Summary
Introduction
Methods - (a) Thin-layer chromatography
(b) Gas-liquid chromatography
(c) High pressure liquid chromatography
Results and discussion - The big lots seized in 1975
All samples seized during 1976
Conclusions
TABLE I
TABLE II Division into subgroups of 14 samples containing ethylcarbonate
Acknowledgements
Author: H. H. A. J. STEENSTRA. HUIZER LOGTENBERG
Pages: 65 to 74
Creation Date: 1977/01/01
All heroin samples seized in the Netherlands in 1975 and weighing more than 10 grams together with all heroin samples seized during September 1976 were analysed by thin-layer chromatography, gas-liquid chromatography and liquid-solid chromatography. Most samples consisted of the so-called Hong Kong No. 3 type. Also the samples found on addicts were as potent as those seized from dealers and traffickers. We attempted to classify the samples into groups or batches, but only in a few cases could we conclude that two or more samples originated from the same batch.
In the Netherlands all drugs confiscated by law enforcement (or judicial) authorities are analysed at the Forensic Science Laboratory of the Ministry of Justice or at the Laboratory of the Municipal Police of Amsterdam. In the former laboratory, 4,200 drug samples were analysed in 1975, a number which rose to about 5,000 samples in 1976. Of all samples analysed the proportion of those corresponding to heroin rose from 15 per cent in 1975 to 30 per cent in 1976.
Qualitative analysis of heroin is based on the morphological properties of the product, the purple reaction it yields with the Marquis reagent, and thin-layer chromatography. A quantitative analysis was performed only when it was estimated that important information could be obtained for the prosecution. However, all heroin samples seized in 1975 and weighing more than 10 grams together with all those seized in September 1976 underwent quantitative analysis with a view to obtaining an indication of the quality of the product. In addition, we were interested to see whether the smaller amounts found on addicts were less potent than the larger quantities found with dealers and traffickers, and whether relatively harmful substances had been added. Finally, we hoped to be able to classify the samples into groups, thereby identifying those which originated from the same source.
The samples were dissolved in alcohol (50 per cent) and spotted on pre-coated thin layer plates (silica gel 60 F 254 Merck).
The developing chamber was saturated with a chloroform-cyclohexane-diethylamine (8:10:3) mixture which was used as the mobile phase. Developing time was about 1½ hours, and afted evaporation of the mobile phase in a drying box at 105 °C, the spots were visualized under UV-light and their position on the plate determined. Figure la shows the chromatogram of an average heroin sample which was found to also contain some barbitone. Next the plates were sprayed with a neutral iodoplatinate solution (l g H 2 Pt Cl 6.6H 2O + 10 g KI in 500 ml of water) and as a result, the alkaloids became visible in the form of blue spots. Also O 6-monoacetylmorphine which was not separated from caffeine became visible as a separate spot, as caffeine does not react with the spray (figure 1b).
When we analysed the larger lots seized in 1975, we were interested in the contents of heroin and caffeine only. Therefore, we could make use of our standard chromatographic equipment which consists of a Perkin Elmer F 30 gas chromatograph fitted with glass columns containing 3 per cent OV-17 on Chromosorb WHP. Column temperature was 260 °C and detection was achieved with a flame ionization detector. About 100 mg of the homogenised heroin sample were dissolved in 10.0 ml of chloroform containing squalans (1 g per litre) as internal standard and 0.3 μl of this solution were injected into the gas chromatograph. The peak areas were measured by an electronic integrator and the system's response was found to be linear in the expected concentration range.
Later, when we became interested in the quantitative determination of the other alkaloids, it seemed more convenient to use high pressure liquid chromatography than to modify the gas chromatographic method.
The separation was carried out on a Siemens S100 liquid chromatograph equipped with a Zeiss spectrophotometer as detector.
|
acetylcodeine
|
|
heroin
|
|
strychnine
|
|
monoacetylmorphine
|
|
caffeine
|
|
barbitone
|
|
morphine
|
Thin-layer chromatogram of a heroin sample containing barbitone ( a) under U.V. light; ( b) after spraying with the iodoplatinale solution.
In order to detect all the expected components in one run, we chose a wavelength of 250 nn. Use was made of steel columns packed with Lichrosorb Li 60, 5?. The mobile phase consisted of a mixture of diethyl ether-isooctane-methanol-diethylamine (52.8:35:12:0.2, by volume) and the elution rate was 1.5 ml/min. at a pressure of 175 bar ( [ 1] ).
