Procedures for assured identification of morphine, dihydromorphinone, codeine, norcodeine, methadone, quinine, methamphetamine, etc., in human urine




Author: Masako ONO , Beatriz Ferreira ENGELKE , Charles FULTON
Pages: 31 to 40
Creation Date: 1969/01/01

Procedures for assured identification of morphine, dihydromorphinone, codeine, norcodeine, methadone, quinine, methamphetamine, etc., in human urine

Masako ONO Consultant, Addiction Services Agency; on loan from the Narcotics Laboratory, National Institute of Hygienic Sciences, Tokyo, Japan
Beatriz Ferreira ENGELKE
Consultant, Addiction Services Agency Charles FULTON
Chemist, Office of the Chief Medical Examiner, later Senior Chemist, Addiction Services Agency

These methods were introduced by the authors for the Laboratory of the Addiction Services Agency of New York City, at the Office of the Chief Medical Examiner of New York City. Titles stated are those that were held during the developmental period. Since then, the three authors have all left the Addiction Services Agency; others are now responsible for the A.S.A. laboratory work, which continues on a large scale.


This work was done to establish some routine methods for assured identification of basic addictive drugs in human urine. Basic substances are extracted in two portions, by chloroform from strongly alkaline solution (pH 13) and by chloroform-isopropanol from pH about 8.5. The extracts are then tested and the substances separated by thin-layer chromatography. After spraying with iodoplatinate solution the presence of blue or violet spots at certain Rf positions provides strong indications, but not conclusive identifications, of the drugs looked for. The new feature is that microscopic chemical tests (microcrystal tests) are combined with the thin-layer chromatography. Made on simple methanol eluates of the sprayed spots they provide definite tests for confirmation and certain identification. Some of the applicable crystal tests are also individually new.


In the past few years great interest has developed in the examination of human urine for drugs - especially for drugs of addiction in the urine of suspected addicts and supervised ex-addicts. This has grown to a major enterprise in several places, involving hundreds of urine samples each day. So great and sudden has been the demand for such analyses that the chief interest recently has been in developing rapid routine methods for indicatory results, which usually cannot be considered to provide conclusive identifications. Considering that a "positive" result may be introduced in a court case, or may affect the future of an ex-addict under supervision, it would seem that over-riding importance should be given to certainty of identification.

The Laboratory for the Addiction Services Agency was established under the aegis and direction of Dr. Milton Helpern, Chief Medical Examiner of New York City. He and Dr. Michael Baden, who was directly in charge of establishing this laboratory, decided at the outset that the urine analyses for drugs at the Office of the Chief Medical Examiner would be based on results as certain as routine results can be made.

The chief problem of addiction is the use of heroin, and heroin (diacetyl-morphine) rapidly hydrolyzes in the body to morphine. Therefore, procedures, alike of ourselves and others, are chiefly directed to the separation and recognition - or proof - of morphine.

However, other addictive drugs should be within the purview of a method. Quinine is not an addictive nor prohibited drug and its presence may be quite innocent, but it is a common adulterant in illicit heroin, and its presence along with morphine has some evidential value.

The procedures of others for drugs in human urine are diverse, but references to five published articles are given [ 1] , [ 2] [ 3] , [ 4] , [ 5] , and these can be read for comparison of methods. The first dates back 14 years; the others pertain to the recent surge of interest. The methods, experience, and standards of the Racing Chemists are also worth attention and study [ 6] . For over 30 years, in this country, they have been examining horse urine for drugs, including morphine - coming from heroin, which is a stimulant for horses.

For rapid routine work certain shortcuts have to be taken in processing, but assured final identifications should be the heart and backbone of any method. Too often there seems to be an assumption that when the analyst is looking for a particular drug, any "positive" test for it is sufficient, or at least sufficiently acceptable, without adequate proof of identity. Such testing, to be sure, screens out "negatives" for that drug, but for a "positive" it simply assumes what the analyst was asked to prove, the actual presence of a stated drug. An analyst never learns where and how a method may be in error, if he does not consistently try to confirm or verify all of his results that might otherwise carry some element of doubt.

Moreover, most of the urines tested come from suspected addicts and ex-addicts. In this restricted group any indicatory test for morphine usually is due to morphine, but the analyst should be more than ever on guard against the assumption that this is necessarily the case. "Positives" from such a group are too easy to believe even when proof is lacking. Certainly it is all right to assume morphine tentatively from any indication, as a guide to conclusive tests. But an assumption should not be the basis of a report asserting the presence of morphine, however valid it may be in most cases for the sort of people being tested - because when the assumption is wrong it is probably wrong for precisely those for whom it was most unjustified, and who will be hurt most by incorrect implication, or reports.


  1. Use 50 ml urine, if that much is available.

  2. If acidic drugs are in question, slightly acidify the urine with mineral acid and extract with 80 ml ether. Wash the ether with a little dilute NaHCO 3 solution to separate true acids from less acid substances. Evaporate the ether for extract to be tested for barbiturates and glutethimide.

  3. Strongly acidify the urine, to about 1 normal in H 2SO 4. Hydrolyze (autoclave) to free morphine, codeine, etc., from their glucuronides. Cool, refrigerate for an hour, filter. This removes interfering natural substances to a great degree.

  1. Make the urine strongly alkaline to about pH 13, also adding a little sulfite to prevent oxidation, and extract (just once, for routine), using 80 ml CHCl 3. Evaporate for extract B-1. Add a drop or two of tenthnormal HC1 during evaporation to retain methamphetamine or other readily volatile base, if any such is to be included in the scope of the examination.

