount of compound was found reacted with BSTFA even at room temperature. Derivatization at different times and temperatures (room temperature/15 min, 50 °C/10 min, and 70 °C/15 min) showed no significant difference in isomerization, indicating iso-LSD was not isomerized to LSD at the derivatization step. The isomerization was also unaffected by the GC-MS conditions. The percent of LSD remained almost the same when the derivatized iso-LSD was tested at different injection (245–285 °C) and oven temperatures (285–305 °C). Therefore, either the commercial source of solid iso-LSD contained 5–7% LSD or the compound isomerized almost immediately when dissolved in organic solvents.

Iso-LSD rapidly isomerized to LSD when dissolved in organic solvents and stored at 2–6 °C. When a freshly prepared iso-LSD solution was tested, the amount of LSD was 5–7%. In 3 days, 16–18% of iso-LSD was changed to LSD (Table 1 ). The isomerization continued for approximately 10 months until it reached an equilibrium (K = 0.5). No further conversion was observed after this period.

Because of the instability of iso-LSD, the compound cannot be used to quantify iso-LSD in urine with a GC-MS method of analysis. Initially, we attempted to isomerize iso-LSD to LSD by aqueous sodium hydroxide and analyze the product by the procedure described (17). When iso-LSD in 0.1 mol/L aqueous sodium hydroxide was heated at 50 °C for 60 min, a considerable amount of iso-LSD remained unchanged. Moreover, >50% of iso-LSD was lost during the incubation period. The loss was apparently due to the amide hydrolysis of iso-LSD.

Conversion of isolysergic acid to lysergic acid was extensively investigated (9). When isolysergic acid was dissolved in 100 g/L aqueous potassium hydroxide and heated on a steam bath for 1 h, only 60% of the compound was isomerized to lysergic acid. Similar base-catalyzed conversion of lysergic acid amide produced only 45% isolysergic acid amide. Approximately 33% of the compound remained unchanged. The conditions used to catalyze the isomerization were also used to hydrolyze lysergic acid amide or other ergot alkaloids to lysergic acid. Therefore, the remaining 22% of lysergic acid amide was lost because of hydrolysis of lysergic acid amide to lysergic acid. None of the methods effectively converted one isomer to the other.

In iso-LSD, the C-8 hydrogen atom is activated by the adjacent carbonylamide group and the C-9 double bond (Fig. 1 ). The hydrogen atom can be easily removed by a nonaqueous base. When treated with sodium ethoxide in ethanol, iso-LSD formed a C-8 carbanion. Addition of water selectively protonated the carbanion to LSD. The conversion to LSD was found to be almost quantitative. The selective protonation is likely due to steric effects of the N,N-diethylamide group on the C-8 carbanion. The reaction was almost immediate (72.3%) after the addition of sodium ethoxide followed by water, but the optimum yield was found by heating the iso-LSD in 0.5 mol/L ethanolic sodium ethoxide at 50 °C for 10 min (98%) (Table 2 ). When the same condition was used to isomerize LSD to iso-LSD, no detectable peak was observed at the retention time of iso-LSD.

The effect of sodium ethoxide on the internal calibrator LAMPA was also investigated. A known amount of drug and internal calibrator was treated with sodium ethoxide at 50 °C over a period of 60 min. The ion area ratios of LSD and LAMPA remained almost the same during the incubation period, indicating LAMPA as a suitable internal calibrator for quantification of iso-LSD. Deuterated LSD (d₃-LSD) cannot be used as internal calibrator because the trimethylsilyl derivatives of d₃-LSD and LSD appear at the same retention time and produce common ions (17).

In extraction, the urine was made basic and saturated with salt. The drugs were then extracted with a solvent. The organic solution was evaporated to dryness. The dried extract was treated with sodium ethoxide to convert iso-LSD to LSD. The product was purified by using two principles of purification. In the initial solid-phase purification technique, the compounds were fractionated to remove low and high polar compounds. After the initial adsorption of the compound in the solid-phase material, the sorbent was washed with a mixture of solvents to remove the low polar impurities. The drugs were then eluted with a suitable mixture of solvents, leaving behind the polar impurities in the solid-phase column. The isomerization of iso-LSD to LSD before the solid-phase purification was essential, because it allowed a narrow solid-phase chromatographic band that contained the least amount of low polar impurities.