About 40 mg of the homogenised sample were dissolved in 25.0 ml of chloroform containing codeine (160 mg per litre) as internal standard and 10 μl of the resulting solution were injected (figure 2) into the chromatograph. Despite the fact that codeine could be expected to be present in some of the seized heroin samples ( [ 2] ),( [ 3] ) still it was chosen as the internal standard because no other compound with a suitable retention time could be found. As a consequence, all samples had to be screened for the presence of codeine, and whenever it was detected, tile composition of the sample (dissolved in chloroform) was also determined in the absence of the internal standard. The retention times and peak areas were measured with an electronic integrator (Perkin Elmer M2). In addition, all seized samples were analysed twice while standard solutions were analysed three times. The opium alkaloids were dissolved as hydrochloride salts except for O 6-monoacetylmorphine, which was dissolved as the free base because its hydrochloride salt is poorly soluble in chloroform.
It was remarkable that almost all seized samples, including those containing as much as 15 per cent O 6-monoacetylmorphine,were found to be very soluble in chloroform. All opium alkaloids contents were calculated as hydrochloride salts. Strychnine was dissolved as the nitrate salt and its content calculated on that basis.
The detector's response was checked for strychnine nitrate, caffeine and heroin hydrochloride, and was found to be linear for the expected concentrations. The standard deviation was measured by taking 7 samples out of a fully homogenised batch of heroin, and analysing them according to the procedure described earlier. The relative peak area of each component to that of codeine was determined twice (table I, a).
In order to get some indication about the inhomogeneityofgranular heroin, this procedure was carried out with 7 samples taken out of a batch of granular heroin, each sample consisting of about 2 granules (I, b).
During the course of this research, 10 heroin samples were found to have a characteristic composition such that it was very likely they belonged to the same batch. From tile values obtained for these samples also, a standard deviation was determined (I, c).
The high standard deviation shown in table I c for heroin and O 6-monoacetylmorphine contents can be explained in terms of heroin degradation under varying conditions; the changes of the standard deviations of the values obtained for strychnine and acetylcodeine are mainly caused by the fact that these samples had a low content of acetylcodeine and a high one for strychnine.
Thin-layer chromatography and gas chromatography were used in the analyses of the big lots seized in 1975. Of the 50 samples, only two consisted of the so-called No. 4 heroin, a white powder which contains mainly heroin hydrochloride and acctylcodcine hydrochloride. The heroin hydrochloride content of these two samples was 89 and 91 per cent (2,3). One sample had no heroin, but did contain O 6-monoacetylmorphine HCI, morphine HC1, caffeine and strychnine. The appearance of this sample was not different from that of the other granular samples.
The remaining 47 samples were of the so-called No. 3 type, a grey-brown granular product manufactured for smoking purposes (4). Number 3 heroin is chemically characterized by the presence of heroin hydrochloride, O 6-monoacetylmorphine hydrochloride, acetylcodeine hydrochloride and caffeine. Most samples also contained strychnine while barbitone was found in only 5 of them. Other substances identified in some of the samples were procaine, sugar and phenacetin (traces). The distribution of the heroin hydrochloride content in these samples is shown in figure 3. The heroin hydrochloride content in these 47 samples ranged between 8 and 48 per cent, and for about 65 per cent of them it varied between 36 and 48 per cent.
Caffeine concentration ranged between 22 and 69 per cent and in most samples it fell between 40 and 45 per cent. The results point to the use of a recipe whereby the prepared heroin hydrochloride is mixed with an equal amount of caffeine. For obvious reasons some manufacturers were adding more caffeine than others.
As to the strychnine content, it was generally less than 1 per cent, which does not constitute a threat to human life. However, one sample contained 15 per cent strychnine, a concentration that can be considered as extremely dangerous [ (4] ), [ (5)] , (6).
Thin-layer chromatography and high pressure liquid chromatography were used in the analysis of the samples seized in September 1976. During that month, 146 heroin samples were received and 10 per cent of them weighed more than 10 grams.
We also received some samples which resembled heroin, but which did not contain the drug. These samples consisted sometimes of a fertilizer, sometimes of a light-brown cat litter.
All the heroin samples seized during September 1976 were of the No. 3 type; the bigger lots consisting of granular material, the smaller lots often of powder. Also most of these samples contained only traces of morphine, and no detectable amounts of codeine.
