Detection and Characterization of Cocaine and ...

Detection and Characterization of Cocaine and Related Tropane Alkaloids in Coca Leaf, Cocaine, and Biological Specimens J.M. Moore and J. F. Casale Special Testing and Research Laboratory Drug Enforcement Administration U.S. Department of Justice McLean, VA 22102-3494 G.Fodor Department of Chemistry West Virginia University Morgantown, WV 26506-6045 and A. B.Jones School of Pharmacy University of Mississippi University, MS 38677

TABLE OF CONTENTS INTRODUCTION........................................................................................ 78 I.

FORENSIC CHEMISTRY OF COCA AND COCAINE............................ 79

II.

CHARACTERIZATION OF COCAINE AND RELATED TROPANE ALKALOIDS IN COCA LEAF ................................................................... 80 A. Coca and the Origin of Cocaine ........ ..... ...... ... .... ..... ... ......... ....... ... . .. .... . 80 B. Tropane Alkaloids of South American Coca Leaf and Greenhouse

Cul ti vars .. .. .... ....... ...... .... ... ....... .. .... ................. .. ..... ...... ..... .. .. .. .. ... .. ... ... .. 81 III.

SURVEY OF METHODS FOR THE DETECTION OF TROPANE ALKALOIDS AT TRACE LEVELS IN BIOLOGICAL SAMPLES ......... 91 A. Ultraviolet Spectroscopy and Atomic and Derivation Spectrometry..... 92 B. Infrared Spectroscopy ............................................................................ 92 C. Raman Spectroscopy .. ... .... ...... ... ... ....... ....... .... ...... ......... .. .... ..... ......... .... 92 D. Mass Spectrometry ..... ..... .. ...... .... ... ... ... ... ...... ........ ...... .. .... ... ...... ..... .... ... 92 E. Chromatography..................................................................................... 93 F. Immunoassay.......................................................................................... 94 G. Miscellaneous Methods.......................................................................... 95 ACKNOWLEDGMENTS ............................................................................ 95 REFERENCES ............. ................................................................................ 95 ABOUT THE AUTHORS .......................................................................... 101

1042-7201/07-02-9Sn7-101/$12.50 •Copyright© 1995 Central Police University Press

Detection and Characterization of Cocaine and Related Tropane Alkaloids in Coca Leaf, Cocaine, and Biological Specimens REFERENCE: Moore JM, Casale JF, Fodor G, Jones AB: Detection and characterization of cocaine and related tropane alkaloids in coca leaf, cocaine, and biological specimens; Forensic Sci Rev 7:77; 1995. ABSTRACT: Cocaine, atropine and scopolamine are the three most important alkaloids in the tropane group. The detection of these alkaloids and their metabolites, at trace levels in biological matrices, is reviewed. These matrices include human and rat physiological fluids such as blood, urine, and saliva as well as human body tissue and hair. The detection, isolation, and determination of cocaine and related tropane alkaloids in cocaine-bearing leaf of South American and greenhouse-cultivated coca is discussed. The relationship between tropane alkaloids in coca leaf and their presence in illicit refined cocaine is addressed. A survey of modern methods for the detection of tropane alkaloids, including mass spectrometry, ultraviolet, infrared and Raman spectroscopy, gas and high-performance liquid chromatography and immunoassay techniques, is presented. KEY WORDS: Atropine, biological fluid, blood, chromatography, cocaine, coca leaf, drugs, erythroxylum, immunoassay, scopolamine, spectroscopy, tissue, toxicology, tropane alkaloids, urine.

