The Analysis of Heroin


Colour Tests
Microscopic Tests


Pages: 27 to 35
Creation Date: 1953/01/01

The Analysis of Heroin


Heroin is now by far the most important drug of addiction, at least in the Western world. In the Far East, opium itself may be more important; and in some parts of the world cannabis may claim more users, but cannabis addiction is far milder, and often only a stepping-stone to heroin.

The great advantage of heroin, from the point of view of illicit traffickers, is that it seems capable of almost unlimited dilution with other substances, while still the addict obtains from it some gratification for his desire. Nearly every small peddler who resells heroin first mixes his supply with at least an approximately equal amount of lactose or powdered sugar. As his heroin is usually highly adulterated and diluted even before he gets it, it frequently reaches the final user containing no more than five per cent of the actual alkaloid, or even sometimes as little as one per cent.

A word here as to nomenclature. "Heroin," now pronounced by most English-speaking addicts as "heroyn" (two syllables), will be used in this article for the drug as it is found in the illicit traffic and used in addiction-with all its adulterations, impurities, and dilutions. It is also to be understood that this illicit "heroin" is actually a salt of the base, namely the hydrochloride. When the name is used as a real synonym for diacetylmorphine, there is no reason to deny it the regular termination "-ine" (in most cases, including this one, pronounced as in "machine") which is applied to the names of the alkaloids and bases of the same type in English and in French. In this case the base is "heroine" and the usual salt, "heroine hydrochloride." One may still speak of somewhat impure heroine, containing small amounts of acetylcodeine and papaverine derived from the original opium, and of poorly acetylated heroine, containing more or less monoacetylmorphine and residual morphine. In this article when it is necessary to refer quite specifically to the pure substance, it will be called diacetylmorphine. It may be best to caution that these distinctions in nomenclature are not generally recognized.

Heroine is manufactured from morphine by a process of acetylation. Nowadays, the whole process is often illicit. First, the opium is either produced illicitly, or if legally produced in one of the opium-producing countries, is diverted from governmental control. Then the morphine is extracted from it in a clandestine "kitchen laboratory." Finally, the morphine is converted to heroine by a laboratory process usually involving the use of acetic anhydride and benzene.

Generally the operators are not skilled chemists, and cases are known of their blowing themselves up or poisoning themselves with benzene fumes. No manufacture of heroine is lawful in the United States of America, but it is permitted in some countries under the same conditions as for other narcotics under international control. Some of the heroine on the illicit market is manufactured in legal drug factories and then diverted into illicit channels. Such heroine is distinguished by its greater purity and better acetyla-tion; this does not mean that it is any more free from adulterants and diluents by the time it reaches the addict-consumer, but it is relatively free of acetyl-codeine, papaverine, monoacetylmorphine, and morphine, which are unintentional impurities.


In examining a "Heroin" sample in a criminal case, the primary necessity is, of course, to identify the diacetylmorphine with complete certainty, or to identify the narcotic, whatever it may be, if the name "heroin" proves to be a misnomer. The quantitative determination of the narcotic ordinarily does not affect the criminal liability, but is usually desired by the enforcement officers. Moreover, the identification, and, when feasible, the quantitative determination of the various diluents, adulterants, and impurities that may be present, supply information that may be useful to enforcement authorities in tracing connexions in the illicit traffic, and in keeping abreast of developments in the use of narcotics. For example, the discovery of the presence of methadone in certain samples may be of interest in both ways, for methadone is one of the new synthetic analgesics which is also under international narcotics control. This discovery was recently made in New York, and also in Chicago, but the distribution of methadone-containing samples was not very wide, and it is too early to say whether these cases indicate a whole new trend in adulteration or mixing of narcotics for sale to addicts.


The various substances that may be found in a heroin sample, besides diacetylmorphine, may supply, in favourable cases, successive "tracers" possibly connecting up street-sellers, wholesalers, manufacturers, and finally even those who recover the morphine from the original opium.

The substances may be classified as follows:

A. DILUENTS. Generally lactose or powdered sugar, but sometimes mannitol or some other substance may be used. Usually some diluent is added by the wholesaler or even the manufacturer; the street-seller also mixes in some diluent, often a different one. When mannitol was used, it was the first diluent, and subsequent dilutions had been made with lactose or sucrose, or both.