Solid-phase purification alone was insufficient to remove all undesirable compounds from the extract. Although it removed many urinary impurities, it introduced other impurities present in the column materials. Extensive solvent wash of the column materials did not remove all the impurities. The acid–base separation after the solid-phase purification successfully removed most of the remaining impurities. Sometime variation in polarity of the column material required adjustment in the methanol concentration at the final elution step. Concentrations in the range of 1.5–2.0% were found to be most suitable. Excess methanol was avoided because it extracted polar impurities that interfered with the detection process.

LSD bound irreversibly with silicic acid on the glass surface. In extraction, when glass tubes were used and the solution was not basic, TEA was added to the solution. The strongly basic nitrogen in the TEA selectively binds to the active sites of the glass surface and protects LSD from breakdown. Similar loss of LSD derivative was observed when the compound was analyzed in the GC column. Loss of sensitivity of samples injected at the beginning of the run was a major problem. Pure LSD derivatized with BSTFA was injected 2–3 times to condition the column. The trimethylsilyl groups of LSD molecules effectively blocked the active sites of silica inside the GC column. A considerable improvement in instrument sensitivity was observed after the column conditioning.

The recovery of the procedure was determined by adding internal calibrator at the beginning and end of the extraction and purification steps. The overall recovery was found to be 69% ± 4%. A major problem in LSD and iso-LSD analysis is the coeluting impurities. The amount of drugs is so low that minor impurities may seriously interfere in the detection process. A clean extract was considered more important than the overall yield of the drugs. Iso-LSD in urine at concentrations ranging from 25 to 5000 ng/L was analyzed with this procedure. Excellent linearity was observed over the concentration range 50–2000 ng/L. The slope, intercept, and correlation coefficient were 1.1, -8.2, and 1.0000, respectively. On the basis of the curve, when the iso-LSD concentrations at 50 and 2000 ng/L were recalculated, the amounts were found to be 53 and 1992 ng/L, respectively. Interrun quantifications were also within ±10% of the theoretical values. Limit of detection was 50 ng/L. Below this concentration, qualifying ion ratios exceed ±20% of that of reference calibrator at concentration 200 ng/L.

The procedure was successfully applied to analyze iso-LSD in a large number of urine specimens. When specimens contained both LSD and iso-LSD and were analyzed by this process, the LSD concentration represented the total amount of LSD (LSD iso-LSD). The amount of iso-LSD was calculated from the difference of total LSD and the LSD tested without isomerization. Fig. 2 shows a specimen analyzed with and without the isomerization step. Both iso-LSD and LSD were detected in the initial detection of LSD (Fig. 2A ). The concentration of LSD was 220 ng/L. After isomerization, the concentration of the total LSD increased to 416 ng/L (Fig. 2B ). The iso-LSD in the specimen was then calculated from the difference in concentrations of total LSD and LSD (196 ng/L).

View larger version (25K): [in this window] [in a new window]   Figure 2. Ion chromatogram of a specimen without isomerization (LSD 220 ng/L) (A) and of the same specimen after sodium ethoxide isomerization (total LSD 416 ng/L) (B). Retention times of trimethylsilyl derivatives of iso-LSD, LSD, and LAMPA are 6.71, 6.83, and 7.07 min, respectively.

Iso-LSD is a byproduct of illicit preparation of LSD. They are stereoisomers and classified as schedule I drugs. Therefore, detection of any of the compounds or their combination is an indication of LSD use. Although two GC-MS procedures may be required to determine iso-LSD in urine, in forensic investigation only one GC-MS confirmation procedure that detects the iso-LSD and LSD together is recommended because it saves time and is cost effective.

	Acknowledgments

 
We thank Eric Shimomura of AFIP, Rockville, MD and Gregory K. Poch of Navy Drug Screening Laboratory, San Diego, CA for some technical assistance and suggestions, and Lisa K. McWhorter for providing the urine specimens.

	Footnotes

 
Navy Drug Screening Laboratory, San Diego, CA 92134.