INTRODUCTION The importance of tropane alkaloids in biological systems is underscored by their large number, as there were about 151 of them detected in plants as late as 1987 [120]; even more significant is the biological activity some of them exert. Atropine (Structure 1), cocaine (Structure 2), and scopolamine (Structure 3) are three of these alkaloids that have medicinal and/or socioeconomic importance. Among these, cocaine is addictive and, therefore, has a serious impact on society. Though it still has legitimate pharmaceutical application as an anesthetic, its far greater impact on society has been its abuse as an illicit drug. Although cocaine can be prepared synthetically, this methodology is complex and time-consuming and provides rather low yields [12,26]. For that reason, and because the source material is readily available, virtually all of the world's supply of refined illicit cocaine is derived from the leaf of the South American coca plant, Erythroxylum coca var. coca (ECVC).

Structure 1. Atropine

Structure 2. Cocaine

/CH3 N

r

CHPH

o-c~-Q~ II 0

0

Structure 3. Scopolamine

Illicit Cocaine Manufacture Most clandestine cocaine-manufacturing laboratories are found in the South American countries of Colombia, Peru, and Bolivia. The manufacturing process generally occurs in two stages [26]. The first involves extraction of cocaine and other alkaloids from the leaf, followed by refinement of the extract and eventual conversion to the base form. This first stage takes place in rather unsophisticated laboratories located in the immediate vicinity of coca leaf cultivation. Most of these laboratories are found in Peru and Bolivia. The refined base is then subjected to

Forensic Science Review • Volume Seven Number Two • December 1995

79

further purification and finally converted to cocaine hydrochloride, most often in another laboratory where the necessary chemicals and supplies are available. These clandestine facilities are located primarily in Colombia. Numerous other tropane alkaloids of the leaf are coprocessed with cocaine and appear as impurities in the final product. Furthermore, these alkaloids can undergo degradation during the manufacturing process and their byproducts also appear in the refined cocaine. Characterization of minor coca leaf alkaloids is of forensic interest and has important strategic and tactical implications from a drug enforcement perspective (136].

I. FORENSIC CHEMISTRY OF COCA AND COCAINE Coca leaf abuse (chewing) has not been of consequence outside of South America. However, understanding the alkaloid chemistry of ECVC and other coca cultivars is still important from a forensic-chemistry perspective. To understand this assertion more fully, one must consider the significance of the complete characterization of manufacturing alkaloidal impurities and by-products present in refined illicit cocaine hydrochloride and base. All of the compounds in this discussion include tropane alkaloids that have been co-extracted from the leaf unchanged, along with cocaine, or have been chemically altered during the manufacturing process. The latter compounds will be referred to as manufacturing by-products. The characterization of alkaloidal impurities and manufacturing by-products in illicit drugs such as cocaine includes:

5.

6.

for the subsequent restriction regarding their availability; Differentiation between botanically derived cocaine and its synthetically manufactured counterpart (e.g., via the Willstatter synthesis [191]) would be feasible; and The only unequivocal means of detennining whether an impurity in illicit refined cocaine is a leaf alkaloid or a manufacturing artifact (by-product) is to establish the presence/absence of that compound in the plant.

To relate the alkaloid content of coca leaf to alkaloidal impurities and manufacturing by-products in illicit refined cocaine, it was necessary to first review the literature for reports of coca leaf analyses. Surprisingly, in the most comprehensive review of the coca/cocaine literature up to 1988 (185], not a single article regarding the quantitative determination of any alkaloids other than cocaine and the cinnamoylcocaines (Structure 4) could be found. This was confirmed by a literature search through June of 1992 using CAS ONLINE, a computer retrieval program offered by Chemical Abstracts Services. Furthermore, for most of the last 150 years, there have been but a handful of tropane alkaloids reported in the major cocaine-bearing leaf, i.e., ECVC.

Structure 4. trans-Cinnamoylcocaine I. 2. 3.

Detection and structural elucidation of the impurities and by-products; Rationalizing their presence; and Quantitative determination.

These objectives are important for a number of reasons: 1.