  2. Make the urine definitely acid with (1 + 1) H 2SO 4, then slightly ammoniacal (adding also a little more sulfite). Add 2 g NaHCO 3 powder. The pH is about 8.5. Extract with 80 ml chloroform-isopropanol, 3:1. Draw off the organic solvent layer, dry it with anhydrous Na2SO 4 (let stand 3 hours); filter. Evaporate for extract B-2.

  3. Dissolve extracts B-1 and B-2 each in 0.25 ml methanol and spot them on thin-layer plates, silica gel G, 20 x 20 cm, thickness 25 microns, using (usually) 20 to 40% of each extract (the rest is reserved for possible further use). Also spot known morphine on every B-2 plate, codeine and quinine on B-1 plates (and/or other drugs if suspected, such as methadone, methamphetamine, etc.)

  4. Develop with solvent of chloroform-methanol, 160:40.

  5. Let solvent evaporate from the plates and examine them under a UV lamp. Then spray them with iodoplatinate solution. Let B-2 plate stand for 1 hour before final reading, in view of possible slow appearance of a feeble morphine spot.

  6. Scrape any spot from the dried plate that may represent a drug sought, and elute with methanol, evaporating in a 5 ml beaker.

  7. Treat the eluted residue with a small drop of 2% acetic acid, about 0.01 ml. Transfer a droplet to a slide, or distribute 2, 3, or more droplets of about a microliter each, or less, in cavities or on a plain slide, and also make dilutions, as appropriate. Evaporate droplets or cover cavities with plain slides as appropriate to the tests, and apply reagents suited to the substance indicated or sought.

  8. Examine the results under a polarizing microscope, observing test-forms (precrystalline or crystals) for form, aggregation, size, and birefringence, and, when appropriate, for color, dichroism, and sign of elongation, as hereafter described for the drugs of the title. (See Microcrystal Tests.)

  9. Partial deviations in the usual procedure are made for very weak spots that may be morphine, and for dihydromorphinone, quinine, and methamphetamine (see Microcrystal Tests for these substances).

Additional details and comments are given in the following discussion. Numbers in parentheses beginning paragraphs refer to the part of the outline discussed or amplified.


  1. As little as 10 ml urine may be used quite well, all the extract from a sample then being chromatographed, so far as it can be readily transferred to a plate. A relatively large sample gives far more opportunity for finding and proving a small amount of an expected drug, for repeating and confirming results when necessary, or for identifying an unexpected drug.

  2. True acids most likely to be found in the ether extract (and transferred to bicarbonate solution) are hippuric, a natural constituent, and salicylic, from aspirin. These, however, are not significant with regard to addiction. Barbiturates, although definitely acidic, are extractable by ether from bicarbonate solution. Glutethimide is slightly acidic, but nearly neutral.

The individual identification of acidic and neutral drugs is today rapidly becoming more and more important with the multiplication of such drugs, but they will not be further discussed here.

  1. Morphine, codeine, and numerous other drugs are excreted in large part combined with glucuronic acid, and not extractable in the usual solvent shakeout. Hydrolysis frees such "bound" substances; we have carried this out by autoclaving under a pressure of 18 pounds per square inch at 260°F, for half an hour. This heating in strongly acid solution is objectionable for the recovery of some substances, producing more or less hydrolysis when not wanted, but as long as morphine is the drug principally looked for, it is desirable for routine.

[ 4] , [ 5] A single extraction for all the solvent-extractable bases at once would have an obvious advantage for routine. But, in our opinion, there are then too many confusing spots for easily reading the subsequent chromatogram for the most wanted base (morphine). As long as morphine is the principal drug sought, and the addictive one most often found, an extraction should be made particularly for it. Because of its phenolic solubility in alkali solution it can be readily separated from most other bases. There are, to be sure, other phenolic bases among drugs and natural substances, and also amphoteric substances that are not phenols, but most basic drugs can be looked for in a separate extract, made first, from strong alkali.

[ 3] [ 4] [ 5] [ 6] The preparation of the solution and the basic extractions, while in the main the same as already done at the Chief Medical Examiner's and by Engelke of the A.S.A. Laboratory in beginning the urine analyses, have some features taken from work at the Japanese Narcotics Laboratory. The developing solvent finally used for the chromatography was also worked out there [ 7] .


  1. Before spotting a plate, it is convenient to make a preliminary test, particularly for the B-2 extract. Use a filter paper which has been dipped in the iodoplatinate solution and then dried. Try a tiny spot of the extract solution on this paper. The intensity of color of the spot, if one is formed, gives an idea of how much extract should be put on the plate. When no blue spot is evident in this preliminary test with B-2, put on as much of the extract as can conveniently be done in one spot. This will probably still be less than half the amount from 50 ml urine. The chromatography can therefore be repeated with the same or a different solvent system and the same or a different spray, without making a new extraction. The reserve extract is generally discarded as soon as the case has been reported.

  2. Chromatographic Rfs from rather crude extracts are somewhat variable, far more so than is generally admitted. The direct comparison of spot levels of samples with those of known morphine, codeine, etc., certainly cuts down discrepancies greatly, but by no means eliminates them.