In some samples, the acetylation reaction did not go to completion and considerable amounts of morphine and codeine could be detected. All samples contained strychnine, and some showed traces of quinine and/or phenacetin while barbitone was present (15 and 18 per cent) in only two samples.
In 14 samples, a component was discovered which gave an intense blue fluorescence under UV-light and reacted with the iodoplatinate reagent. Thin-layer chromatography, gas chromatography and infra-red analyses proved this component to be quinine ethylcarbonate.
The distribution of the contents of heroin hydrochloride, caffeine, O 6-monoacetylmorphine hydrochloride, acetylcodeine hydrochloride and strychnine are represented in figures 4 through 8. The results show that the samples seized in September 1976 can be classified into three groups. The first, into which fall 70 per cent of the samples, had a heroin hydrochloride content between 36 per cent and 52 per cent, the largest number of samples containing about 43 per cent of the substance. In the second and third groups, heroin hydrochloride ranged between 20 per cent and 32 per cent, and 14 per cent and 18 per cent (figure 4), respectively. Also, no significant differences was found between the powdered and the granular samples. However, two samples whose appearance was not different from the others, did not contain any heroin hydrochloride. Instead they had large amounts of morphine hydrochloride and caffeine.
Figure 5 shows that on the basis of the caffeine content, three groups can be identified: the first and largest contained between 34 per cent and 48 per cent of the substance, the second which did not significantly differ from the first showed a caffeine content between 50 per cent and 56 per cent. The third and smallest group consisting of 10 samples (with a heroin hydrochloride content between 14 per cent and 18 per cent), had a caffeine content between 58 per cent and 64 per cent.
All samples had a O 6-monoacetylmorphinehydrochloride content of less man 20 per cent, the largest number of them containing 5 per cent of the substance. No separate groups could be distinguished.
Figure 7 shows that the amount of acetylcodeine hydrochloride ranged for all samples between 3 and 11 per cent, and no distinct groups could be identified on the basis of the content of this substance.
It can be seen from figure 8 that the distribution of the strychnine content m the samples is such that they can be divided into two groups. The major group shows a strychnine content (calculated as the nitrate salt) of up to 1 per cent, the largest number of samples in this group containing 0.5 per cent of the substance. The minor group contained between 1.6 per cent and 1.9 per cent strychnine, and consisted of the same 10 samples mentioned earlier whose caffeine and heroin
Distribution of the contents of the alkaloids in the samples of September 1976.
hydrochloride contents were 60 per cent and 17 per cent, respectively. This same group of 10 samples had similar contents of O 6-monoacetylmorphinehydroehloride and acetylcodeine hydrochloride.
By means of a computer each sample was compared with every other sample. Were declared identical all samples which did not differ in any component's content by more than 6 times the largest standard deviation, shown in table I. Samples which were found to be alike were placed in one group. It turned out that nearly all samples could be placed into two or more different groups, which indicates that the groups overlap.
Combining the values for all five components of a particular sample with respect to the mean value of those individual components in all samples, we tried to discover separate groups in another way, by using the formula:
|
L=??
|
(x - ?%/)
|
2 |
|
?%
|
in which x is the percentage of a component in a sample, ? is the mean of the percentage of this component in all samples and ? the standard deviation of the mean of this component. This formula resembles that used for calculating the eccentricity in gaussian distributions. This value is independent of the standard deviation used when comparing each sample with every other sample, as mentioned before. From figure 9 it appears that there is a group of 12 samples with L values between 4.5 and 5.5. To this group belong the 10 samples mentioned earlier. The remaining 2 samples were those which did not contain any heroin.
Qualitative analyses showed 14 samples to contain quinine ethyl carbonate. Combining qualitative and quantitative data, we could divide this group into 4 subgroups (table II), with one sample not belonging to any of these subgroups. The results indicate that these 14 samples originated from at least 4 different batches, which implies that the 146 heroin samples considered could have originated from 40 different batches.
With the standard, deviations obtained in these studies, it was not possible to classify 146 samples into 40 batches within rather narrow concentration-ranges. This was because the difference between two consecutive batches would become far smaller than 6 times the standard deviation. Other factors interfering strongly with this classification are the lack of homogeneity within one batch and the chemical instability of the heroin samples.
FIGURE 9 - Distribution of the "L-values" (see text) in the samples of September 1976.
Most heroin samples confiscated in the Netherlands were of the so-called Hong Kong No. 3 type with a wide variation in composition; however, in most of them, the heroin and caffeine contents varied within relatively narrow ranges.