It enables the forensic chemist to compare different

drug seizures to determine if they are of a common origin [141]; this is important in developing drug conspiracy cases; 2. In the case of illicit drugs derived from botanical sources, e.g., cocaine, one may be able to render judgments regarding geographic origin of the botanical source or the site of manufacture; '3. The characterization ofalkaloidal impurities and manufacturing by-products allows for the differentiation of the illicit drug from its legitimate, pharmaceutical counterpart; 4. It may be possible to understand the manufacturing process more fully. This would result in the identification of solvents and chemicals used, which would assist law-enforcement efforts in monitoring these chemicals

In recent years, a number of gas chromatographic and high-performance liquid chromatographic methods have been developed that generate so-called "cocaine signature impurity" profiles for refined illicit cocaine samples (29,30,37,56,93,109,110,123, 124,126,135, 137, 138,140,141,144]. These have been used to compare illicit cocaine seizures to determine their commonality of origin (141] and to give insight into the manufacturing process (26]. These profiles are based upon the detection and determination of trace-level tropane alkaloids and their degradation products in refined cocaine [ 136]. Such methodology has been reviewed in a recent report (136]. Over the past few years, the analysis of biological samples for the presence of cocaine and/or its metabolites has received considerable interest. This is primarily because of the increased use of cocaine in the drug-user community and the increased interest in workplace testing for drugs of abuse. Cocaine is one of the five drugs or

Moore, Casale, Fodor, & Jones • Detection and Characterization of Cocaine and Related Tropane Alkaloids

,.

80

drug classes for which workplace drug testing is conducted in Federal agencies of the U.S. government and selected private industries which are required under Federal regulations to conduct workplace drug testing [52]. Cocaine is used both in its salt form as cocaine HCl and as the free-base form. Various purities of this drug are found, ranging from coca paste (an intermediate product in the cocaine refinement process having cocaine concentrations typically ranging from 40 to 80%) to nearly pure cocaine HCl or "crack" cocaine (a free-base smokable form of the drug). The drug can be ingested, injected, snorted, or smoked; the latter two usages are the most popular [23]. When smoked or snorted, cocaine is rapidly absorbed into systemic circulation via the pulmonary system or the nasal mucosa. Cocaine in the body is rapidly and extensively metabolized by liver and plasma enzymes, primarily to benzoylecgonine (Structure 5) and ecgonine methyl ester (Structure 6). Both of these metabolites are excreted in the urine along with some unmetabolized cocaine. The Federal programs in the U.S. use specific methodology directed toward benzoylecgonine in urine, the compound with the longer half-life (benzoylecgonine can usually be detected for 2-3 days after a single dose). Specifically, these programs use an immunoassay method as an initial screening procedure. This is usually followed by gas chromatographic/mass spectrometric (GC/MS) analysis to confirm any positive results obtained by the initial test [52]. These Federally directed programs use cutoff concentrations, below which a sample is called negative, even though the analyte may be present. The immunoassay screening procedures respond to several cocaine metabolites; thus, the cutoff concentration is typically 300 ng/mL, whereas the GC/MS confirmation procedures are specific for benzoylecgonine and typically have cutoff concentrations of 150 ng/mL [52].