B. ADULTERANTS. While the diluents may also be called adulterants in the usual meaning of the term, it is convenient to distinguish the substances which are merely diluents from the drugs of more or less activity which are added either to conceal the comparative lack of real heroine, or to enhance its effect. Quinine is the most common adulterant and probably occurs in more than half the heroin samples in the United States and Canada. It is probably added, for one reason, to enhance the bitter taste of the mixture, in case the addict tries to assay the strength by tasting. It is most often added as the hydrochloride; sometimes as the sulfate. Some years ago, procaine was very commonly used. The amount of quinine usually equals or exceeds that of heroine, but the procaine was only about one-sixth to one-third the amount of the heroine, and possibly it was used chiefly for its effect as a local anaesthetic. Other adulterants that have been found are barbiturates, caffeine, acetphenetidin, methadone, and amphetamine. The adulterant is generally added by the manufacturer or the wholesaler, and thereafter mere dilution is the rule.

C. IMPURITIES OF MANUFACTURE. These include monoacetylmorphine and morphine, remaining from imperfect acetylation. The absence of such impurities may be significant of diversion from legal manufacture. Some years ago a series of samples analyzed in New York and said to be heroin coming from the eastern Mediterranean were found to contain a small percentage of calcium carbonate. This probably resulted from the manufacture, but it was impossible to say whether it was incidental or accidental, or a deliberate addition.

D. IMPURITIES OF ORIGIN. These include codeine (present as acetylcodeine after the acetylation), papaverine, meconic acid, and brown colouring matter. The absence of these substances may not mean much, but their presence proves that a rather crude process was used, not only in the manufacture and purification of the heroine, but before that, in the extraction of the morphine. Some opiums are so low in papaverine or codeine that a relatively high content of one or the other of these alkaloids may even tell something about the kind of opium used. Meconic acid can only be found in rather crude heroine. A brown heroine generally means crude manufacture, but it is possible that a brown colour is left in deliberately in some cases. Even a nearly pure diacetylmorphine hydrochloride is not a pure shining white like lactose. The brown heroine can be considerably diluted with lactose and look more like "pure heroine" than if it were shining white.



Colour Tests

  1. Marquis' reagent (Formaldehyde in concentrated or (6 + 1) H 2SO 4)1 -purple red changing to purple.

  2. Frohde's reagent (Molybdate in concentrated H 2SO 4)2-strikes violet,quickly changing to strong purplish red, fading out to weaker brown or brownish, then developing green.

  3. Mecke's (or Lafon's) reagent (Selenious acid in concentrated H 2SO 4)3,4- green,quickly greenish blue, changing to blue,slowly to bluish green with yellow-brown edge, then olivaceous green.

    Comment on tests 1 to 3: These sulfuric acid reagents provide highly characteristic colour tests for the spot-plate, and most adulterants and diluents do not seriously interfere. Rarely, a sample may contain so little heroine that charring of the diluent sugar with the concentrated acid will obscure the colours. Of course in that case, a separation is necessary. Imperfect acetylation does not matter as morphine and monoacetylmorphine give the same colours. The initial colour with the hydrochloride is slightly different from that with the free alkaloid or sulfate, bluer with Frohde's reagent and more yellow-brown with Mecke's; and this effect is increased if additional chloride is present. Chloride or hydrochloride causes some effervescence with concentrated sulfuric acid. The inexperienced analyst should compare the colours with those yielded by known diacetylmorphine hydrochloride.

  4. Nitric acid-light yellow solution, gradually bright green.5Concentrated HNO 3 is usually used, but a (4 + 1) acid is somewhat better (4 parts concentrated HNO 3mixed with 1 part H 2O).

    Comment: This highly characteristic test is not very sensitive and often enough it may not be obtained on adulterated, diluted samples. Morphine gives an orange red color fading to yellow; sometimes with a little morphine present a red-orange is obtained at first and later the green of diacetylmorphine develops. Monoacetylmorphine is similar to morphine.

  5. Copper test. To a little of the powder on the spot plate add several drops of water, 2 or 3 drops of 3% H 2O 2, a drop or two of NH 4OH and stir with a piece of copper. A pink to red color is produced.

    Comment: The reaction was originally discovered by Deniges; the form of the test used here was introduced by Oliver.6,7 The test is given by various phenols and their acetyl derivatives including morphine and diacetylmorphine.8The stirring with copper should cause the formation of bubbles of oxygen, and in the absence of diacetylmorphine or other reactive compound the solution should gradually turn blue; if this does not occur, the peroxide solution was too weak.

Microscopic Tests

  1. Platinum chloride, H 2PtCl 6. A little of the powder being examined is dissolved in a drop of water, or better in diluted acetic acid, on the microscpoe slide, and a drop of the 5% reagent solution is added, then crystals are looked for under the microscope. They form gradually and are needles in rosettes, like cokleburs; they form best at a dilution near the limit of amorphous precipitation. The crystals grow larger and form blades in rosettes in the presence of acetic acid, which also diminishes the interface of quinine, and may be used up to (1+1) strength. (Lews, New York International Revenue Laboratory.)