Strychnine, which is an essential component of No. 3 heroin, was present in almost all samples, but with the exception of one case, its concentration was relatively low.
The smaller amounts found on the addicts were not less potent than the bigger lots seized from dealers and traffickers. Also no difference was found between the powdered and the granular samples, the latter being considered as the original product from Asia. The results described here indicate that in the Netherlands heroin has not been cut by dealers until now.
We attempted to classify the heroin samples seized in September 1976 into groups or batches. However, only in a few cases where samples were characterized by special additives or by an unusual quantity of one or more components could we conclude that they belonged to the same batch. It turned out, in many cases, that differences in composition between samples can lead to the conclusion that they do not belong to the same batch. Finally, the results of these analyses gave no indication about the number of production units involved.
|
Relative standard deviation measured from |
Caffeine |
Heroin |
Acetyl-codeine |
Mono-acetyl-morphine |
Strychnine |
||
|---|---|---|---|---|---|---|---|
|
a
|
7 |
determinations of a homogenized sample
|
1.8 | 1.8 | 1.3 | 1.4 | 7.9 |
|
b
|
7 |
determinations of a granular sample
|
1.4 | 1.7 | 2.1 | 2.7 | 7.5 |
|
c
|
10 |
samples probably belonging to one batch
|
1.7 | 4.1 | 5.3 | 7.0 | 4.7 |
|
Case No. |
Quinine ethyl-carbonate |
Phenacetin |
Compound D (unidentified) |
Compound E (unidentified) |
Caffeine |
Heroin |
Acetyl-codeine |
Mono-acetyl-morphine |
Strychnine |
Subgroup |
|---|---|---|---|---|---|---|---|---|---|---|
| 23891 |
+
|
+
|
+
|
+
|
38.3 | 37.2 | 7.9 | 10.1 | 0.9 |
a
|
| 23895 |
+
|
+
|
+
|
+
|
37.6 | 36.9 | 7.8 | 11.3 | 0.8 |
a
|
| 24047 |
+
|
52.8 | 24.4 | 2.6 | 11.5 | 0.3 |
b
|
|||
| 24049 |
+
|
+
|
54.2 | 26.1 | 5.8 | 8.8 | 0.3 |
c
|
||
| 24050 |
+
|
+
|
53.0 | 26.5 | 5.9 | 9.0 | 0.4 |
c
|
||
| 24052 |
+
|
51.8 | 23.4 | 2.5 | 12.5 | 0.4 |
b
|
|||
|
24141-1
|
+
|
52.5 | 25.0 | 2.9 | 12.1 | 0.5 |
b
|
|||
|
24141-2
|
+
|
51.6 | 25.1 | 2.6 | 12.1 | 0.4 |
b
|
|||
|
24307-1
|
+
|
+
|
54.6 | 26.7 | 6.3 | 8.1 | 0.5 |
c
|
||
|
24307-2
|
+
|
54.1 | 17.0 | 3.5 | 14.7 | 0.9 |
?
|
|||
|
24307-3
|
+
|
+
|
53.1 | 26.7 | 6.3 | 7.9 | 0.3 |
c
|
||
| 24562 |
+
|
+
|
+
|
+
|
37.8 | 41.3 | 6.6 | 5.1 | 0.8 |
d
|
|
24575-4
|
+
|
+
|
+
|
+
|
35.5 | 38.4 | 6.1 | 4.9 | 0.8 |
d
|
|
24575-5
|
+
|
+
|
+
|
+
|
36.0 | 38.7 | 6.2 | 5.1 | 0.8 |
d
|
The authors wish to thank Mr. C.I. Agasi of the "Verenigde Pharmaceutische Fabrieken" for supplying O [ 6] -monoacetylmorphine hydrochloride and acetylcodeine hydrochloride for reference purposes.
R. Verpoorte and A.B. Svendsen, J. Chromatog., 100, 231 (1974).
002"Heroin Manufacture in Hong Kong", confidential report of the Royal Hong Kong Police Narcotics Bureau, September 1973.
003I. Dainis, "Opiates in Australia", confidential report of the Australian Bureau of Narcotics, June 1975.
004D. Eskes and J.K. Brown, Bulletin on Narcotics, XXVII: 1, 67, 1975.
005S. Moeschlin, Klinik und Therapie der Vergi/tungen, Georg Thieme Verlag, Stuttgart, 1972.
006R.H. Dreisbach, Handbook of Poisoning, Lange, 1974.