Structure 5. Benzoylecgonine

II. CHARACTERIZATION OF COCAINE AND RELATED TROPANE ALKALOIDS IN COCA LEAF A. Coca and the Origin of Cocaine To enhance the detection/determination and rationalization of trace-level tropane alkaloid impurities/by-products in refined illicit cocaine, the in-depth analysis of coca leaf is a prerequisite. In the following discussion, existing methodologies for coca leaf analysis are reviewed as well as more recent developments in the characterization of new tropane alkaloids in that matrix. Throughout this review, the relationship between the alkaloids in the leaf and their presence, along with degradation by-products, in illicit cocaine will be mentioned. The review herein will encompass mostly the coca/ cocaine literature published since 1980. Recommended reading that provides a good overview of alkaloids in the Erythroxylum genus (including plants with little or no cocaine) has been provided by Novak et al. [151], Evans [59], and El-Imam et al. [55]. Another invaluable reference for the cocaine investigator is the comprehensive annotated cocaine bibliography by Turner et al. [185]. Although the focus of this review is concerned with the characterization of tropane alkaloids in coca, the relationship between a given alkaloid of the leaf and its presence in illicit refined cocaine will also be addressed. The same will be done regarding the alkaloid in the leaf and manufacturing by-products in illicit cocaine. Also, analytical data for alkaloids present in greenhouse-cultivated coca, as well as in coca grown at a tropical site other than South America will be reviewed briefly. Though there are four major coca cultivars, most attention will be focused upon ECVC as this is the source of most refined illicit cocaine. South American coca leaf is cultivated from Colombia through Bolivia, primarily along the eastern slopes of the Andes mountains and in adjacent valleys. According to Plowman [157,158], there are four cocaine-bearing and taxonomically distinct coca cultivars found in South America. All have measurable and physiological quantities of cocaine. The most prevalent of these is from the family Erythroxylaceae, genus Erythroxylum, species coca and variety coca. Referred to as E. coca var. coca (ECVC), or "Huanuco coca," this variety is found in Colombia, Ecuador, Peru, and Bolivia. E. coca var. coca is believed responsible for most of the world's supply of illicit cocaine. It is cultivated largely in the Upper Huallaga Valley and Cusco area of Peru and in the Chapare and Yungas regions of Bolivia.

Structure 6. Ecgonine methyl ester Forensic Science Review • Volume Seven Number Two • December 1995

81

Another variety of coca is E. coca var. ipadu (ECVI). Also known as "Amazonian coca," it is cultivated on a relatively small scale by Indian tribes of the upper Amazon in parts of Colombia, Peru, and Brazil [157,158]. Although closely related to E. coca var. coca, the leaf of E. coca var. ipadu has the lowest cocaine content of the four cultivars. A third cocaine-bearing plant reported by Plowman, and of a different genus, is E. novogranatense var. truxillense (ENVT). Also known as "Trujillo coca," it is found in north-central and coastal Peru. It has been perhaps best known as the traditional source for the flavoring in the soft drink Coca-Cola (sans cocaine). The fourth coca cultivated in South America is E. novogranatense var. novogranatense (ENVN). Also known as "Colombian coca," it is found in the interAndean valleys of Colombia, along the Caribbean coast and in the more-moist parts of the Colombian Andes. Unlike the other three cultivars, ENVN is quite stable under widely varying ecological conditions. Although ENVN and ENVT have significant levels of cocaine, they are not believed to be a major source of this illicit drug. This may be in part because both varieties have reduced areas of cultivation and significantly increased levels of other coca alkaloids; the latter could presumably render any cocaine product less than suitable for marketing without further, relatively expensive, processing.

B. Tropane Alkaloids of South American Coca Leaf and Greenhouse Cultivars The previous and continuing characterization of tropane alkaloids in South American coca can be divided into three sections as described here. Section I includes the characterization of well-established alkaloids, namely cocaine, cinnamoylcocaine, ecgonine methyl ester and aand ~-truxilline (Structure 7). Virtually all of that work was done using South American ECVC (and Java coca) and occurred in the middle-to-late 19th and early 20th centuries. It wasn't until 1977 that the first significant and accurate quantitative data for cocaine and the cinnamoylcocaines began appearing in the literature [31,37,46,87, 139,159,164,176,186, 187,197]. This data will also be reviewed in Section I. Section II describes the characterization of new major and tra,ce-level alkaloids that were first detected and structurally identified in illicit, refined cocaine samples, and then confirmed in South American coca leaf. These include both the cis- and trans-isomers of cinnamoylcocaine, an additional nine isomeric truxillines, the socalled hydroxycocaines and the trimethoxy-substituted analogues of cocaine and related alkaloids. Also de-

scribed in this section is new methodology for the quantitation of selected alkaloids for both South American and greenhouse-cultivated coca leaf. These aforementioned studies were all reported between 1987-1994. Section III reports coca studies that began around 1993 and are on-going. In this work, new methodologies are described for the detection, isolation, and/or structural characterization of about 125-150 new, trace-level coca alkaloids. Virtually all of this work is associated with Peruvian and Bolivian ECVC and their extracts. 1. Section I Studies

a.