  3. With the solvent given, the chromatographic separations, besides the usual ones, include codeine from its metabolite norcodeine, morphine similarly from normorphine, meperidine from normeperidine, methadone from its metabolite of unknown constitution, imipramine from desipramine, and quinine from fluorescent metabolites. Thus the original, still unchanged drugs are recovered in much greater purity than by methods which may make other separations but not these in particular. Also norcodeine, normeperidine, and occasionally normorphine can be eluted from their own spots and independently identified.

  4. The B-1 plate is examined under UV particularly for quinine fluorescence. There are often other spots that fluoresce more or less well, probably usually due to natural components of urine. Along with a good quinine spot there are also fluorescent spots which appear to be due to quinine metabolites. One such is on the B-2 plate. Some substances on the other hand tend to quench the violet light given off by a UV lamp. After a little experience the general appearance under UV shows whether the preparation of solution and the extractive separations were done carefully and well.

  5. The value of the extractive separation of morphine from most other bases is seen at once in the B-2 chromatograms. Samples on the B-2 plates often show no iodoplatinate spot at all, and a morphine spot, when it occurs, is usually very plain and a good dark blue, even when quite, small. Still, confirmation of the morphine is needed. And on the other hand, a spot similarly placed but not a good blue may contain enough morphine for verification, but have the spot-color affected by some other substance, also present.

Also the blankness of the B-2 chromatogram with most normal urines makes it possible to see very faint spots, the color of which is hard to judge, but which may be due to morphine. For such a spot, it is usual with us to repeat the chromatography, and if the spot appears again, the two spots are combined for elution. Many such spots have been confirmed as morphine, but most (though not all) of spots at the right Rf that fail of morphine confirmation, are weak.

There are at least three possible sources of non-morphine spots at the morphine Rf on the B-2 plates (most often weak ones in the first two cases):

  1. An "overflow" of a B-1 substance. Since only one B-1 shake-out is made, this is certainly possible if there is a large spot in the corresponding region of the B-1 plate.

  2. A substance (whether a drug or, at times, a natural urine component) may not be extractable by chloroform alone - perhaps because of too high watersolubility - but may be sufficiently extracted by chloroform-isopropanol to show a spot.

  3. Some substances, amphoteric like morphine, may also give spots at or about the same Rf. So far, dihydromorphinone - which on spraying with iodoplatinate gives a violet spot instead of the dark blue of morphine - is the only current drug now known to us to do this. But so far, by no means all of the possibly extractable phenolic or amphoteric drugs and other bases have been tried by us (nor covered by others, as far as we have found in published articles).

In the crystal tests it is frequently found that other substances, or "impurities ", are present along with morphine, in the same spot. In most cases they give amorphous precipitation but not crystals with the reagents used for morphine, and simply make the morphine identification more difficult. Some do give crystals with the M-2 reagent. Some of these may not be even partially responsible for a colored spot - not showing up at all with the iodoplatinate spray, if alone and pure. However, such a substance, moving with the same Rf as morphine, might attract reactive impurities and so be the cause of a faint spot in the morphine region, in the absence of morphine.

All in all, while a spot in the right place, especially a strong dark blue one, may be provisionally assumed to be morphine, pending further test, it still needs good confirmation. True, even some weak spots that cannot be confirmed may actually be due to morphine, but either below the limit of confirmation (in the amount recoverable from the sprayed spot) - no result in the confirmatory test - or too impure for the method described below (amorphous precipitation only). But cases of the latter kind in themselves mean that something else is present at the same Rf, and may sometimes be alone there; and other cases remain in which even a fairly strong spot in the right place is definitely neither morphine nor dihydromorphinone. One substance, not yet identified, is occasionally present, which gives a strong purple spot in the morphine region. When the developing solvent is changed to chloroform-methanolammonia (28%), 160:35:5, its Rf becomes higher and quite different from morphine or dihydromorphinone and the latter two are also separated from each other.

The situation as regards most B-1 spots is of course even more difficult. A spot may be in the right place (and have the right color) for a particular substance, but this is only indicative, with much less force than in the case of morphine, inasmuch as any of a multitude of drugs might show up on the B-1 plate. Individual and decisive confirmation is therefore required for spots of supposed codeine, methadone, methamphetamine, etc.

Different sprays are often used successively on the same plate, either to reveal other substances, or else to bolster up chromatographic conclusions about spots already shown. But, in our opinion, in the latter case this procedure is not sufficiently decisive even at best, and at worst may lead to erroneous conclusions, because two different substances may be (and often are) present in the same spot.


  1. The Rf values are so variable with factors such as temperature, quantity of drug, etc., and whether obtained with the pure drug, an extract of the pure drug from urine solution, or from samples (the drug having undergone more or less metabolism, and being always somewhat impure) - and so variable also in the hands of different analysts - that the attempt to specify the Rfs exactly, to two significant figures, can have little value, except perhaps as averages after many trials, or under rigorously prescribed conditions. The order, however, with the given solvent, on the standard silica gel plates, is unlikely to change. The trial of a very few knowns will then show on approximately what level to look for other drugs that have been tested. Even with known pure solutions only, to establish the order with certainty numerous trials are necessary. until drugs of similar Rf have had direct comparison on the same plate, in all cases. Subject to possible correction in a few cases, it is believed that the following lists of drugs in order of their Rfs from low to high, will be helpful. Lines through the two columns indicate Rfs of just under 0.2, 0.4 and 0.5. The higher Rfs seem to be more subject to discrepant values.

Solvent, chloroform-methanol, 160:40

B-1 drugs

B-2 drugs

Rf about .18
Rf about .38
Rf about .48
1,7-Dimethylxanthine *
Theobromine *
Theophylline *
Caffeine *

* These are best shown with strongly acid H 3BiI 6 solution as the spray.