¹ Nonstandard abbreviations: LSD, lysergic acid diethylamide; LAMPA, lysergic acid methylpropylamide; SPE, solid-phase extraction; BSTFA, bis(trimethylsilyl)trifluoroacetamide; GC-MS, gas chromatography–mass spectrometry; MSD, mass selective detector; and TEA, triethylamine.

The opinions expressed herein are those of the authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense.

	References

Top Abstract Introduction Materials and Methods Results and Discussion References

 

1. Isabell H. Comparison of the reactions induced by psilocybin and LSD-25 in man. Psychopharmacologia 1959;1:29-38.
2. Wolbach AB, Jr, Isabell H, Miner EJ. Cross tolerance between mescaline and LSD-25 with a comparison of the mescaline and LSD reactions. Psychopharmacologia 1962;3:1-14.
3. Pradhan SN, Hollister LE. Abuse of LSD and other hallicinogenic drugs. Pradhan SN Dutta SN eds. Drug abuse clinical and basic aspects 1977:274-289 The C.V. Mosby Co. St Louis. .
4. Hoffer A. D-Lysergic acid diethylamide (LSD): a review of its present status. Clin Pharm Ther 1965;6:183-255. [Medline] [Order article via Infotrieve]
5. Curry AS. Advances in forensic and clinical toxicology 1972:41-49 CRC Press Boca Raton, FL. .
6. Stoll A, Hofmann A. Die alkaloide der ergotoxingruppe: ergocristin, ergokryptin und ergocornin. Helv Chim Acta 1943;26:1570-1601.
7. Hofmann A, Tscherter H. Isolierung von lysergsaure-alkaloiden aus der mexikanischen zauberdroge Ololiuqui (Rivea corymbosa L. Hall. f). Experientia 1960;16:414.
8. Smith S, Timmis GM. The alkaloids of ergot. Part VI. Ergometrinine. J Chem Soc 1936;:1166-1169.
9. Smith S, Timmis GM. The alkaloids of ergot. Part VII. Isoergine and isolysergic acids. J Chem Soc 1936;:1440-1444. [Medline] [Order article via Infotrieve]
10. Rutschmann J, Stadler PA. Chemical background. Berde B Schild HO eds. Handbook of experimental pharmacology, ergot alkaloids and related compounds, Vol. 49 1978:29-85 Springer-Verlag Berlin. .
11. Stenlake JB. Dissociation constants of the lysergic acids. J Chem Soc 1955;:1626-1628. [ISI][Medline] [Order article via Infotrieve]
12. Leemann HG, Fabbri S. Uber die absolute konfiguration der lysergsaure. Helv Chim Acta 1959;42:2696-2709.
13. Foltz RL, Foltz RB. Lysergic acid diethylamide (LSD). Baselt RC eds. Advances in analytical toxicology 1989;Vol. II:140-158 Year Book Medical Publishers Chicago. .
14. Axelrod J, Brady RO, Witkop B, Evarts EV. The distribution and metabolism of lysergic acid diethylamide. Ann N Y Acad Sci 1957;66:435-444.
15. Siddik ZH, Barnes RD, Dring LG, Smith RL, Williams RT. The fate of lysergic acid di[¹⁴C]ethylamide ([¹⁴C]LSD) in the rat, guinea pig and rhesus monkey and of [¹⁴C]iso-LSD in rat. Biochem Pharmcol 1979;28:3093-3101.
16. Nelson CC, Foltz RL. Determination of lysergic acid diethylamide (LSD), iso-LSD, and N-demethyl-LSD in body fluids by gas chromatography/tandem mass spectrometry. Anal Chem 1992;64:1578-1585. [Medline] [Order article via Infotrieve]
17. Paul BD, Mitchell JM, Burbage R, Moy M, Sroka R. Gas chromatography–electron impact mass fragmentometric determination of lysergic acid diethylamide in urine. J Chromatogr 1990;529:103-112.
18. Isabell H, Miner EJ, Logan CR. Relationships of psychotomimetic to anti-serotonin potencies of congeners of lysergic acid diethylamide (LSD-25). Psychopharmacologia 1959;1:20-28.