Cocaine and cinnamoylcocaines Arguably, it was Gaedcke [64,77] or Wohler [194] who was the first to extract cocaine from coca leaf in the 1850s or 60s. Around 1860, Albert Nieman isolated a white crystalline substance from coca which he called cocaine [112,194, 195]. The structure of cocaine was not confirmed until 1923, by Willstatter and co-workers via total synthesis [191]. It wasn't until 1953 that Findlay elucidated the stereochemistry of cocaine [62]. The first reports of cinnamoylcocaine in coca were by Liebermann [115, 116] and Giesel [66,67], who isolated it from Java coca leaves. In those early studies there was no differentiation of the cis- and trans-isomers associated with cinnamoylcocaine. It was not until 1973 that the presence of both geometric isomers of cinnamoylcocaine were reported in refined illicit cocaine, and then confirmed in a crude coca leaf extract [138]. Reliable quantitative data for cocaine and the cinnamoylcocaines in coca did not appear until the late 1970s. One of the most comprehensive of those studies was by Plowman and ~ivier [158], who determined the cocaine and cis- and trans-cinnamoylcocaine content for the four major South American cultivars (ECVC, ECVI, ENVT, ENVN), collected in Bolivia, Peru, Ecuador, and Colombia. These results, produced by GC-MS with selected-ion monitoring, are summarized in Table 1.

Structure 7. a-Truxilline

Moore, Casale, Fodor, & Jones • Detection and Characterization of Cocaine and Related Tropane Alkaloids

82

Table 1. Summary of cocaine and cinnamoylcocaine content of dried leaves of the cultivated cocas (E. coca and E. novogranatense) a No. of samples

E. coca var. coca

30

0.23-0.96

0.63

0.0011-0.532 0.068

E. coca var. ipadu

6

0.11-0.41

0.25

0.0-0.0084

E. novogranatense var. novogranatense

3

0.55-0.93

0.77

14

0.42-1.02

0.72

E. novogranatense var. truxillense a b

Cocaine b Range Mean

Total cinnamoylcocaines b Range Mean

Leaf variety

Cis-b Range

Mean

Trans-b Range Mean

0.0-0.44

0.05

0.0-0.11

0.0183

0.005

0.0-0.0084

0.005

0.0

0.0

0.107-0.65

0.379

0.072-0.53

0.287

0.035-0.12

0.092

0.0-0.93

0.231

0.0-0.68

0.154

0.0-0.43

0.0775

Data taken from Ann Bot 51:641; 1983 [158]. Data reported as mg/100 mg dry mass.

Using a deuterated cocaine standard and GC-MS, Holmstedt et al. [87] reported the cocaine content of South American ECVC, ENVT, and ENVN. The average driedleaf cocaine levels of 12 ECVC samples from Peru was 0.73%. A single sample of Peruvian ENVT had a cocaine content of 0.71 %. Two samples of ECVC cultivated in Bolivia had a cocaine content ofO. 70% and 0. 74%. Leaves oflOColombianENVNsamplesvariedfrom0.17-0.76%, with an average of 0.47%. In a GC-flame ionization detector (FID) study of ECVC leaf from Peru, Turner et al. [ 186] reported cocaine levels of 0.60, 0.57, and 0.60% in samples cultivated in Cuzco, Trujillo, and Tingo Maria, respectively. In a subsequent study by the same investigators, the cocaine and cinnamoylcocaine content of Peruvian ECVC leaf grown in three disparate locations was determined [187]. They found that there were differences in the cinnamoy lcocaine content from each site, and that the relative ratios of the total cinnamoylcocaines to cocaine varied with sample origin. In a study by Rivier [164] in 1981,it was observed that the variation of alkaloid content in coca leaf can depend upon environmental conditions and leaf age, and that intra-plant variation of cocaine and cinnamoylcocaines in leaf was significant. Finally, it was concluded that the variation of the alkaloid content in individual leaves was too great to allow the use of the ratio of cocaine to the cinnamoylcocaines as a taxonomic marker. The lower cocaine levels of ECVI leaf ("Amazonian coca") [159] was confirmed in a study by Plowman and Rivier [158], ranging from 0.1-0.4% w/w, based on airdried leaf. Glass and Johnson [68] reported a high performance liquid chromatography (HPLC) method for cocaine

