This file is a part of the Rhodium site archive. This Aug 2004 static snapshot is hosted by Erowid
as of May 2005 and is not being updated. > > Back to Rhodium Archive Index > >
[] [] [Chemistry Archive]

Illicit Production of Cocaine

Casale JF, Klein RFX

Forensic Science Review 5, 95-107 (1993)

HTML by Rhodium


The predominant methods currently used for illicit production of cocaine are described. For illicit natural cocaine (i.e., from coca leaf), this includes production of coca paste from coca leaf via both the solvent and acid extraction techniques, purification of coca paste to cocaine base, and conversion of cocaine base to cocaine hydrochloride. For illicit synthetic cocaine (i.e., synthesized from precursor chemicals), the classic five-step synthetic route used in all clandestine laboratories seized to date is summarized. The origins of the most common alkaloidal impurities and processing/synthetic by-products typically identified in illicit natural, illicit synthetic, and pharmaceutical cocaine are discussed. Forensic differentiation of exhibits arising from the various production methods are addressed both in terms of overall product purity and the presence/absence of these impurities and byproducts.

Table of Contents

  1. Table of Contents
  2. Introduction
  1. Growing and Harvesting of Coca Leaf
  2. Illicit Cocaine Production
    1. Illicit Natural Cocaine
      1. Coca Paste
      2. Coke Base
      3. Cocaine Hydrochloride
    2. Illicit Synthetic Cocaine
  3. Licit (Pharmaceutical) Cocaine Production
  4. Forensic Differentiation of Licit Versus Illicit Cocaine
    1. Illicit Natural Cocaine
    2. Illicit Synthetic Cocaine
    3. Pharmaceutical Cocaine
  1. References


Throughout the 1980s and into the 1990s, cocaine (Structure 1) has been the most widely used "hard" drug of abuse in the United States64. Although recent drug abuse monitors have suggested that illicit cocaine usage in the United States is declining63, worldwide use is still rapidly increasing due to expanding markets in Europe, South America, and the Far East65,66. Because of the disastrous socioeconomic consequences associated with the widespread abuse of cocaine, the United Nations, the United States, and other developed nations continue to commit extensive resources to research and combat this problem. A significant percentage of this effort is directed toward interdiction of cocaine production and smuggling. Such efforts require detailed knowledge of typical production techniques and the analytical profiles of the final products.

However, open scientific research into cocaine production has been severely restricted due to the inherently sensitive nature of the topic. First, cocaine is under strict worldwide legal controls, and requires special permits to possess and/or work with. Secondly, although an extensive amount of research has already been commissioned and completed, the results are often either proprietary or sealed under varying levels of government classification. As a result, there is a critical lack of current, accurate information in the open scientific literature concerning both licit and illicit cocaine processing. This information gap has resulted in extensive duplication of already researched topics and/or misdirection of many research initiatives.

In order to partially address these issues, the authors report detailed descriptions of the most common illicit (i.e., natural and synthetic) cocaine production techniques in current use. Although certain aspects of illicit cocaine processing have been previously summarized (e.g.,21,57,62), to the authors' knowledge this is the first comprehensive, in-depth study of this topic. In addition, the authors briefly discuss analytical profiles for cocaine produced via these techniques which allow for forensic differentiation of seized cocaine exhibits.

I. Growing and Harvesting of Coca Leaf

(-)-Cocaine (cocaine) is a naturally occurring alkaloid found in certain varieties of plants of the genus Erythroxylum. There are over 200 distinct species of Erythroxylum, of which only two, Erythroxylum coca and Erythroxylum novogranatense, contain significant amounts of cocaine. In South America, two varieties within each of these two species are cultivated; these are Erythroxylum coca var. coca (ECVC), Erythroxylum coca var. ipadu (EM), Erythroxylum novogranatense var. novogranatense (ENVN), and Erythroxylum novogranatense var. truxillense (ENVT)7,54,55. Coca cultivation is distributed throughout the central and northern Andean Ridge, with approximately 60% in Peru, 30% in Bolivia and the remainder (in approximate order of importance) scattered throughout Columbia, Ecuador, Venezuela, Brazil, Argentina, and Panama1.

Each of the cultivated varieties of Erythroxylum has a distinct total alkaloidal profile and agricultural range. Of the four, ECVC is the most common cultivar and the source from which most cocaine, both licit and illicit, is derived54. Therefore, its cultivation and harvest are described in detail in the following section.

ECVC contains cocaine (range 0.3 to 1.5%, average 0.8% relative to dry leaf weight) as the principle alkaloid, with approximately 10 to 15% cis- and trans-cinnamoylcocaine and 2 to 3% truxillines relative to cocaine47,56. ECVC, which is botanically classified as a shrub, is readily cultivated in widely varied climates and soil conditions. Its primary agricultural range is throughout the montane tropical forests along the eastern slopes of the Andes, principally from 500 to 1,500 m altitude55. It can live up to 50 years and can grow to a height of up to 3 m, but cultivated plants are commonly pruned to from 1 to 2 m for ease of harvest. After 5 to 10 years, the plants are usually uprooted or cut back to near ground level, reportedly due to decreasing cocaine content in the renewed leaf growth. The highest cocaine contents are generally found in fresh leaves harvested from plants grown at higher, cooler altitudes. In some areas, the plants are commonly interplanted with other crops (corn, yucca, etc.) or in "fallow" fields (i.e., mixed with indigenous grasses and weeds). In addition, various agricultural enhancements, e.g., fertilizers, pesticides, herbicides, irrigation, etc., are sporadically used. The overall effect of such efforts on leaf yield, harvest interval, or cocaine content are currently unknown.