  1. he purpose of the elution is of course tied to the possibility of microcrystal tests. Elutions from paper chromatography, but on a larger scale and with the base from a band recovered "free" in a solvent, had been previously used in toxicological work for the Chief Medical Examiner. However, it was not at all obvious that a simpler elution from the small TLC spots could be satisfactory. The combination of the thin-layer chromatography and the microcrystal tests was initiated by Engelke and proved successful from the start.

  1. With a spatula, scrape the spot area completely from the sprayed dried plate. (In some cases, a spot can, with advantage, be scraped from the unsprayed plate, which can afterward be sprayed for other substances.) Transfer the powder into a 5 ml beaker and add 1 ml methanol. Warm on the steam bath while moving gently. Filter through a small folded filter paper of 3 cm diameter, just previously moistened with methanol, into another 5 ml beaker. Again treat the residue in the first beaker with 1 ml of methanol, shaking gently. Filter into the beaker containing the first filtrate. Repeat once more, then wash down the filter paper with 1 ml methanol. Evaporate the combined methanol filtrates.


(10-12) The most definite and certain identifications that can be readily made in cases of this kind are with microcrystal tests [ 8] . Sometimes several different tests can be run on the eluate from one spot, and often dilutions of the eluate solution can be used to advantage. On the other hand it is occasionally necessary to concentrate the eluate solution on the test slide, as well as possible, by repeated transfer of droplets, evaporated in the same place for a final single test.

In the toxicologic work, a chromatographic band on paper is eluted with chloroform in the presence of a small volume of solution of borax and sodium sulfite. UV spectrophotometry and both color and microcrystal tests can then be used on the eluted residue. There is a wealth of crystal tests and a particularly good one can generally be found for nearly any drug, particularly a basic one. Unfortunately the simple methanol eluate of a sprayed TLC spot contains iodide from the spray, which interferes with most of the good color tests (the best are with concd. H 2SO 4 reagents), and with many of the crystal tests. Potassium ion from the spray and sodium ion from the plate itself may also interfere, but the iodide interference is the most serious. However, a considerable number of the best crystal-producing reagents already contain iodide, and these can be used with the simple methanol eluate. The best tests for morphine, codeine, methadone, and many other drugs are thus available for use with TLC. Some other crystalproducing reagents, not reacting with the iodide, are also available. Quinine spots can be noted by fluorescence for elution without spraying. Volatility tests can be used for methamphetamine and numerous other volatile drugs. Tests are described here for the drugs mentioned in the title: these have been used successfully on eluted residues not merely with spots of "knowns" but also with spots from urine samples.


  1. The best test for morphine is an old one, the simple aqueous test with K 2HgI 4. It is best obtained on a neutral solution of morphine salt or a solution only slightly acid with dilute acetic acid, and with a reagent saturated with HgI 2 (not properly Mayer's reagent, which has excess KI). It is sensitive to less than 0.05 microgram in a microliter drop tested (with no more than an equal volume of reagent solution added), and is best with a very dilute solution just strong enough to give slight but immediate amorphous precipitation. Strangely, considering that the test goes back at least to 1885 [ 9] , a number of chemists trying out crystal tests have reported the result simply as amorphous. They no doubt failed to observe the precrystalline forms properly, or even see them (too much light), and then let the drops dry up before the definite crystallization occurred; and they may have tried only too-high concentrations. The test has been much used by the Racing Chemists, however [ 6] .

In testing a sample eluate, place a droplet in a cavity slide, and partly transfer, diluting with a tiny droplet of 2 % acetic acid in another cavity; or in two or three others successively, if the spot was more than minimal. Then test the greatest dilution by mixing in a tiny droplet of reagent solution; and proceed to test the more concentrated droplets until a concentration is reached which gives a distinct immediate precipitate (or until the whole series of droplets has been tested). To prevent any rapid evaporation, cover each test with a plain slide (which is generally removed for a few moments while looking microscopically, since vapor from the solution fogs the slide above). Examine at about 100 x.

In a short time a very dilute solution - even well beyond the limit of immediate precipitation - shows at the edge of the droplet characteristic, very transparent, rather vague dendrites. Look for these with minimum light; with too strong light they cannot be seen. With a little greater concentration the slight amorphous precipitate quickly changes to characteristic smudge rosette forms out in the solution, also best observed with minimum light. A little heavier precipitate gives these rosettes small and in great number, producing "speckling ". These forms are dark with slightly lowered focus, bright or "silvery" with slightly raised focus. In a concentrated solution the precipitate gelatinizes.

There seems to be a doubt whether these dendrites and rosettes are truly crystals; they may be considered precrystalline forms. Gradually, with standing (perhaps in 10 or 15 minutes with fairly pure morphine, but allow at least two hours for samples) unmistakable crystals form as brushes, fans, rosettes, etc., of fine needles, appearing somewhat brownish to brown by transmitted light. Sometimes they occur as loose cluster-dendrites of very long fine needle-hairs.

The precrystalline forms usually show no birefringence and the needles little or almost none with fairly pure morphine. With impurities - always present to some extent with samples - the precrystalline forms occur with little difficulty but the change to needle crystals is delayed, incomplete, and with minute amounts or high impurity will not occur at all. Also in such cases the precrystalline rosettes may show weak birefringence and any needle crystals that slowly form may be fairly bright with the polarizers crossed. When the sign of elongation can be distinguished with the red plate it is positive for the rosette forms, negative for the bright needles.