quantitation in coca and compared the results with a GC procedure. The cocaine content of South American ECVC was determined by Solon and Sperling [ 176] using a simple methanol extraction (75 °C) of the leaf, followed by GCFID quantitation. This method, used currently by the U.S. Drug Enforcement Administration, also appeared suitable for the determination of cis- and trans-cinnamoylcocaine [139]. In the most comprehensive study done to date, conducted from 1993-1995, the cocaine and cinnamoylcocaine levels of more than 1000 ECVC leaf samples from Peru and Bolivia were determined using the DEA methodology [31]. This data is summarized in Table 2. In unique studies, cocaine/cinnamoylcocaine ratios were determined in individual coca leaves using massanalyzed ion kinetic energy spectrometry [46,197]. The analysis required as little as 1 mg of sample and required no solvent extraction. That study determined that coca plants from different geographic regions could be distinguished on the basis of alkaloid content. In summary, it is seen that the level of cocaine for most air-dried South American ECVC leaf fell between about 0.60% and 0.75% w/w with an average content near 0.70%. From the most comprehensive survey [31], the total cinnamoylcocaine (cis- +trans-) content, relative to cocaine, was in the range of 8-15%, with an average ratio of cis-ltrans- =(l.5-2.0)/1. It should be noted that during the manufacture of refined illicit cocaine from the leaf, potassium permanganate is used in one of the steps; when done properly, this results in the oxidation of most of the cinnamoylcocaines. Although these alkaloids can still be found in most illicit refined cocaine samples, their levels (relative to cocaine) are significantly reduced, to generally below 4% w/w [136].

Forensic Science Review • Volume Seven Number Two • December 1995

83

Table 2. Cocaine and cinnamoylcocaines content in South America coca leaf

Country/(Region) a Bolivia/(Chapare Valley) Bolivia/(Yungas) Peru/(Upper Huallaga Valley) 0

b c d

No.of sample

Cocaine

338

0.62

42 668

Cis-

Cinnamoylcocaine

Trans-

0.054 90 undance 4000

95

~~3

'11~0

100 I

I

100

9

105

115

110

120

105

138

125

130

135

140

145

150

155 155

COCAINE

96

B

2000 122 lOO

i

110 114

23

132

l45kz

1~1 2~ 55

140

150

0+.;.+++t-l-+-l'-T-t-~.~.~,++./+iC.,....,..,....,..T""""".-HT"°-r-+-~,~.....,...rr-,.........,..,...,.....,-t-r.....,.:.......-.,...,....,...,..,,~.+.-..:.+-,,..,...,._

/Z

-> 90

100

105

1

110

115

120

125

130

135

145

155

mlz Figure 3. Comparative electron ionization mass spectra (vertical and horizontal expansion (90-160 amu) of: (A) pseudococaine and (B) cocaine, acquired from a mass selective detector (MSD). (Reproduced with permission from J Forensic Sci 39:1537; 1994 [35].)

1

5

REFINED COCAINE BASE

2.0e

1.Be

Pscuooco

1.6e4-

CAINE--..