Leaf harvesting is usually not a periodic, "set-piece" operation similar to traditional farming techniques; rather, it is a continuous, ongoing operation usually extending over the entire year - thus providing the farmer with a continuous source of income and a hedge against market fluctuations, which can be severe. Individual plots (i.e., a specific small field or several rows in a large field) are harvested on an average of four times a year. The leaves are comprehensively stripped from the plants by hand. Harvested leaves are usually immediately sun dried on an open-air patio until dry enough to be readily broken up between the fingers. This normally takes 1 to 2 days, depending on the prevailing weather conditions. If the leaf is destined for a nearby illicit laboratory, the drying stage is sometimes skipped. The leaves are frequently raked and turned to aid the drying process, and care is taken to get them undercover immediately if the weather turns threatening. The leaves will ferment (rot) very quickly if they are not dried immediately, especially if they get rain-soaked during the drying process54. Upon sun drying, the fresh leaf loses from two-thirds to three-quarters of its weight due to evaporation of water13; this reduced weight aids eventual transportation. The immediately dried leaf is reasonably stable with respect to cocaine content and decomposition if kept dry and cool5; however, improper handling and/or excessive heat and humidity will result in rapid decomposition56. Dried leaf is typically packaged in 50-pound bags and immediately transferred to a coca market or an illicit laboratory.

Illicit market prices for coca leaf closely track the licit market, but are usually slightly lower. Prices can fluctuate dramatically, not only with normal supply and demand pressures and seasonal supply, but also with the current level of local interdiction efforts by law enforcement. Diversion of leaf from coca markets to illicit cocaine production is common.

Taxonomic studies have shown that ECVI, ENVN, and ENVT each originally derived from ECVC7. ECVI is primarily seen in the lowlands of the western Amazonian basin55,58. It has a much lower cocaine content (average ca. 0.25%56) than ECVC and until recently was primarily cultivated only for chewing by local natives; however, rapidly increasing cultivation has signaled a recent switch into illicit cocaine production54. It has a very low percentage of the cinnamoylcocaines relative to cocaine (approximately 2%56) and probably a correspondingly negligible percentage of the truxillines41. ENVN is primarily seen in Colombia, and is much more tolerant of diverse ecological conditions versus the other cultivars54. Its cocaine content is comparable to ECVC (average ca. 0.8%56); however, it also contains a much higher percentage of the cinnamoylcocaines and truxillines (each typically 40 to 60% relative to cocaine47,55,56). ENVT is primarily seen in the arid northwest areas of Peru54, and is quite similar to ENVN in alkaloid content41,47,55,56. It also has a relatively high percentage of flavonoids versus the other three cultivars, and - although currently supplanted by ECVC - it was cultivated for decades for the soft-drink industry54,55; because of its distinct, non-bitter taste, it remains a very popular leaf for chewing. The extraction and processing of illicit cocaine from ENVN or ENVT is reportedly more difficult than from ECVC or ECVI54 and their cultivation for this purpose is therefore less common.

II. Illicit Cocaine Production

A. Illicit Natural Cocaine

Production of illicit natural cocaine involves three steps:

  1. Extraction of crude coca paste from the coca leaf;
  2. Purification of coca paste to coke base; and
  3. Conversion of coke base to cocaine hydrochloride.

Classically, each of the individual processing steps are accomplished in separate so-called "paste," "base," and "crystal" laboratories (separate meaning anywhere from several meters to several thousand kilometers apart). More recently and increasingly, however, the traditionally separate, sequential paste and base operations are being condensed into direct leaf-to-base laboratories, skipping the isolation of coca paste.

Paste, base, and direct leaf-to-base laboratories represent a deeply entrenched, widespread cottage industry, with thousands of individual operations located throughout the coca-producing regions of South America. In contrast, crystal laboratories are generally much larger, more sophisticated and centralized operations, varying up to semi-industrial pilot-plant type laboratories involving extensive chemical and engineering expertise. They are usually located in remote locales in order to avoid enforcement efforts.

It is important to recognize that there is no one method for obtaining coca paste, coke base, or cocaine hydrochloride. On the contrary, there are numerous procedural variations from lab to lab, especially in the substitution of alternate chemicals. In addition, illicit manufacture of cocaine is not a static situation, but rather is constantly evolving - an evolution that has, in fact, been forcibly accelerated by recent, successful enforcement initiatives. Experimentation with new procedures designed to evade controls on essential chemicals or develop more convenient/less expensive methodologies is common and, in contrast to past secretiveness, new procedures are commonly widely shared. To date, however, the critical elements of cocaine processing remain common to all variants.

1. Coca Paste

There are currently two general methods for processing coca leaves into coca paste, hereafter referred to as the solvent extraction technique and the acid extraction technique. The solvent technique (the traditional methodology) was directly derived from one of the original commercial processes developed in the early 20th century23, and remains the most commonly used method in Peru, Colombia, and Ecuador. The acid technique (a much more recently developed methodology) is a considerably more labor-intensive procedure also directly derived from yet another, even older commercial process59. It requires relatively little organic solvent (which is controlled in certain areas of South America), and is currently the most commonly used method in Bolivia. It should be noted that, to the authors' knowledge, all previous literature reports to date summarizing illicit cocaine processing have only detailed out versions of the solvent technique, i.e., this is the first detailed report of the acid technique.

a. The Solvent Extraction Technique (Scheme 1)

Scheme 1.
Illicit production of coca paste via the solvent
extraction technique (see text for details).

The coca leaves are macerated, dusted with an inorganic base (usually lime or a carbonate salt), dampened with a minimal amount of water, and placed in a maceration pit - typically either a 55-gallon drum or large plastic barrel, a large metal trough or a staked-out pit lined with heavy-duty plastic. Alternately, an aqueous solution of the inorganic base is pre-mixed, then poured over the macerated leaves. If fresh (i.e., not sun-dried) leaf is used, the operators may not add any water. The addition of the inorganic base ensures that the cocaine is in its free base form. A water-immiscible organic solvent (usually kerosene, less commonly diesel fuel or gasoline) is added to the dampened coca leaf slurry and the mixture is either vigorously mixed for several hours or left standing with occasional stirring for up to 3 days, thereby extracting the cocaine free base into the solvent. The efficiency of the extraction is highly dependent on how much time the leaves spend in contact with the solvent and how much effort the operators have put into macerating the leaves (the finer the leaves have been chopped up, the more efficient the transfer of cocaine base to the solvent). Mechanization of the maceration (e.g., with leaf mulchers) and extraction processes (e.g., with washing machines or cement mixers, etc.) is common. In addition, in certain operations the leaves are reportedly repeatedly extracted to ensure more quantitative recovery of cocaine.