When a complete test-result is obtained - not only the characteristic vague dendrites and rosettes but also brushes or other characteristic aggregates of the fine needles - this one test alone definitely identifies morphine. When the amount of morphine is so small or the proportion of impurities so great that the formation of definite needles does not occur, the precrystalline forms still constitute an excellent test which is considered to confirm morphine, taking into consideration the method of extraction and the chromatographic spot at the right Rf.

When confirmation of morphine is sought from a small, vague, and doubtful spot, and no reaction is obtained with the first droplet of eluate solution transferred to a cavity, transfer all the test of the residue, so far as possible, to a cavity, washing down the beaker with another little drop of dilute acetic acid, and evaporating a droplet at a time in the cavity until the last, then the test is made. If even the characteristic dendrites at the edge are not then obtained, there is no good evidence of morphine regardless of the spot Rf.

  1. K 2CdI 4, aqueous test. This is very similar to the K 2HgI 4 test but requires a more concentrated solution (usually the next more concentrated than gives a good K 2HgI 4 test, if there were several successive dilutions). The vague dendrites at the edge are larger but with fairly pure morphine even more transparent and more difficult to see. However, with rather impure morphine they are easily seen and the smudge rosettes are more likely to show weak birefringence and distinguishable + elongation than those obtained with K 2HgI 4. The great advantage of this test, sometimes outweighing its lesser sensitivity, is that definite crystallization in brushes, fans, sheaves and rosettes of fine needles occurs comparatively readily. Also, these needles always exhibit enough birefringence with strong light to show negative elongation with the red plate.

When a distinct but small spot is tested and the precrystalline forms are just obtained with K 2HgI 4 in a test on the original eluate solution, it may be advisable to concentrate all the rest of the eluate in one cavity for the K 2CdI 4 test, as the latter will offer a better chance of going to needle crystallization. For eluates of larger spots, K 2CdI 4 readily provides an excellent second test. It is a distinct but not a completely independent test, due to the relation between Cd and Hg in the periodic table, and the very similar forms obtained in the case of morphine.

  1. I-KI reagent M-2, added with cover-glass to a dry deposit [ 10] . Place a droplet of the solution to be tested on a plain slide, and if it is at all concentrated make one or two dilutions on the slide; then evaporate the droplets and test the residues. This is a very sensitive test with pure morphine but interferences due to impurities or other substances present make the practical sensitivity with samples less than with K 2CdI 4. Crystals due to other substances than morphine are often obtained in part. Most common are little flakes generally ovoid, highly dichroic, colorless or light ochre to black, with crossed polarizers bright white or ochre. Sometimes there are good rectangular blades of the same dichroism; probably due to a different substance.

The morphine-iodine-iodide crystals are themselves of several kinds, varying with the concentration and with the amount of extra iodide in the eluate. The "most dilute" and often the only ones obtained are long fine black needles to brown hairs in clusters, having scarcely any birefringence. The "concentrated" crystals are red, red-brown, or red-black blades, plates, bars, often in clusters, birefringent. These two kinds are the normal forms of morphine with this reagent. With extra iodide in the eluate there are quite small rosettes of short dark brown needle-blades of low birefringence; these require just a little greater concentration than the long fine needles. They occur frequently because of iodide from the spray, and are often the only kind of crystals that form in the tests of many samples. With a concentrated sample residue there are sometimes dark rods, often branching or splitting at one or both ends, with brownish-orange birefringence. These are also caused in part by extra iodide contributed by the spray.

When both normal kinds of crystals are obtained the test is considered specific for morphine in itself. Any of the crystals with morphine are highly characteristic and add to the proof of identity. However, the small rosettes, the relatively short, dark brown needleblades, are closely related to the "dilute" kind of crystals obtained with nalorphine and the same reagent M-2.


  1. If a spot may be dihydromorphinone, rechromatograph another part of the extract, developing with chloroform-methanol-ammonia, 160:35:5. Include knowns of morphine and dihydromorphinone on the same plate. They are now separated; morphine Rf about .45, dihydromorphinone about .57. The notyet-identified substance previously mentioned as giving a purple spot in the same region now goes much higher.

If the second chromatogram also indicates dihydromorphinone, confirm it on the eluates of both spots separately, if they seem large enough, or on the eluate of the two spots together if they are small, by means of a test with solution of reinecke salt. Also elute the known dihydromorphinone for a comparative test.

Dihydromorphinone gives reineckate crystals readily, rosettes of pointed blades, white with polarizers crossed, + elongation. With TLC eluate a small amount may give at first rosettes that are rather vague (or composed of delicate fine needles) of negative elongation. In any event compare with the result on the known eluate under as nearly as possible the same conditions.

Reinecke salt gives crystals with a great many bases. These reineckates are not so distinctively different from each other as the crystals with some other reagents, but the test is sometimes the best one available, as seems to be the case with TLC eluate of dihydromorphinone. The two chromatographic results together with the method of extraction, and plus the crystal test, place the identity beyond a reasonable doubt. The proof is further strengthened by obtaining the same crystals from the spots separately, if possible.

Morphine, oxymorphone, dihydrocodeine, norcodeine, and normeperidine are among the many bases giving crystalline reineckates.