    Comment: This test was probably originated by Putt in 19129; it is the best known crystal test for diacetyl-morphine, and has long been a favourite for identification. The crystals are very characteristic, but the test is much more easily spoiled by the presence of other alkaloids than either of the two following.

  2. Mercuric iodide in HCl. (The solvent solution contains about 27% by volume of concentrated HCl, or 10% by weight of HCl, and is saturated with Hgl 2.) This reagent is applied to a little of the dry substance on the microscope slide. The crystals are branching threads and splinter-plates.10,11

    Comment: The test is extremely sensitive and usually succeeds even with highly adulterated samples, though since the crystals are colourless, care must be taken to distinguish them from any undissolved particles of the powder. The reagent can also be applied to the aqueous solution.

  3. Gold H Bromide in (2 + 3) H 2SO 4. A drop of reagent is added to a drop of aqueous solution of the substance on the microscope slide. The crystals are fine needles, mostly scattered, but often partially in sheaves or rosettes. The interference of quinine is much diminished by the presence of acetic acid. Use a very little of the powder scattered on the slide; add a drop of 20% (or stronger) acetic acid solution; do not stir, but after a moment add a drop of the reagent.11

    Comment: The test is extremely sensitive, and is best obtained on a very dilute solution, containing only just enough diacetylmorphine for a light amorphous precipitate when the reagent is added. Caffeine yields somewhat similar crystals with this reagent.

Other recommended microcrystal tests for diacetyl-morphine can be made on the aqueous solution with 1:20 HAuCl 4 in (1 + 1) H 2SO 4,11 half-saturated sodium picrate, and either Na 3PO 4 or Na 2CO 3. The last mentioned reagents precipitate the free base, which soon crystallizes.

More tests have been mentioned than are really necessary. It is considered that with good results in tests 1, 2, 3, 7, and 8, the diacetylmorphine may be considered as positively identified, without the necessity of separating it from accompanying impurities, adulterants, and diluents.


  1. Refractive indices. The best method of positively identifying at least the predominant diluent is by determining its refractive indices, using the polarizing microscope.12,13 The ordinary lactose is the alpha hydrate, having indices of 1.517, 1.542, and 1.555. Characteristic, rather large pieces are usually seen which are roughly triangular and show bright colour bands with crossed nicols; these show the highest index lengthwise and the lowest crosswise. Many other pieces show 1.555, but few except these triangular pieces show the lowest index. Beta lactose was found in one sample examined at the Chicago Internal Revenue Laboratory; the indices are 1.542, 1.572, and 1.585.13 Sucrose indices are 1.538, 1.565, and 1.571; the highest and lowest are easily found on bright pieces. Mannitol indices 1.520, 1.555, 1.558.14 It occurs in rods, which uniformly show the highest index lengthwise. The lowest index can be found on a few of the brightest rods, crosswise. The middle index can be found crosswise on crystals appearing a weak gray with crossed nicols; although the crosswise index of such crystals sometimes seems to be a little less (about 1.550), with an index solution of 1.555 numerous crystals can be found with both indices nearly equal to that of the liquid. By the refractive index method mannitol can be positively identified even in small percentages in the sample.

  1. Reduction. with Fehling's solution. Dissolve about 20 mg (weighed-not more) of the powder in 4 to 5 cc water. Split into two parts and add 2 drops (1 + 1) HCl to one. Heat both, then add 2 cc Fehling's solution to both and heat. Hydrolyzed sucrose-complete reduction; unhydrolyzed sucrose-none. Lactose gives considerable but incomplete reduction, only a little more in the solution heated with acid than in the other.

  2. Georgia and Morales test.15 The polyhydroxy alcohols, such as mannitol and sorbitol, yield formaldehyde on oxidation, and a test for them can be made with essentially the same procedure used for methanol in distilled spirits-oxidation by permanganate, decolorization by oxalic acid in the presence of sulfuric acid, and restoration of the colour of Schiff's reagent (decolourized fuchsine solution).


  1. Quinine or other fluorescent compounds. A little of the powder is dissolved in dilute sulfuric acid in a test tube or on the microscope slide and observed under ultraviolet light. The General Electric 250 watt "Purple X" lamp is especially useful for such observations. Nupercaine also fluoresces strongly.

  2. Herapathite crystals16 for quinine identification. A little of the powder is treated on the microscope slide with a drop of iodine reagent C-3, and a cover glass applied. If quinine is present the herapathite crystals usually form quite readily and are easily recognizable by their properties of pleochroism, especially with the polarizing microscope. The reagent, a variant of Iodine reagent C-l,17 and easier to make, contains: Concentrated I:KI solution (10 grams iodine plus 10 grams KI dissolved in a little water and then diluted to 100 cc; this can be kept as a stock solution)-0.4 cc, glacial acetic acid-2.5 cc; water-2.6 cc, (1 + 3) H 2SO 4-0.5 cc. (Total 6 cc.) This reagent will keep for a week or so in a rubber bulb flask.