After completion of the extraction procedure, the solvent is removed from the mixture either by pressing, filtering, draining from a plug, siphoning or other similar means. The resulting solution is usually completely organic, but may contain a small aqueous layer underlying the organic layer. If necessary, the liquid is re-filtered to remove any remaining vegetable matter and, if two layers remain, the lower (aqueous) layer (which is extremely basic due to dissolved lime or carbonate) is separated by pour-off and siphoning and discarded.

The large volume of organic solvent resulting from the leaf extraction(s) is then back-extracted with a much smaller volume of dilute sulfuric acid, which is added directly to the organic solvent, mixed vigorously for 2 to 10 minutes, then allowed to sit and re-separate. The acid converts the cocaine free base to cocaine sulfate, which dissolves in the aqueous layer. The organic solvent is then separated, leaving only the dilute sulfuric acid solution of cocaine sulfate. This latter yellowish-brown solution is commonly referred to as "agua rica" or "guarapo" (agua rica). The organic solvent is usually re-used indefinitely, with additions of fresh solvent to make up natural attrition due to handling and irrecoverable absorption into the leaf mulch.

In the final phase of coca paste isolation, an excess of base, usually lime, carbonate, or caustic soda, is slowly added to the agua rica solution with stirring. The base neutralizes the sulfuric acid and converts cocaine sulfate back to the free base, which precipitates out of the solution as a gummy, yellowish solid. This solid is coca paste, which is filtered, dried, packaged, and shipped to a base lab.

The cocaine content of coca paste generated by the solvent extraction technique varies from 30 to 80%. It contains numerous additional components other than cocaine, including other coca alkaloids and inorganics. However, most of the free carboxylic acids have been removed because of their limited solubility in dilute acid and solubility in dilute alkali solutions. The dried material usually has a "cakey" consistency and usually will not free-flow easily. Although kerosene and diesel fuel are the extraction solvents of choice, many other water-immiscible organic solvents or solvent mixtures may be substituted. Similarly, while any soluble inorganic base may be effectively used for the neutralization of the agua rica solution, carbonate salts are traditionally the most popular because they act as their own visual endpoint indicators. The addition of any carbonate salt to the acidic solution causes vigorous foaming from the release of carbon dioxide gas; thus, the neutralization endpoint is where the addition of carbonate no longer causes foaming of the reaction mixture. This visual endpoint indicator is very useful to operators without access to sophisticated equipment.

b. Bazuco

A variant of the solvent technique involves the production of bazuco, a crude preliminary run of coca paste with a low cocaine content. Bazuco is often given to paste laboratory workers as payment or co-payment. It is commonly mixed with tobacco and smoked by the user, and represents a very rapidly growing abuse and addiction problem throughout the cocaine-producing regions of South America2,24. In the most common variant, bazuco is obtained by mixing an insoluble diluent (e.g., flour or ground maize) into the dilute sulfuric acid solution prior to back-extraction of the organic solvent. Following extraction, the diluent-slurred aqueous layer is separated from the organic solvent in the previously described manner, and a base is added to the solution just to the point where some initial precipitation is observed. The solution is allowed to stand a few minutes and is then filtered to co-capture the diluent and this initial crude precipitate of coca paste, which is then air dried to give bazuco. Additional base is then added to the filtrate to precipitate the remainder of the coca paste in the usual manner. Chemically, the preparation of bazuco serves two purposes:

  1. The diluent-slurred aqueous solution makes an excellent visual indicator of the interface boundary between the two layers; and
  2. The first precipitate reportedly contains a relatively high content of the cinnamoylcocaines.

Thus, isolation of bazuco reduces the amount of oxidizing agent required in the next step for the production of coke base (vide infra). Coca paste obtained following preliminary isolation of bazuco is purer and usually whiter in appearance.

c. The Acid Extraction Technique (Scheme 2)

Scheme 2.
Illicit production of coca paste via the acid
extraction technique (see text for details).

The coca leaves are placed directly in a maceration pit (almost always a staked-out pit lined with heavy-duty plastic, commonly referred to as a "pozo") containing just enough dilute sulfuric acid to cover the leaves. The leaf/dilute sulfuric acid mixture is vigorously macerated, typically by workers who get in the pit and forcefully stomp the leaves for 1 to 2 hours. The acid converts the cocaine free base in the leaves to cocaine sulfate, which dissolves in the aqueous solution. As with the solvent extraction technique, the efficiency of the extraction depends on how much time the leaves spend in contact with the dilute sulfuric acid solution and how much effort the workers put into stomping the leaves. After the stomping is complete, the acidic coca juice is removed (usually by bucketing) and poured through a coarse filter (to remove any remaining vegetable matter) into a separate decant pit (commonly referred to as a "chiquero"). At this point, an excess of lime or carbonate is added to the isolated dilute sulfuric acid solution with vigorous stirring, thus neutralizing the cocaine sulfate and any remaining sulfuric acid and precipitating a very crude curdled coca paste. The endpoint of the base addition is monitored via spot-testing of small aliquots of the solution with an ethanolic solution of phenolphthalein (called "punto"). The curdled coca paste in the solution is not collectable as such, but is rather back-extracted with a much smaller volume of kerosene, which is thoroughly mixed in for 2 to 10 minutes and allowed to re-separate. After isolation, the kerosene fraction is then handled exactly as in the solvent technique; i.e., the kerosene is back-extracted with a yet smaller volume of fresh dilute sulfuric acid, again generating an agua rica solution.

The acid technique always involves multiple (3 to 5) extractions of the leaves; i.e., the already stomped leaves are treated with another fresh solution of dilute sulfuric acid and re-stomped. Each pozo extract is handled identically in turn, except that the same agua rica solution is used to back-extract all of the kerosene extracts (thus continually enriching its cocaine content). Following processing of the final pozo extract, the isolated agua rica solution is again handled exactly as in the solvent technique; i.e., made basic via addition of an inorganic base, thereby precipitating coca paste.

Coca paste generated by the acid technique is essentially equivalent to that produced via the solvent method, and similarly contains from 30 to 80% cocaine. The advantage of the acid versus solvent technique is the use of a minimal volume of organic solvent; however, it is considerably more labor-intensive. This variant is used extensively throughout Bolivia, where personal possession of large volumes (more than 50 liters) of organic solvents (e.g., kerosene) in the coca-growing regions is illegal.