  1. (3)HgI 2 - (7)KI, aqueous test. Codeine does not readily crystallize simply with K 2HgI 4 (KI solution saturated with HgI 2), but does so when a large excess of iodide is present. The Mayer formula contains excess iodide but not enough for the best crystallization. A reagent previously used has had 3 g HgI 2 with 15 g KI (per 100 ml), for crystals with codeine and other alkaloids requiring excess KI for good results [ 11] .

However, the codeine crystallization is rapid enough, and the sensitivity a little better, with a little less than half this much KI - still a large excess, as only about 2 g KI is required to dissolve 3 g HgI 2. Make the test on a little droplet in a cavity and cover with a plain slide to prevent any rapid evaporation.

As with morphine there are precrystalline forms, which with codeine from the usual samples are spherulites, looking much like drops with transmitted light, but showing low bluish birefringence, radially positive. With eluate from small spots the change to truer crystallization does not occur. However, as the test drop dries down, and best seen shortly before complete dryness, the spherulites change to a brighter yellow birefringence, and are radially negative. Sometimes this change can be seen at the center of a spherulite, with its rim remaining bluish and positive. If the drop is allowed to dry up completely and then water or dilute acetic acid added, the dimmer bluish spherulites are restored. This test even without better crystallization at least supplies good confirmation of codeine in a spot at the right Rf.

With purer codeine, or generally with eluate from a large spot, smudge rosettes, radially positive, will form in part, or instead of the spherulites. These may crystallize in tiny grains. Also larger irregular crystals form, more or less in clusters, of bright birefringence, inclined extinction.

  1. HgI 2.KCN + (9)KI or HgI 2.NaCN + (8)NaI, added to the eluate solution in dilute acetic acid, gives similar results, generally with brighter birefringence of the spherulites first formed.

  2. K 2CdI 4 gives somewhat similar results, but the spherulites first formed have little or often no distinguishable birefringence. True crystals of bright birefringence, but of no very definite characteristic shapes, form somewhat more readily, perhaps, than in the first test above.

  3. I-KI reagent 0-2, added with cover-glass to a dry deposit. Clusters of somewhat chisel-like crystals, tending to be triangular, attached by their apices, form gradually. They are very brightly birefringent and quite dichroic, yellow to red, highly characteristic and unmistakable. The practical sensitivity is less than for the preceding tests but this one can very often be obtained on samples - generally, when the amount is large, as is usually the case when an addict takes codeine.


  1. (3)HgI 2 - (7)KI, added to the solution in dilute acetic acid. Gives crystals readily, brightly birefringent; with crossed polarizers they look mostly like rods, but by transmitted light are seen to be clusters of plates, and when well formed are seen to be triangles attached by their apices. This test alone will confirm norcodeine in a spot at the proper level.

  2. To a droplet of the eluate in dilute acetic acid, add a grain or two of NaC1. If the amount of norcodeine is not too small, the dissolving grain is quickly surrounded by needles to rods, bright white with polarizers crossed, negative elongation. While not as sensitive as the preceding test, the insolubility of the hydrochloride is a remarkable characteristic. The crystals can also be obtained by adding (1 + 11) HCl. From a very small yield of norcodeine they are sometimes only obtained when the drop dries. In fact, simply drying a droplet of the eluate solution will often yield some similar crystals, but doubtless of the hydriodide. A good test can be obtained by adding KI instead of NaCI, but the crystals with the latter are more unusual.

  3. The reinecke test gives small clusters of irregular blade crystals, often slightly curved, of rather low birefringence, positive elongation.

  4. I-KI reagent 0-2, added with cover-glass to a dry deposit. This reagent gives crystals with known norcodeine which are highly characteristic, but the test is not sufficiently sensitive for the small TLC spots so far obtained with samples. Brown blades, more or less rectangular, dichroism yellow to brown, or light brown to black-brown, birefringent, often in "paddle-wheel" aggregates.


  1. I-KI reagent Q-6 added with cover-glass to a dry deposit. The amorphous precipitate quickly crystallizes in colorless (white) small needles and irregular blades to feathers, often in clusters, with bright white birefringence, negative elongation and highly characteristic narrow Xs with arms of blades, colorless or light yellowish, with low bluish birefringence, ± elongation (perhaps more often +, but sometimes differing on parts of the same X). When an X grows large enough an interference figure can be obtained, either an optic axis or an acute bisectrix with small separation of the isogyres. The sign of the crystal is positive.

  2. I-KI reagent 0-2, used in the same way, gives at the more concentrated parts of the deposit the same kinds of crystals as Q-6; and in the more dilute parts narrow straight splinters and brown blades with + dichroism, black to light brown, with crossed polarizers brown-orange. This reagent is even more sensitive than Q-6, but if only the "more dilute" kind of crystals is obtained, it is not so highly characteristic. On the other hand, when all three types of crystals are obtained this test must be considered specific by itself.


  1. A quinine spot can be noted before spraying with iodoplatinate by fluorescence at the proper Rf. Use an eluate from such an unsprayed spot.

  1. H2PtBr6 in (4 + 1)HBr,(35), added with coverglass to a dry deposit. Orange plate crystals soon form, very brightly birefriengent, hexagonal, often clustered; or clusters of orange rods. The reagent reacts with iodide.

  2. I-KI reagent Q-6, added with cover-glass to a dry deposit. First forms characteristic brown smudge rosettes, birefringent, radially + with the red plate. On standing, if pure enough, deep red to black grains form, some showing deep red with crossed polarizers, but the more perfect ones are squares or cubes which are not birefringent in a flat position. The brown rosettes constitute an excellent test even if the change to redblack grains does not occur. This test can be used with eluate from a sprayed spot, but results are better from an unsprayed spot.