  3. Sanchez test for procaine and other primary aromatic amines.18,19. A solution of furfural in acetic acid, applied to the dry powder on the spot plate, yields a bright red colour with any speck of procaine or other primary aromatic amine. One part of freshly distilled furfural in 3 parts alcohol does not become badly discoloured during storage in a refrigerator and may be kept as a stock solution. One part of this solution mixed with 4 parts glacial acetic acid makes a good reagent.

  4. Chromium sulfate-chloride reagent for methadone.20This reagent can be applied either to a drop of solution or directly to a little of the dry powder on the microscope slide, and will readily yield the methadone crystals even in the presence of much more heroine or quinine or both.

Other adulterants are less common but have been found and identified, in certain samples, in the course of quantitative examinations.


In discussing the nitric acid test for diacetylmorphine, it has already been mentioned that this may show the presence of morphine or monoacetylmorphine by an orange colour. The following tests may also be used:

  1. FeCl 3. To the solution on a spot-plate add 1 drop of 10% FeCl 3 solution. Morphine or monoacetylmorphine yields a green to blue colour, but it is not a very sensitive test. The test is due to the free phenolic hydroxyl, which is possessed by monoacetylmorphine and morphine but not by diacetylmorphine, which yields no reaction in this test.

  2. Iodic acid tests. To two "spots" of the solution add a drop or two of 5% HIO 3 solution, and to one of the "spots" add 2 or 3 drops of H 20 2solution; then add NH 4OH to both. The one without peroxide turns dark brown (Lefort test),21and the one with peroxide, red.22These are phenolic reactions, to which morphine and monoacetylmorphine are sensitive. In diacetylmorphine the phenolic hydroxyl is acetylated, and it gives nothing, or at most a slight reaction due to partial hydrolysis.


  1. Ferric chloride test for meconic acid. Some of the powder is dissolved in water, the solution strongly acidified with HCl and shaken two or three times with ether. The separated ether extracts are shaken with a little dilute ferric chloride solution, which turns red if appreciable meconic acid was present.

There do not seem to be any simple tests in use for codeine (acetylcodeine) or papaverine in the presence of the other constituents of samples. These have been identified after separation in the course of quantitative determinations of the alkaloids present.


  1. Chlorides and sulfates. Test a drop of the solution of the "heroin" on the microscope slide with AgNO 3 solution, to confirm the presence of chlorides (the heroine seems always to be present as hydrochloride). Test another drop with BaCl 2solution to see if any sulfates are present.

  2. Carbonates. If any carbonate is present, the powder will effervesce with dilute acid.

  3. Calcium. Treat some of the powder on the microscope slide with several drops of water and add a drop of concentrated H 2SO 4 at the side of the aqueous solution. As the liquids mix, needle crystals soon form, some of which gradually grow into highly characteristic plates.23


The percentage of diluents is ordinarily determined merely by difference; of course allowing for the fact that the alkaloids are present as their salts, often including water of crystallization, e.g., diacetylmorphine hydrochloride 1H 2O.

There are so many possible adulterants that it is not possible to prescribe any general method of analysis which can always be used to determine them, and methods are generally varied in accordance with the qualitative indications. In trying to make a fairly complete analysis, it is advisable first to get some measure of the total of active ingredients, consequently some methods of this nature are given first.


  1. Titration of amine-combined acids. Treat a weighed portion of the heroin sample, usually a gramme or two of the mixture with adulterants and diluents (preferably enough to contain approximately 0.3 to 0.4 gramme of alkaloidal salts), with 15 to 25 cc water (recently boiled); then dilute with 100 cc neutral ethanol. Titrate with n/10 NaOH, using alpha naphthol benzein indicator (0.1 to 0.2 cc of 1% alcoholic solution). In alcoholic solution, and with a suitable indicator, the salts of amines, and ammonium salts, titrate just like free acids.24The actual free acidity can be determined by another titration, in water, using methyl red indicator. This merely serves as a slight correction to the amine titration, unless some acidic adulterant is disclosed.

  2. Titration of chlorides. This is accomplished in the usual way with standard silver nitrate solution, and potassium chromate as indicator. It may yield the total equivalent of the alkaloids, if all are present as hydrochlorides, or it may show the equivalent of the heroine in mixtures containing heroine hydro-chloride and quinine sulfate. Of course neither this nor the preceding titration discloses such an adulterant as caffeine.

  3. Weight of alkaloidal residue. A solution of a weighed quantity of the powder is made ammoniacal and extracted with chloroform, or if much morphine is present, with 3: 1 chloroform-isopropanol. The solvent is evaporated off on the steam bath, the residue cooled and weighed. This will not include barbiturates, which have to be extracted from acid solution. In the case of a complex mixture it is advisable to acidify the aqueous solution and reextract to see what may be obtained. (See the following procedures 4 and 5.)