Chemically, coca paste from either extraction procedure has a gummy consistency and a limited shelf-life. If continuously exposed to excessive heat and humidity, it will slowly self-dissolve, turning into an oily liquid with a pungent, unpleasant odor. This drawback is well known to the clandestine operators; for this reason, coca paste is usually immediately processed to coke base. If this is not possible, it is usually stored as agua rica until further processing is possible.

2. Coke Base (Scheme 3)

Scheme 3.
Illicit production of coke base from
coca paste (see text for details).

Conversion of coca paste to coke base is a purification procedure. As was noted above, the cocaine purity level of coca paste varies from 30 to 80%, depending on the extraction technique, variety of coca, and competence of the operators. The remainder consists of inorganic salts and various alkaloidal impurities, notably cis- and trans-cinnamoylcocaine, which are co-extracted from the leaves. Failure to remove these impurities results in a final product (i.e., cocaine hydrochloride) of poorer quality with respect to cocaine content and especially color and appearance. This is well known among laboratory operators, and as a result, this step is rarely skipped.

Coca paste is first re-dissolved in a small amount of dilute sulfuric acid (thus reconstituting a fresh agua rica solution); as previously noted, the solution has a yellowish-brown color similar to beer. Some operators then slightly increase the pH of the solution with careful addition of base. The solution is then titrated against a concentrated aqueous solution of potassium permanganate, a powerful oxidizing agent. Potassium permanganate gives an intensely purple solution when dissolved in water; as it reacts with the oxidizable alkaloidal impurities in coca paste, it is reduced to manganese dioxide (an insoluble, brown-black solid), which precipitates out of solution. While many operators just add a set volume of concentrated aqueous permanganate to a given weight of coca paste/volume of agua rica (as determined by experience), the more usual method is to slowly add the solution with vigorous stirring, wait a few minutes, and then check to see if the solution has any yellowish-brown color remaining. This is determined by visual inspection of the solution after waiting for the precipitated manganese dioxide to settle out; if the solution is still colored, the addition of the permanganate solution is continued until the solution is finally colorless. Thus, potassium permanganate also acts as its own visual endpoint indicator. Over-addition or too rapid addition of permanganate is known to result in decomposition and loss of cocaine, so the operators work carefully to get it just right.

When the permanganate addition is judged to be complete, the solution is filtered to remove the precipitated manganese dioxide. The resulting colorless, slightly acidic solution (still commonly referred to as agua rica, hereafter oxidized agua rica) is again treated with a solution of base (usually dilute ammonia at this stage) with stirring. Again, the ammonia neutralizes the cocaine sulfate and any remaining sulfuric acid, thereby precipitating purified coke base, which is filtered, dried, packaged, and transferred to a crystal laboratory.

a. Direct Leaf-to-Base Laboratories

In a recently developed and currently quite common variant, both solvent and acid extraction laboratories are being extended to production of coke base. In this alternate, coca paste is never isolated; rather, the unoxidized agua rica solution recovered from back-extracting the kerosene solution is filtered, adjusted (if desired) to higher pH with a carbonate or bicarbonate salt, and then treated directly with the potassium permanganate solution. This is a short-cut technique directly converting coca leaf to coke base, and offers several advantages to the clandestine operators:

  1. There is a net savings of whatever inorganic base is being used to precipitate coca paste and the sulfuric acid required to reconstitute the agua rica;
  2. The previously described difficulties associated with the poor shelf-life of coca paste are avoided (coke base is much more stable than coca paste); and
  3. The operators save a lot of time.

Coke base generally varies from 80 to 95% cocaine. Since potassium permanganate oxidation tends to remove both the cinnamoylcocaines and other colored impurities typically found in coca paste, the appearance of coke base is usually much lighter, varying from light tan to white; in addition, it has a drier, more mobile (free-flowing) consistency versus coca paste.

If too little potassium permanganate is used, an individual coke base exhibit may retain significant levels of cinnamoylcocaines (varying as high as 15% relative to cocaine for coke base derived from ECVC). Conversely, if improper mixing, poor pH control, or excess permanganate is used, cocaine itself may be oxidized to N-formylcocaine, which in turn can be hydrolyzed to N-norcocaine8,10,26,33,60. N-norcocaine can also undergo an intramolecular transamination reaction, giving N-benzoyl norecgonine methyl ester26,60. Thus, poor potassium permanganate oxidation techniques contribute directly to the relative amounts and types of impurities found in the coke base and eventually in the resulting cocaine hydrochloride (i.e., high cinnamoylcocaines with low N-norcocaine and N-formyl cocaine contents or low cinnamoylcocaines with higher N-norcocaine, N-formylcocaine, and N-benzoyl norecgonine methyl ester contents).

b. Alternate Oxidizing Agents

Although potassium permanganate is the most popular oxidizing agent (primarily because of its ready availability and the color change associated with its use), several alternate oxidizing agents have been increasingly reported. The efficacy of these latter reagents is under current investigation at this laboratory.

3. Cocaine Hydrochloride (Scheme 4)

Scheme 4.
Illicit production of cocaine hydrochloride
from coke base (see text for details).

As was previously noted, crystal laboratories mark the switchover from the cottage industry of paste, base, and direct leaf-to-base laboratories to much larger, more sophisticated and centralized operations. Crystal laboratories are usually supplied with coke base either from a specific network of feeder base laboratories or from open-market middlemen. As was previously noted, the quality of the coke base is directly reflected in the corresponding quality of the final product; therefore, all coke base is spot-checked prior to conversion to the hydrochloride. Poor quality base is either returned to the suppliers or re-oxidized (i.e., resubmitted to permanganate oxidation) either on-site or in separate, large-scale re-oxidation laboratories. In some operations, all coke base is re-oxidized as a normal matter of course.

The illicit production of cocaine hydrochloride is not handled in large batches, but rather as a very large number of small batches. Nearly all operations work on a 1 kg scale, with a few varying up to as much as 5 kg/batch. A very large crystal laboratory may have hundreds of individual batches running simultaneously in a 24 h/day operation.