The crystal tests are sensitive, but not as sensitive as the fluorescence and the iodoplatinate spray. Consequently, not all quinine spots that may be noticed can be confirmed - even if they really are quinine. In narcotic cases, but only those in which a confirmed spot of morphine has been found, the quinine indication has been reported by us on the basis of fluorescence and sprayed Rf, without attempting further confirmation. Quinine is not a narcotic or prohibited drug anyway, but the spot has some significance as more definitely indicating heroin addiction when the morphine is certainly present.

However, if the presence of quinine is an essential point, it needs confirmation no less than other chromatographic spots, even though its fluorescence makes it considerably more probable from the Rf than other B-1 drugs. Urine contains natural fluorescent substances, and there are also other drugs and substances that fluoresce that may possibly be present. Thus there is some danger of "spot hunting ": i.e., looking for a quinine spot and calling any fluorescent spot in the neighborhood "quinine ". If there is no check such guesses tend more and more to be accepted as correct in all cases. Also, quinidine gives the same kind of fluorescence, metabolites, and Rf as quinine. Their important microcrystal tests are quite different and supply the chief practicable distinctions. Quinidine is used as a drug but has never been reported as an adulterant of heroin, to the writers' knowledge.


  1. This is said to be currently a much abused drug, colloquially known as "Speed ". The chromatographic spot can be readily identified by subjecting the eluate to volatility microcrystal tests. These can, indeed, be carried out on the scraped spot without making a methanol elution, but since not very much powder can be put in a cavity without touching the hanging drop on a plain slide above it, use of the eluate is generally better.

Put a droplet of the eluate in a cavity, make alkaline with a small drop of 10% NaOH, and cover with a plain slide bearing a small hanging drop of reagent or of (1 + 99) HCl. In the latter case, after the drop has been exposed to the vapours for up to about 45 minutes, or for a suitable time as shown by the development of precipitation in a parallel hanging drop of reagent, reinvert the plain slide and allow the drop of captured hydrochloride solution to dry up. Examine the residue microscopically and then test, usually with a reagent added to the dry deposit with a cover-glass. There are over a dozen readily volatile amphetaminetype drugs (plus nicotine and at least four common volatile bases of decomposition, including phenethylamine, one of the closest relatives of amphetamine); and microcrystal tests to distinguish them are absolutely essential [ 12] .

  1. HAuCl 4 in (l+2)H 3PO 4, as hanging drop reagent. An important test for most of the volatile sympathomimetics, including amphetamine. Methamphetamine readily gives crystals which are mainly needles to "toothpicks" and rods, sometimes dendritic, sometimes feathered, very bright birefringence (rising to colors), negative elongation. When sufficient precipitate has formed, whether in drops or crystals, reinvert the plain slide and let stand for observation. When the amount being tested is minute, expose to the vapor for 30-45 minutes, reinvert the reagent drop, and let it air-dry down to strong H 3PO 4. This gives a sensitive test. dl-Methamphetamine gives similar crystals to the d- with this reagent.

  2. H 3BiI 6 in (1+7)H 2SO 4, as hanging drop reagent. d-Methamphetamine gives long orange splinters and needles, and deep red angular grains, dl-Methamphetamine gives characteristic red rods with slanting ends.

  3. HAuBr 4 in HOAc-3(2+3)H 2SO 4,(40), added with cover-glass to the dry hydrochloride deposit captured by hanging drop of (1 + 99) HCl - irregular orange-brown blades to needles usually form first, no dichroism, bright with crossed polarizers, negative elongation, good distinction from mephentermine. Later there are crystals with strong dichroism of yellow to red, usually negative absorption but sometimes with sign reversed, in long strands and blade-rods, or plateblades which tend to be rectangular. Mephentermine also gives strongly dichroic crystals with this and the following reagent. dl-Methamphetamine does not readily give dichroic crystals (see next test).

  4. HAuBr 4 in 2H 3PO 4 - 1(2+3)H 2SO 4,(90), added with cover-glass to the dry hydrochloride deposit - very good for the dichroic crystals with the d-, colorless or yellowish to red, generally segmented and square-cut but quite elongate, dl-Methamphetamine gives mainly non-dichroic orange-brown rhomboidal crystals.


  1. Iodoplatinate spray: 5% platinum chloride solution (H 2PtCl 6.6H 2O crystals - 1 gram, in 20 ml H 2O) - 2 ml, KI solution (4 g/100 ml) - 25 ml, H 2O to make 50 ml.

The following microcrystal reagents can conveniently be kept in small amounts in 8 ml dropping bottles which have screw cap, pipet, and rubber bulb.

  1. K 2HgI 4. Dissolve 2 g KI in 100 ml water and saturate with HgI 2 (nearly 3 g HgI 2 required). Leave a little excess HgI 2 in contact with the solution.

  2. K 2CdI 4. Dissolve 5 g CdI 2 and 4.5 g KI in 100 ml water.

  3. Reinecke salt in glycerol-water (1+4) [ 12] . Dilute 1.0 ml glycerol-water (1 + 1) with 1.5 ml H 2O and add a pinch of reinecke salt. Shake and let stand a short time until approximately saturated. Use fresh (same day as made). Add to an aqueous droplet on a plain slide and let air-dry down to glycerol--the reagent will not crystallize out.

  4. (3)HgI 2- (7)KI. Dissolve 3 g HgI 2 and 7 g KI in 100 ml water.