  1. Extraction of alkaloids with separation of caffeine, morphine, and barbiturates. A solution of aweighed quantity of the powder is made ammoniacal and extracted with CCl 4. The solvent is evaporated, the residue cooled and weighed. When necessary, this residue can be further examined by methods 6, 7, 8, given below. This accounts for most alkaloids, but not morphine25 or caffeine. The aqueous solution is then acidified and extracted with CHC1 3. If a small portion of the extract, on being evaporated, leaves a residue, the CHCl 3 is shaken with dilute ammonia, then evaporated. Any residue now left is likely to be caffeine. If barbiturates were extracted from the acid solution they would be held by the ammonia, and can be recovered by again acidifying and extracting. If the qualitative tests indicated morphine, the original solution can be made ammoniacal again and extracted with chloroform-isopropanol (3: 1). This procedure can be somewhat expanded to correspond to the following one, using CCl 4 instead of ether.

  2. Ether-chloroform separation into extractive classes. This procedure is suggested for highly adulterated samples to discover what substances may be present.

    1. Treat the powder with dilute HCl and extract with ether. Shake the combined ether extracts with dilute NH 4OH, then evaporate the ethereal solution. Look for such substances as barbital or phenobarbital in the aqueous ammonia, substances such as acetphenetidin in the residue from the ether.

    2. Make the original HCl solution ammoniacal and shake out with ether. Evaporate the ether and weigh the residue. This is the main alkaloidal residue and contains the diacetylmorphine and most other alkaloids that may be present. It may have to be examined further by other methods (e.g., method 6 following; or method 7, below).

    3. Make the original solution again acid and extract with CHCl 3. Examine for substances such as caffeine.

    4. Make the solution again ammoniacal and extract with 3: 1 chloroform-isopropanol. Examine for morphine.


  1. Hydrolysis method for morphine derivatives. Often the most practicable way to determine diacetylmorphine in the presence of some other alkaloid is to hydrolyze it to morphine, which can then be separated by means of its phenolic property of solubility in alkali solution. This method can be used to determine the morphine equivalent of the total of diacetylmorphine, monoacetylmorphine, and morphine, in the usual adulterated diluted samples. As carried out by Shaffer at the U. S. Internal Revenue Laboratory at Chicago, it is as follows: Treat a weighed portion of the sample with 20 cc 1 n H 2SO 4. Reflux 1 hour (or heat in a beaker on the steam bath, replacing evaporated water occasionally). Transfer to a separatory funnel. Add 1.5 cc 1: 1 NaOH. Shake out 3 times with equal volumes of CHCl 3 to remove alkali-insoluble alkaloids. Wash the combined CHCl 3 extracts once with 5 cc 2% alkali solution. Shake the main solution once more with CHCl 3 and shake the CHCl 3 with the same aqueous wash. Evaporate the CHC 3 and weigh the residue, if the non-morphine alkaloids are to be determined. Combine the aqueous solutions. Add concentrated HCl to acidity and add 8 drops in excess. Add concentrated NH 4OH till basic, then 2 to 4 drops in excess. Shake three times with equal volumes of 3:1 chloroform-isopropanol. Wash the combined extracts with 1 to 2 cc water. Extract a fourth time with chloroform-isopropanol and shake the extract with the same wash water. Filter the chloroform-isopropanol extracts through a pledget of cotton into a dish or an Erlenmeyer flask. Evaporate the solvent. Dissolve the morphine residue in methanol. Titrate with n/10 H2SO4 until methyl red indicator begins to show red. Add boiled water and finish the titration.


  1. Separations based on chloroform extraction from HCl solution. The following method was originally devised chiefly as a means of separating diacetyl-morphine as such from quinine (without hydrolyzing the former to morphine). However, its utility goes far beyond this separation.


  1. Put a weighed quantity of the sample, estimated to yield not over 0.2 gramme diacetylmorphine (generally 1 to 5 grammes of the usual adulterated and diluted sample), into a separatory funnel.

  2. Treat with 5 to 15 cc cold diluted HCl (10% by volume of concentrated HCl). The larger amount is used when it is necessary to take 5 to 10 grammes of the sample because of an extremely low content of diacetylmorphine. Some undissolved lactose can be disregarded.

  3. In a second separatory, put 2 to 2.5 cc cold diluted HCl; in a third separatory, 25 cc water plus 0.5 cc concentrated NH 4OH solution; in a fourth separatory, 25 cc water.

  4. Extract 5 times, with 30 to 20 cc chloroform each time. Filter through cotton in the stem of the separatory into the second separatory.