Procedures often vary dramatically from laboratory to laboratory, especially with respect to solvent use. In the classic variant, for each batch, the coke base is dissolved into diethyl ether, filtered or decanted from any remaining insoluble impurities, and an equal volume of acetone containing a stoichiometric quantity of concentrated hydrochloric acid added to the filtrate with stirring. The hydrochloric acid immediately ion-pairs with the coke base to give cocaine hydrochloride, which begins to precipitate out of the solution as shiny white, flaky crystals. The use of excess concentrated hydrochloric acid is avoided due to the development of a distinct yellow color (especially in acetone), which in turn can be partially conferred upon the cocaine hydrochloride; this is unacceptable from a marketing viewpoint. If time is not a critical factor, the resulting solution is allowed to sit from 3 to 6 hours in order to complete the crystallization process. If the laboratory operators are rushed, however, the individual batches are placed in a hot water bath (called a "baño María"), which reduces the total reaction time to approximately 30 min. Use of the baño María technique reportedly results in cocaine hydrochloride of slightly reduced quality with respect to appearance. After completion of the crystallization process, the product is filtered, dried under heat-lamps and/or microwave ovens, pressed, packaged, and shipped to distribution networks. Spent solvents are usually recycled, either on-site or at a separate recycling facility. The insoluble impurities filtered off from the initial diethyl ether solution are not discarded, but rather are re-dissolved in dilute sulfuric acid, precipitated via addition of dilute ammonia and handled as bazuco (vide supra).

As was noted before, diethyl ether/acetone 1:1 is the classic solvent combination for the crystallization process. However, due to the current difficulties in obtaining acetone and (especially) diethyl ether in South America, use of alternate solvents or solvent mixtures for the above A + B addition procedure is quite common. The critical factors in solvent mixture composition are:

  1. Solubility of coke base in solvent A;
  2. Miscibility of solvent B with concentrated hydrochloric acid; and
  3. Insolubility of cocaine hydrochloride in the combined A + B solvent mixture.

Unsubstantiated reports suggest that laboratory operators select solvent mixtures based on density; i.e., by attempting to match the "ideal" densities of diethyl ether (0.715 g/mL), acetone (0.795 g/mL) and diethyl ether/acetone 1:1 (ca. 0.755 g/mL). The most common solvents currently identified in illicit cocaine include (in approximate order of importance): methyl ethyl ketone, toluene, methylene chloride, ethyl acetate, aliphatic hydrocarbons (hexanes, etc.), acetone, benzene, methyl acetate, isobutyl alcohol, and diethyl ether4,28,32. Use of standard industrial, cleaning, or processing solvent mixtures, e.g., ESSO 10/20, is also common. The overall effects of the use of these alternate solvents on the impurity profile of the resulting cocaine hydrochloride is under current investigation at this laboratory.

Illicit, unadulterated cocaine hydrochloride generally varies from 80 to 97% purity, and can vary in appearance from an off-white powder to white, iridescent crystals virtually indistinguishable (visually) from pharmaceutical cocaine. Not unexpectedly, most of the alkaloidal impurities present in the starting coke base are carried through the crystallization procedure and appear in the final product.

Fig. 1. Illicit synthetic cocaine, step 1-312:

  1. Production of 2-carbomethoxytropinone;
  2. Its conversion to Methyl Ecgonine; and
  3. Benzoylation to Cocaine.

Only single enantiomers depicted for simplicity.

B. Illicit Synthetic Cocaine

The classic total synthesis of cocaine involves three synthetic, one enantiomeric resolution and one diastereomeric purification steps (Figure 112,22), and requires a significantly high level of synthetic expertise and well-equipped laboratory facilities. The synthesis will produce a pair of racemic diastereomers (of which only one, i.e., (-)-cocaine, is physiologically active) if the enantiomeric resolution and diastereomeric purification steps are omitted. To date, there have been only three seizures of illicit synthetic cocaine laboratories in the United States. All three followed the classic synthesis; however, none of the three performed the enantiomeric resolution step. Two of these laboratories were run by clandestine operators with advanced chemical training, and successfully produced very low yields of racemic cocaine.

The first step involves a ring coupling Mannich reaction using methylamine, succindialdehyde, and acetonedicarboxylic acid monomethyl ester in high dilution in a buffered, aqueous solution at 25°C. After 2 days, the reaction mixture is made basic and extracted with chloroform to give racemic 2-carbomethoxytropinone; tropinone is the major impurity. Enantiomeric resolution of the racemate can be accomplished at this point with (+)- and (-)-tartaric acid; however, as noted above, none of the operators of the three clandestine laboratories seized to date attempted such a resolution.

In step two, the 2-carbomethoxytropinone is dissolved in a minimal volume of ice-cold dilute sulfuric acid and reduced to methyl ecgonine with a 1 to 1.5% Na/Hg amalgam at pH 3.5 and 5°C. Reaction conditions are critical; poor pH and/or temperature control results in both decarboxylation of 2-carbomethoxytropinone to tropinone (which is, in turn, reduced to tropine and pseudotropine) and C-2 epimerization of methyl ecgonine to pseudoecgonine methyl ester. After several hours, the reaction is made basic, extracted with chloroform, and evaporated to an oil containing methyl ecgonine and pseudoecgonine methyl ester in an approximate 3:1 ratio. Additional impurities usually include tropinone, tropine, pseudotropine and unreacted 2-carbomethoxytropinone. The majority of pseudoecgonine methyl ester is precipitated from the oil by the addition of diethyl ether and removed via filtration. The filtrate is evaporated to dryness, dissolved in diethyl ether and converted to the hydrochloride. None of the operators of the three clandestine laboratories seized to date attempted to purify their methyl ecgonine any further than the pseudoecgonine methyl ester precipitation step.