  5. (a)HgI 2.KCN + (9)KI. Dissolve 0.7 g KCN in I00 ml water and saturate with HgI 2 (use 4.6 g). Filter, then add 9 g KI.

  1. HgI 2.NaCN + (8)NaI. Dissolve 0.5 g NaCN in 100 ml water and saturate with HgI 2 (use 4.6 g). Filter, then add 8 g NaI.

  1. I-KI reagent M-2. First make a solution of 5 g iodine and 30 g KI with water to make 100 ml. Mix 1.0 ml. of this with 1.5 ml concd. HCl and 1.5 ml syrupy H 3PO 4 (85 %). Best kept in the refrigerator.

  2. I-KI reagent O-2. Mix 0.4 ml I-KI solution (10:35 g/100 ml) with 2.0 ml glacial acetic acid, 1.6 ml water, and 2.0 ml syrupy H 3PO 4. Best kept in the refrigerator.

  3. I-KI reagent Q-6 [ 12] . Mix 0.4 ml I-KI solution (5:30 ml/100 ml H 2O) with 1.0 ml glacial acetic acid, 0.6 ml H 2O, and 1.0 ml magnesium acetate solution (40 g/100 ml H20). Best kept in the refrigerator.

  4. H 2PtBr 6 in (4+1)HBr,(35) [ 12] . 1 g H 2PtCl 6H 2O crystals dissolved in 28 ml HBr (40 %) plus 7 ml H 2O.

  5. HAuCI 4 in (l+2)H 3PO 4,(20). Dissolve 1 g HAuCI 4.3H 20 crystals in 20 ml of a mixture of 1 part syrupy H 3PO 4 with 2 parts water.

  6. H 3BiI 6 in (l+7)H 2SO 4. First prepare a concentrated Bi(NO 3) 3 solution by dissolving 50 g bismuth subnitrate in 70 ml (l+1)HNO 3, and diluting to 100 ml with water. (This is used for mixing various H 3BiI 6 reagents). Then mix 0.5 ml of this with 2.0 ml water, 1.25 g KI, 2.5 ml (l+3)H 2SO 4, and 0.05 g Na hypophosphite. Best kept in the refrigerator.

  7. HAuBr 4 in HOAc-3(2+3)H 2SO 4,(40) [ 12] . First make HAuBr 4 in (2+3)H 2SO 4,(30), with 1 g HAuC1 4.3H 2O crystals, 1.5 ml HBr (40%), and (2+3) H 2SO 4 to make 30 ml. (This is used as an excellent reagent in its own right, as well as for mixing others.) Mix 3 ml of this with I ml of glacial acetic acid.

  8. HAuBr 4 in 2H 3PO 4-1(2+3)H 2SO 4,(90). Mix 2 ml of the same HAuBr 4 in (2+3)H 2SO 4,(30), as in the preceding formula, with 4 ml syrupy H 3PO 4.



Gilbert J. Mannering, Arthur C. Dixon, Nicholas V. Carroll, and Ogle B. Cope, "Paper Chromatography Applied to the Detection of Opium Alkaloids in Urine and Tissues." Journal of Laboratory and Clinical Medicine, 44, 292-300 (1954).


J. Cochin and J. W. Daly. "Rapid Identification of Analgesic Drugs in Urine with Thin-Layer Chromatogr\aphy." Experientia 18, 294-5 (1962).


K. D. Parker and C. H. Hine. "Manual for the Determination of Narcotics and Dangerous Drugs in the Urine." Psychopharmacology Bulletin, July 1966; Bulletin on Narcotics (United Nations) 19, No. 2, 51-7 (1967).


Vincent P. Dole, Wan Kyun Kim, and Ilze Eglitis. "Detection of Narcotic Drugs, Tranquilizers, Amphetamines, and Barbiturates in Urine." Journal of the American Medical Association, 198, 349-52 (1966).


Bernard Davidow, Nicolo Li Petri, Babington Quame, Bernard Searle, Eugene Festlich, and Joseph Savitscky. "A ThinLayer Chromatographic Screening Test for the Detection of Users of Morphine or Heroin." The American Journal of Clinical Pathology, 46 (No. 1) 58-62 (1966).


Harry Peterson . Methods for the Collection and Analysis of Horse Saliva and Urine for the Detection of Drugs. 9th Revision, 1964. New York State Racing Commission Laboratory. Not a public document but the methods are not held secret from chemists outside the Association of Official Racing Chemists who have related official work.


M. Ono and H. Asahina. "Detection of Narcotic Drugs in Biological Fluid IV." (In Japanese ). Bulletin of the National Institute of Hygienic Sciences, 83, 16 (1965). Also contributions II, III and V by the same authors in 1963, 1964 and 1966; and I by H. Asahina and M. Ono in 1961.


Charles C. Fulton. 10 articles under "Chemical Microscopy" in The Encyclopedia of Microscopy, edited by George L. Clark, Reinhold Publishing Corporation, 1961.


Theodore G. Wormley. Microchemistry of Poisons, 2nd edition J.B. Lippincott Co., Philadelphia, 1885.


Charles C. Fulton. "Crystal Tests for Minute Amounts of Morphine ." Journal of Laboratory and Clinical Medicine, 23, 622-5 (1938).


Charles C. Fulton, "The Precipitating Agents for Alkaloids." American Journal of Pharmacy, 104, 244-71 (1932).


Charles C. Fulton . Modern Microcrystal Tests for Drugs, to be published by Interscience Publishers, New York.