  5. Shake each extract in turn with the wash HCl solution in the second separatory and filter into the third separatory. Also shake each with the ammo-niacal solution and filter into the fourth separatory. Also shake each extract with the wash water, and finally filter into a weighed glass dish or beaker.

  6. Finally, shake the wash HCl in separatory #2 with an additional 20 cc chloroform, and take this extract along to separatory #4, withdrawing all the chloroform from #3 and finally from #4, and filtering into the weighed dish or beaker.

  7. Evaporate the chloroformic extracts. The residue is primarily diacetylmorphine; it frequently also contains acetylcodeine, and may also contain papaverine, methadone, or caffeine. (See E.)

B. (8) Barbiturates, if present, will be in the ammoniacal solution. They may be separated by acidifying and extracting with chloroform.

C. (9) If the determination of quinine is the prime consideration at this stage, it is only necessary to combine the HCl solutions, make alkaline with NaOH, and extract the quinine with chloroform. The aqueous solution can then be reacidified, made ammoniacal, and any morphine present extracted with 3:1 chloroform-isopropanol. Such morphine is actually the morphine equivalent of the monoacetyl-morphine and morphine of the original sample. Codeine, if present as such in the original sample, would be mostly left with the quinine, and extracted with it, but this very seldom occurs, as apparently the codeine is practically all acetylated in most samples. Procaine, if present, is left in the HCl solution by the extractions under A. It can be determined by method 8, below.

D. (10) For poorly acetylated samples, combine the HCl solutions after the extraction given under A, but instead of making alkaline with NaOH as in C, neutralize the acid gradually with NH 4OH while cooling, and make slightly ammoniacal. Then extract with CCl 4. This will extract the monoacetyl-morphine, leaving the morphine in the aqueous layer. The morphine can then be extracted with 3:1 chloroform-isopropanol. Quinine, if present, will come out with the monoacetylmorphine, and after weighing or titrating the residue the quinine can be extracted from alkali solution, and the monoacetylmorphine (or morphine, if it has become hydrolyzed) extracted from ammoniacal solution with chloroform-isopropanol, titrated, and the titration calculated as monoacetylmorphine. Procaine, if present, would also come out in the CCl 4 in this step.

E. (11)In order to determine the various alkaloids in the residue from the chloroform extracts of procedure A, the hydrolysis method is used. When the sample has evidently been made from an impure morphine, the papaverine and codeine are of interest (the latter comes from acetylcodeine reconverted to codeine by hydrolysis), and they are separated from the morphine by shaking out the solution, after making it alkaline, with CCl 4. Methadone, if present, is separated from the morphine resulting from the hydrolysis in the same way. No detailed studies have yet been made on its separation from codeine and papaverine.

(12) In the case of codeine and papaverine, the alkaloids are recovered from the CCl 4 by shaking with dilute H 2SO 4 solution. Papaverine is then shaken out of the dilute H 2SO 4 solution with chloroform. The chloroform extracts are washed with dilute ammonia solution and water, and evaporated. The H 2SO 4 solution is then made alkaline and extracted for codeine.

(13) Usually, it is sufficient to determine the diacetylmorphine by difference, by merely subtracting the weights found for papaverine and codeine from the weight of residue found at step (7). The aqueous solution of step (12) can be reacidified, and shaken with CHCl 3 to remove any caffeine that may still be present (after acid hydrolysis). (To determine caffeine, however, it is better to process the original sample as in procedure 4 or 5.) The aqueous solution can then be made ammoniacal and extracted with 3: 1 chloroform-isopropanol to recover the morphine, from which the original diacetylmorphine can be calculated.


  1. Procaine and volatile bases from alkali.26 Usually the best way to determine procaine in heroin samples is by distillation from alkali. Procaine decomposes and a volatile amine distills over. Weigh a suitable quantity of the sample into a flask, add 150 cc water, a few pieces of granulated zinc, and 3 cc of 1:1 NaOH. Distill at least 100 cc into a measured volume of standard acid, in excess of the volatile bases, with delivery tube below the surface of the solution. Rinse the condenser and delivery tube with a little water. Titrate the excess of acid with standard alkali, using methyl red indicator. Each cc of 0.1 n acid consumed is equivalent to 0.0236 gramme of procaine or 0.0272 gramme of procaine hydrochloride. Stovaine, beta-eucaine, or pethidine will interfere, without yielding quantitative results; however, these do not seem to be normal heroin adulterants. Pethidine in small quantity can be quantitatively determined by distilling from an alkaline sodium-chloride solution. It is not decomposed by this procedure.


A number of identifying reactions have already been mentioned, particularly for diacetylmorphine itself. Some others that may be necessary or useful are given here. They require separation of the substance from most of the others present.