4

1.4e

3 1.2e

2

08-1701 1.0e

8000

6000

g

10

11

12

13

14

Retention Time (Min) Figure 4. Capillary gas chromatographic-flame ionization detection chromatogram (DB-1701) of an ionpair eluate from an illicit refined cocaine base sample containing pseudococaine. Peak identification: 1 = tropacocaine (9.3 min); 2 = dextromethorphan internal standard (11.5 min); 3 =N-norcocaine (13.0 min); 4 = pseudococaine; 5 =cocaine. (Reproduced with pennission from J Forensic Sci 39:1537; 1994 [35].) Forensic Science Review • Volume Seven Number Two • December 1995

89

g.

New Quantitative Methodology for Coca leaf Alkaloids As reviewed in Section I, most of the previously reported methodologies for the quantitation of coca leaf alkaloids pertained only to cocaine and the isomeric cinnamoyicocaines. Then in 1994, Moore et al. reported new coca leaf methodology for the concomitant isolation and determination of cocaine and the cinnmoyicocaines along with ecgonine methyl ester, tropacocaine, benzoyltropine, cuscohygrine, the truxillines, and hygrine [139]. This method was applied to three of the four South American cultivars and also coca leaf cultivated in greenhouses and at a non-South American tropical setting. Some of this data is presented in Table 7. Finally in 1994, Casale and Moore described the determination of trimethoxy-substituted tropane alkaloids [34] and pseudococaine [35] in South American ECVC leaf. 3. Section III Studies A necessary condition associated with cocaine signature profiles, especially if used in cocaine comparison/ conspiracy court cases, is that they be able to withstand scrutiny with regard to the specificity of the methodologies used. Ideally, a cocaine signature profile of a given sample should be considered a "fingerprint" of that sample, and be comparable with the uniqueness of human fingerprinting and human DNA genetic profiling. Although that rigorous standard can be currently met by using a combination of different chromatographic cocaine profiling methods [29,34,132,135, 136, 140, 141], improvement/refinement of all methodologies is a continuing goal. To

accomplish this objective, it is necessary to characterize additional coca alkaloid impurities/by-products, usually . at trace levels, in refined illicit cocaine, and then develop methodologies for their quantitative determination. To facilitate this, additional in-depth studies of coca leaf are necessary. This section reviews methodologies developed in 1994-1995 that have allowed for the detection of many new trace-level alkaloids in South American ECVC [27,28,145]. It should be noted that virtually all South American ECVC alkaloids at levels > 0.5% (relative to cocaine) have now been characterized in the leaf, mostly during the past ten years; this includes both qualitative and quantitative analyses. The most recent methodology, discussed in this section, was for the isolation and identification of alkaloids in coca at levels below 0.5% (relative to cocaine). The initial results from this study indicated the presence of at least an additional 125-150 new coca alkaloids, most of which had a 2-carbomethoxy-3-oxotropane or 3-oxo-tropane moiety. In this new methodology [27,28,145], the major and trace-level coca alkaloids were extracted/isolated in their totality from the bulk leaf matrix by extraction with toluene followed by "trap" column chromatography, using a dilute sulfuric acid/Celite column packing. The trace-level alkaloids were isolated from the bulk of the major alkaloids (e.g., cocaine, the cinnamoy icocaines and the truxillines) by partitioning between chloroform and a pH 4 buffer. Both phases were subjected to refinement by either repeated ion-pairing and adsorption (alumina) column chromatography and/or recrystallization. Some fractions were subjected to additional purification via pre-

Table 7. Quantitative results for cocaine and other coca alkaloids in greenhouse- and tropical-cultivated coca leaves a.b Alkaloid Cocaine

Ecvcc 0.54

Ecgonine methyl ester

57

Cuscohygrine

57 0.3 d

Tropacocaine cis-Cinnamoylcocaine trans-Cinnamoylcocaine a b

c

d

7.2 18

ENVT