In step three, the methyl ecgonine hydrochloride is benzoylated with benzoyl chloride in pyridine near 0°C. After 24 h, the reaction mixture is allowed to warm to room temperature and is diluted with diethyl ether, which precipitates a cocaine HCl/pyridine HCl complex. This precipitate is filtered and washed with additional ether to remove excess pyridine, dissolved in water, and extracted with additional ether to remove benzoic acid. The resulting aqueous solution is made basic with dilute ammonium hydroxide (causing dissociation of the cocaine HCl/pyridine HCl complex), and repeatedly extracted with methylene chloride. The combined extracts, which also contain the remaining free pyridine, are evaporated to dryness to give cocaine base, which is re-dissolved in diethyl ether/acetone 1:1 and converted to the hydrochloride via addition of a stoichiometric amount of concentrated hydrochloric acid. As noted above, the clandestine manufacture of illicit synthetic cocaine is extremely unusual. This is not surprising, because - even when attempted by a skilled chemist - the preparation of (-)-cocaine via total synthesis proceeds in less than 10% overall yield. This is clearly economically infeasible in view of the relatively low cost and ready availability of illicit natural cocaine.

III. Licit (Pharmaceutical) Cocaine Production

Pharmaceutical cocaine is a by-product from the industrial extraction from coca of flavoring agents used in the soft-drink industry. The isolation process is proprietary and cannot be detailed in this study; however, it is known to proceed through numerous recrystallization and purification steps. The final product, cocaine hydrochloride, is generally of better than 99.5% purity.

IV. Forensic Differentiation of Licit Versus Illicit Cocaine

Illicit natural cocaine accounts for more than 99.99% of all seized exhibits. Exhibits of illicit synthetic cocaine are extremely rare. Pharmaceutical cocaine is rarely seen and is invariably the result of licit drug diversion or illegal prescriptions. The individual processes used to obtain each type of cocaine are distinct and give products that are chemically unique with respect to the presence and/or relative enhancement or diminution of various impurities. Therefore, detailed forensic analysis can differentiate between all three types.

A. Illicit Natural Cocaine

As previously detailed, the purity of illicit natural cocaine typically varies from 80 to 97%. Virtually all unadulterated illicit natural cocaine contains numerous impurities at levels readily detected by chromatographic and spectrometric techniques3,6,8-11,14,16-20,24-27,29-31,33-53. These impurities include co-extracted coca alkaloids, processing chemicals, and solvents. Additional impurities may also be introduced via chemical modification of cocaine or other coca alkaloids during processing and environmental degradation due to heat and humidity. Finally, various inorganic salts (especially bases) may also be present. Alkaloidal impurities that have been identified at significant levels in illicit natural cocaine include N-acetylnorcocaine, 2,3-didehydroecgonine, 2,3-didehydroecgonine methyl ester, benzoic acid, benzoyl ecgonine, N-benzoyl norecgonine methyl ester, trans-cinnamic acid, cis- and trans-cinnamoylcocaine, cis- and trans-cinnamoylecgonine, ecgonine, methyl ecgonine, N-formylcocaine, N-norcocaine, N-norecgonine, tropacocaine, all five diastereoisomeric truxillic acids, all eleven diastereoisomeric truxillines, and all six diastereoisomeric truxinic acids. Cut samples, of course, may contain a wide variety of additional adulterants and/or diluents. The in-depth chromatographic analysis of illicit natural cocaine was recently reviewed47.

B. Illicit Synthetic Cocaine

The purity of uncut illicit synthetic cocaine can vary dramatically depending on the skill of the clandestine operator performing the synthesis. Illicit synthetic cocaine will not contain many of the alkaloidal impurities commonly identified in illicit natural cocaine, e.g., trimethoxycocaine, the cinnamoylcocaines or the truxillines, but can include any of a wide variety of synthetic by-products (some of which match naturally occurring alkaloidal impurities). Of these, pseudococaine, benzoyltropine and tropacocaine, resulting from benzoylation of pseudoecgonine methyl ester, tropine and pseudotropine, respectively, are the most likely. Additional impurities which are indicative of synthetic cocaine include 3-benzoyloxy-2-carbomethoxytropidine (2,3-didehydrococaine), 3-benzoyloxytropidine (2,3-didehydrotropacocaine), and 2-carbomethoxy-3-methylaminotropidine22. 2,3-Didehydrococaine and 2,3-didehydrotropacocaine result from the benzoylation of unreduced 2-carbomethoxytropinone and tropinone, respectively, and 2-carbomethoxy-3-methylaminotropidine from the irreversible rearrangement of the 2-carbomethoxytropinone/methylamine imine formed during the initial Mannich condensation reaction.

C. Pharmaceutical Cocaine

Pharmaceutical cocaine usually has a purity better than 99.5% and typically has little (if any) coca-related impurities. For example, none of the cinnamoylcocaines or truxillines (the most common alkaloids co-extracted with cocaine from coca leaf) have been detected in pharmaceutical cocaine. The most commonly identified impurities include benzoylecgonine, cocaethylene (ethyl cocaine), ecgonine, methyl ecgonine, and norcocaine. The hydrolytic impurities, i.e., benzoyl ecgonine, ecgonine, and methyl ecgonine, are not a result of the production process itself, but rather arise from degradative hydrolysis of cocaine hydrochloride over time. Cocaethylene results from transesterification of the C-2 carbomethoxy moiety during the initial industrial extraction of the coca leaf15, while norcocaine results from the overoxidation of cocaine base during one of the purification steps.


Authors' Note: The comprehensive reference list for this review would easily surpass 500 citations. Indeed, an "annotated" bibliography on cocaine published in 1988 (current through 1986) lists over 5,000 citations61, and, as an even cursory glance at Chemical Abstracts would confirm, at least half again that many articles have been added to the literature over the last 7 years. The reference list below is therefore suggestive only, and primarily emphasizes more recent advances. Note that many of the selected references include extensive citation lists.