  1. Monoacetylmorphine. The alcoholic hydroxyl is the one acetylated; the phenolic hydroxyl is free and consequently the colour tests are practically the same as with morphine. The best microcrystal tests, which distinguish this compound easily from diacetylmorphine and morphine, are as follows:

  2. a. Reagent applied directly to a little of the dry substance:

    1. 1:60 HAuC1 4 in (1 + 1) H 2SO 4-rosettes of feathered crystals.

    2. 1:60 HAuCl 4in concentrated HCl-small yellow plates, frequently triangular.

    3. Iodine reagent C-3-large red-black needles or slender rods.

    Reagent applied to the aqueous solution:

    1. 1:20 HAuCl 4 in concentrated HCl-curving branched threads in rosettes, also feathered pale yellow plates.

    2. Bouchardat No. 1 (1 gm of iodine to 1 gm Kl in 100 cc solution)-branching threads in double sheaves with a bar between; dark brown and bluish-black colours.

  1. Codeineis easily distinguished by its strong reactions with Marquis' and Mecke's reagents, along with its negative or quite feeble reaction with Frohde's reagent. The best microcrystal test is probably the following:

    1. Iodine reagent C-3, applied to a little of the dry substance, soon yields brownish-yellow plates of characteristic appearance in shape and colour-shading, and rather dichroic with polarized light.27

  2. Papaverine

    1. Fuming Frohde's reagent gives the best colour test and one that is extremely sensitive. The colour produced is intense violet. It is convenient to keep Buckingham's reagent-10 gm ammonium molybdate in 100 cc concentrated H 2SO 4-and for "Fuming Frohde's", dilute a portion with about 10 times as much fuming H 2SO 4(15 to 25% free SO 3).

    2. Microcrystals: Papaverine gives a number of sensitive tests in which the crystals are small square plates, quite transparent. When they are somewhat distorted they form as 4-pointed stars. The best reagents are probably ZnCl 2 & HCl (reagent containing 40% concentrated HCl by volume), HgCl 2 & HCl (15% of concentrated HCl), and K 2ZnI 4.

  3. Procaine.The colour test with furfural and acetic acid and the following microcrystal tests are good.29The reagents are added to a drop of the aqueous solution.

    1. H 2PtBr 6-with a full drop of reagent, small dark rosettes of needles form soon. If a very tiny drop of reagent is added, there is immediate crystallization in light-coloured feathered crystals.

    2. HAuCl 4 and HCl (25% by volume of concentrated HCl)-large irregular coarsely feathered yellow plates. Very characteristic but less sensitive than the other tests given here.

    3. HAuBr 4 and HCl-small orange plates, scattered and in rosettes.

  4. Caffeine.

    1. Murexide colour test. In an evaporating dish mix a couple of milligrams of the powdered substance, a minute crystal of KClO 3, and 2 drops diluted HCl (1: 1). Evaporate to dryness on the steam bath and continue heating several minutes longer. (If the residue remains completely colourless heat cautiously with a small flame until it becomes pinkish or brownish.) Cool and treat with 2 drops dilute NH 4OH solution. A strong purple-red colour is produced.28

    2. Microcrystals: 1: 20 HAuCl 4in concentrated HCl applied directly to a little of the dry substance-needles in sheaves and rosettes. On standing, if the concentration of caffeine is fairly high, yellow cubical grains or square-cut plates form. This test distinguishes caffeine well from theobromine and theophylline.

  5. Amphetamine

    1. Microcrystals: 1: 20 HAuCl 4in (2 + 1) H 3PO 4, applied directly to a very little of the dry substance (use only about 0.1 - 0.05 mg of the pure substance). Small characteristic yellow plates or flakes soon form.


The heroin traffic is an international one, but too little is known in any country about the characteristics of illicit heroin elsewhere. Knowledge of the adulterants found in illicit heroin in the various countries may be of value to enforcement officers everywhere. Accounts of other methods found of value for identification, determination, or particularly for detection of certain adulterants by direct tests on the heroin sample will be welcomed.


Marquis, Pharmazeutische Zentralhalle für Deutschland 1896, 814. (The Merck Index, 5th edition, Merck & Co., Rahway, New Jersey, 1940.) ("Chemical, Clinico-Chemical Reactions, Tests and Reagents by the Authors' Names.")

Frohde, A., "Zum Nachweis des Morphiums," Zeitschrift für Chemie 1866, 378-379 (from Archiv der Pharmazie (2) 126,54).

Lafon, Ph. "Action des seleniates et des selenites sur les alcaloides, Nouvelle reaction de la codéine." Comptes rendus hebdomaires des séances de l'Académie des Sciences 100, 1543 (1885).