  1. Abruzzese R: Coca leaf production in the countries of the Andean Subregion; Bull Narc 41, 95 (1989)
  2. Agreda, RF: Drug abuse problems in countries of the Andean Subregion; Bull Narc 38, 27 (1986)
  3. Allen AC, Cooper DA, Kiser WO, Cottrell RC: The cocaine diastereomers; J Forensic Sci 26, 12 (1981)
  4. Avdovich HW, LeBelle MJ, SavardC, Wilson WL: Nuclear magnetic resonance identification and estimation of solvent residues in cocaine; Forensic Sci Int 49, 225 (1991)
  5. Balick MJ, Rivier L, Plowman T: The effects of field preservation on alkaloid content of fresh coca leaves (erythroxylum spp.); J Ethnopharmacol 6, 287 (1982)
  6. Baugh LD, Liu RH: Sample differentiation: cocaine example; Forensic Sci Rev 3, 102 (1991)
  7. Bohm BA, Ganders FR, Plowman T: Biosystematics and evolution of cultivated coca (erythroxylaceae); Syst Bot 7, 121 (1982)
  8. Brewer LM, Allen AC: N-Formyl cocaine: a study of cocaine comparison parameters; J Forensic Sci 36, 697 (1991)
  9. By AW, Lodge BA, Sy WW: Characterization of cis-cinnamoylcocaine; J Can Soc Forensic Sci 21, 41 (1988)
  10. Casale JF: N-Acetylnorcocaine: a new cocaine impurity from clandestine processing. I.; Journal of the Clandestine Laboratory Investigating Chemists Association 1, 23 (1991)
  11. Casale JF: Detection of pseudoecgonine and differentiation from ecgonine in illicit cocaine; Forensic Sci Int 47, 277 (1990)
  12. Casale JF: A practical total synthesis of cocaine's enantiomers; Forensic Sci Int 33, 275 (1987)
  13. Casale JF, Meyers RP, Klein RFX: An in-depth study of cocaine base processing in the Chapare Valley, Bolivia; manuscript in preparation.
  14. Casale JF, Moore JM: Determination of pseudococaine in coca leaves and illicit cocaine exhibits; J Chromatogr; manuscript submitted.
  15. Casale JF, Moore JM: An in-depth analysis of pharmaceutical cocaine: cocaethylene and other impurities; J Pharm Sci; manuscript submitted.
  16. Casale JF, Moore JM: 3',4',5'-Trimethoxy-substituted analogs of cocaine, cis-/trans-cinnamoylcocaine and tropacocaine: characterization and quantitation of new alkaloids in coca leaf, coca paste and refined illicit cocaine; J Forensic Sci; in press.
  17. Casale IF, Waggoner RW: A chromatographic impurity signature profile analysis for cocaine using capillary gas chromatography; J Forensic Sci 36, 1312 (1991)
  18. Casale JF, Watterson JW: A computerized neural network method for pattern recognition of cocaine signatures; J Forensic Sci 38, 292 (1993)
  19. Casale JF, Watterson JW: A neural network method for pattern recognition of chromatographic signature patterns of forensic trace evidence; Proceedings of the International Symposium on the Forensic Aspects of Trace Evidence; U.S. Federal Bureau of Investigation Academy: Quantico, VA; in press.
  20. Chiarotti M, Fucci N: HPLC analysis of cocaine diastereomers by chiral stationary phase; Forensic Sci Int 44, 37 (1990)
  21. Cooper DA: Clandestine production processes for cocaine and heroin; in Proceedings of the International Conference on Assessment of Drug Control Issues of Controlled Substance Analogs; Rabat, Morocco; p 95 (1987)
  22. Cooper DA, Allen AC: Synthetic cocaine impurities; J Forensic Sci 29, 1045 (1984)
  23. Duilius: Cocaine; Chem Ztg 54, 31 (1930) [ Chemical Abstracts 24, 3322 (1930) ]
  24. ElSohly MA, Brenneisen R, Jones AB: Coca paste: chemical analysis and smoking experiments; J Forensic Sci 36, 93 (1991)
  25. Ensing JG, de Zeeuw RA: Detection, isolation and identification of truxillines in illicit cocaine by means of thin-layer chromatography and mass spectrometry; J Forensic Sci 36,1299 (1991)
  26. Ensing JG, Hummelen JC: Isolation, identification and origin of three previously unknown congeners in illicit cocaine; J Forensic Sci 36, 1666 (1991)
  27. Easing JG, Racamy C, de Zeeuw RA: A rapid gas chromatographic method for the fingerprinting of illicit cocaine samples; J Forensic Sci 37, 446 (1992)
  28. Fortuna J: Statistical analysis of cocaine head-space vapors; Substance Detection Systems, Vol 2092, Europto Series, Presentation - International Symposium and Exhibition on Substance Identification Technologies; Innsbruck, Austria; Oct. 1993.
  29. Gill R, Abbott RW, Moffat AC: High-performance liquid chromatography systems for the separation of local anaesthetic drugs with applicability to the analysis of illicit cocaine samples; J Chromatogr 301, 155 (1984)
  30. Jane I, Scott A, Sharpe RWL, White PC: Quantitation of cocaine in a variety of matrices by high-performance liquid chromatography; J Chromatogr 214, 243 (1981)
  31. Janzen KE, Walter L, Fernando AR: Comparison analysis of illicit cocaine samples; J Forensic Sci 37, 436 (1992)
  32. Kram TC: Hydrogen-1 nuclear magnetic resonance spectroscopic analysis of organic solvents implicated in illicit cocaine processing; manuscript in preparation.
  33. LeBelle MJ, Callahan SA, Latham DJ, Lauriault G: Identification and determination of norcocaine in illicit cocaine and coca leaves by gas chromatography-mass spectrometry and high-performance liquid chromatography; Analyst 113, 1213 (1988)
  34. LeBelle MJ, Callahan SA, Latham DJ, Lauriault G, Savard C: Comparison of illicit cocaine by determination of minor components; J Forensic Sci 36, 1102 (1991)
  35. LeBelle MJ, Lauriault G, Callahan SA, Latham DJ, Chiarelli C, Beckstead H: The examination of illicit cocaine; J Forensic Sci 33, 662 (1988)
  36. Lewin AH, Parker SR, Carroll Fl: Positive identification and quantitation of isomeric cocaines by high-performance liquid chromatography; J Chromatogr 193, 371 (1980)
  37. Lukaszewski T, Jeffery WK: Impurities and artifacts of illicit cocaine; J Forensic Sci 25, 499 (1980)
  38. Lurie IS, Moore JM, Cooper DA, Kram TC: Analysis of manufacturing by-products and impurities in illicit cocaine via high-performance liquid chromatography and photodiode array detection;
    J Chromatogr 405, 273 (1987)
  39. Lurie IS, Moore JM, Kram TC, Cooper DA: Isolation, identification and separation of isomeric truxillines in illicit cocaine; J Chromatogr 504, 391 (1990)
  40. Medina F: Anhydroecgonine methyl ester in cocaine seizures; Microgram 12, 139 (1979)
  41. Moore JM: personal communication; U.S. Drug Enforcement Administration Special Testing and Research Laboratory: McLean, VA; 1993.
  42. Moore JM: The application of chemical derivatization in forensic drug chemistry for gas and high performance liquid chromatographic methods of analysis; Forensic Sci Rev 2, 79 (1990)
  43. Moore JM: Determination of total truxillines in illicit cocaine seizures using capillary gas chromatography-electron capture detection; in Proceedings of the International Symposium of the Forensic Aspects of Controlled Substances; U.S. Federal Bureau of Investigation Academy: Quantico, VA; p 191; 1988.
  44. Moore JM: The application of derivatization techniques in forensic drug analysis; in Klein M, Kruegel AV, Sobol SP (Eds): Instrumental Application of Forensic Drug Chemis-try; U.S. Department of Justice, Drug Enforcement Administration: Washington, DC; p 180 (1978)
  45. Moore JM: Gas chromatographic detection of ecgonine and benzoylecgonine in cocaine; J Chromatogr 101, 215 (1974)
  46. Moore JM: Identification of cis- and trans-cinnamoylcocaine in illicit cocaine seizures; J Assoc Off Anal Chem 56, 1199 (1973)
  47. Moore JM, Casale JF: The in-depth chromatographic analyses of illicit cocaine and its precursor, coca leaves - recent advances; J Chromatogr; in press.
  48. Moore JM, Cooper DA: The application of capillary gas chromatography-electron capture detection in the comparative analyses of illicit cocaine samples; J Forensic Sci 38, 1286 (1993)
  49. Moore JM, Cooper DA, Lurie IS, Kram TC, Carr S, Harper C, Yeh J: Capillary gas chromatographic-electron capture detection of coca-leaf-related impurities in illicit cocaine: 2,4-diphenylcyclobutane-1,3-dicarboxylic acids, 1,4-diphenylcyclobutane-2,3-dicarboxylic acids and their alkaloidal precursors, the truxillines;
    J Chromatogr 410, 297 (1987)
  50. Moore JM, Meyers RP, Jimenez MD: The anatomy of a cocaine comparison case: a prosecutorial and chemistry perspective; J Forensic Sci 38, 1305 (1993)
  51. Noggle FT Jr, Clark CR: Liquid chromatographic analysis of samples containing cocaine, local anesthetics, and other amines; J Assoc Off Anal Chem 66, 151 (1983)
  52. Noggle FT Jr, Clarke CR: Liquid chromatographic identification of cis- and trans-cinnamoylcocaine in illicit cocaine; J Assoc Off Anal Chem 65, 756 (1982)
  53. Olieman C, Maat L, Beyerman HC: Analysis of cocaine, pseudococaine, allococaine and allopseudococaine by ion-pair reverse-phase high-performance liquid chromatography; Recl Trav Chim des Pays-Bas 98, 501 (1979)
  54. Plowman T: The ethnobotany of coca (erythroxylum spp., erythroxylaceae); Adv Econ Bot 1, 62 (1984)
  55. Plowman T: Amazonian coca; J Ethnopharmacol 3, 195 (1981)
  56. Plowman T, Rivier L: Cocaine and cinnamoylcocaine content of erythroxylum species;
    Ann Bot 51, 641 (1983)
  57. Schlesinger HL: Topics in the chemistry of cocaine; Bull Narc 37, 63 (1985)
  58. Schultes RE: Coca (erythroxylon coca ipadu, ethnobotany, cultivation, processing) in the Northwest Amazon; J Ethnopharmacol 3, 173 (1981)
  59. Sperber O: Cocaine manufacture in Peru; Tropenpflanzer 15, 684 (1911)
    [Chemical Abstracts 6, 1494 (1912)]
  60. Stenberg VI, Narain NK, Singh SP, Parmar SS: An improved synthesis of norcocaine;
    J Heterocycl Chem 13, 363 (1976)
  61. Turner CE, Urbanek BS, Wall GM, Waller CW: Cocaine; University Press of Mississippi: Jackson, MS; 1981.
  62. U.S. Department of Justice, Drug Enforcement Administration, Office of Intelligence: Coca cultivation and cocaine pocessing: An overview; Arlington, VA; 1993.
  63. U.S. Department of Justice, Drug Enforcement Administration, Intelligence Division: The National Narcotics Intelligence Consumers Committee (NNICC) Report, 1992; Arlington, VA; 1993 (and prior editions dated 1985-1991).
  64. U.S. Department of Justice, Drug Enforcement Administration, Intelligence Division: The Illegal Drug Situation in the United States; Arlington, VA; 1993.
  65. U.S. Department of Justice, Drug Enforcement Administration, Intelligence Division: Cocaine Situation in Europe; Arlington, VA; 1992.
  66. U.S. Department of State, Bureau of International Narcotics Matters: International Narcotics Control Strategy Report; Washington, DC; 1993.

About the Authors

J.F. Casale & R.F.X. Klein

John F. Casale earned his B.S. in chemistry from Appalachian State University, Boone, NC, in 1981, and immediately entered the North Carolina State Bureau of Investigation Forensic Drug Laboratory in Raleigh. He joined the Special Testing and Research Laboratory in 1992 as a Senior Forensic Chemist with the Research Group. His current interests lie in cocaine signature analysis and the isolation and identification of tropanoid alkaloids in coca and cocaine.

Robert F. X. Klein earned his Ph.D. in synthetic organic chemistry from Georgetown University, Washington, DC, in 1985. His thesis detailed novel syntheses of the S-pseudoazulene thialene and various oxazole-based nonsteroidal anti-inflammatory agents. Dr. Klein joined the Special Testing and Research Laboratory in 1987 as a Senior Forensic Chemist, and is currently the Supervisory Chemist of the Research Group. His current interests lie in illicit cocaine, heroin, and methamphetamine processing and the synthesis of designer drugs.