Mecke, Zeitschrift für Öffentliche Chemie 1899, 351. (The Merck Index, 5th ed., 1940)

Zernik, Berichte der deutschen pharmaceutischen Gesellschaft 1903, 67. (The Merck Index, 5th ed., 1940)

Deniges, Georges.

(a) "Nouvelle reaction de la morphine."

Comptes rendu hebdomaires des Séances del l'Académie des Sciences 151, 1062-63 (1910)

(b) "Nouvelle reaction de la morphine."

(Bulletin des travaux de la Société de pharmacie de Bordeaux, Nov. 1910) Répertoire de pharmacie (3) 23, 10-11 (1911) (Extrait)

Oliver, T. H., "A New Test for Morphine and its Salts" The Medical Chronicle, Manchester, 27, 221-222 (1914). Journal of the American Medical Association 63, 513-514 (1914) (Abstract).

Fulton, Charles C., "A Test for Phenols and for Copper" American Journal of Pharmacy, January 1933, 25-29.

Putt, Earl B., "Micro-Chemical Tests for the Identification of Some of the Alkaloids" Journal of Industrial and Engineering Chemistry, 4, 508-512 (1912).

Williams, G. D., and C. C. Fulton, "The Microscopic Identification of Heroine," American Journal of Pharmacy, September 1933, 435-439.

Fulton, Charles C., and John B. Dalton, "Micro-Crystal Identification Tests for Morphine, Heroin, Dilaudid, and Cocaine," American Journal of Police Science; incorporated in The Journal of Criminal Law and Criminology, 32, 358-365 (1941).

Hartshorne, N. H., and A. Stuart, Crystals and the Polarizing Microscope; a handbook for chemists and others. London, E. Arnold & Co., 1934.

Winchell, Alexander N., The Optical Properties of Organic Compounds. Madison, The University of Wisconsin Press, 1943.

Walton, C. F., and C. A. Fort, "Mannite and Dextran in the Jellying of Molasses from Juice of Frozen and Deteriorated Cane," Industrial and Engineering Chemistry 23, 1295 (1931).

Georgia, F. R., and Morales, Rita, "Detection of methanol in alcoholic beverages," Industrial and Engineering Chemistry 18, 304-306 (1926).

Herapath, Journal für praktische Chemie 61, 87 (1854) (The Merck Index, 5th ed., 1940).

Fulton, Charles C., "Iodosulfate Microchemical Identification Tests for Cinchona Alkaloids," Industrial and Engineering Chemistry, Analytical Edition, 13 , 848-850 (1941).

Sanchez, J. A., "Reactions for procaine" (translated title) Revista farmacéutica (Buenos Aires) 1917, 699 (Reference)

Sanchez, J. A., "Reaction for primary cyclic amines" (translated title) Anales de la asociación química, Argentina 23,431 (1925) ( Chemical Abstracts 20, 2300).

Fulton, Charles C., "The Identification of Methadon by Microcrystals," In United Nations document E/CN.7/117,14 April 1948, part II,B, pp. 17-31.

Lefort, M. J., "Etudes chimiques et toxicologiques sur la morphine suivies d'observations sur son passage dans l'économie animale," Journal de pharmacie et de chimie (3) 40, 97-113 (1861-2 e partie)("Morphine et acide iodique," pp. 103-107).

Fulton, Charles C., "Some New and Improved Tests for Morphine and Related Alkaloids," Journal of Laboratory and Clinical Medicine, 13, 750-765 (1928).

Chaurot, Emile Monnin, and Clyde Walter Mason, Handbook of Chemical Microscopy, Volume II. Chemical Methods and Inorganic Qualitative Analysis. N. Y., John Wiley & Sons (London, Chapman & Hall), 1931. "Detection of Calcium, Strontium, Barium. A-By Means of Dilute Sulfuric Acid," pp. 96-100, with Figs. 56 & 57.

Reimers, F., "Titration of alkaloidal salts" (translated title) Dansk Tids. Farm. 5, 42 (1931) ( Chemical Abstracts 26, 257).

Doran, James M., "The rapid separation of heroin and morphine," Journal of the American Pharmaceutical Association, 5, 163-165 (1916).

Matchett, John R., and Joseph Levine, "A Method for the Determination of Procaine." Journal of the Association of Official Agricultural Chemists, 23 , 776-777 (1940).

Fulton, Charles C., "Opium alkaloids-Codeine," In United Nations document E/CN.7/117/Add.1, 22 September 1948.

Mulliken, Samuel Parsons, A Method for the Identification of Pure Organic Compounds. Vol. II, New York, John Wiley and Sons (London: Chapman and Hall) 1916. (Murexide reaction, p. 31.)

Fulton, Charles C., "The Identification of Cocaine and Novocaine," The American Journal of Pharmacy, July and August 1933.


Full size image: 139 kB