A Color Change Was Not Necessarily a Positive Cannabis Test in a Duquenois Study
A chemical reaction changing color is not automatically a positive marijuana identification. A laboratory study of Duquenois-Levine chemistry recorded red, pink and yellow reactions from several non-cannabinoid compounds, while its method defined a purple hue as positive. That distinction limits what the experiment can establish about false-positive cannabis tests.
The experiment examined chemistry, not people
Rubiano, St Firmin and Coticone tested substituted benzene compounds to investigate how their molecular structures affected reactions with Duquenois reagents. This was bench chemistry using chemical compounds, not a human exposure study, animal experiment or cell-culture assay. It did not analyze blood, urine or saliva from suspected cannabis users.
The method described mixing the compounds with the reagent, waiting three minutes, adding hydrochloric acid and recording the observed color immediately. The authors were investigating a proposed reaction mechanism rather than estimating a field kit’s diagnostic accuracy in a representative series of seized substances.
The actual colors matter
Table 1 reported red for aniline, pink for phenol and yellow for acetanilide. Toluene and bromobenzene were recorded as clear. Those observations should retain their original color descriptions: relabeling all visible reactions as purple would change the evidence.
Table 2 showed that the positions of chemical groups also mattered. The 1,3-dihydroxybenzene compound produced pink, while the 1,2 and 1,4 forms showed no color change. Likewise, 3-methoxyphenol produced dark red, whereas 2-methoxyphenol was clear. The unchanged results are part of the findings, not inconvenient observations to omit.
The authors interpreted the pattern as support for the role of electron-donating groups and the arrangement of substituents in producing colored reaction products. However, the report supplied qualitative observations rather than replicate counts, confidence intervals or a formal statistical comparison of color categories.
A mechanism study is not a false-positive-rate study
The paper’s introduction described an organic-solvent extraction step associated with the Duquenois-Levine procedure. Its experimental section described reagent addition, acid addition and immediate color observation, without describing that extraction step for the reported compound tests. The tables therefore should not be presented as a validation of every stage of a complete commercial test procedure.
Nor can these few compounds establish how frequently officers or laboratories would misidentify unknown material. A false-positive-rate estimate would require a defined decision rule, an appropriate set of known negative materials and documented classification results. This paper did not report that kind of diagnostic-performance experiment.
The fair conclusion works in both directions. Non-cannabinoid compounds producing colors warrant attention to reaction specificity and interpretation. But a pink or red result is not, by itself, proof that the same compound produced the procedure’s stated purple endpoint or passed an unreported extraction-stage criterion.
Keep substance identification separate from biological exposure
A test of a chemical material also answers a different question from a test of a person’s specimen. In separate controlled human research, Newmeyer and colleagues collected oral fluid after cannabis smoking and measured cannabinoids using two-dimensional gas chromatography–mass spectrometry. That biological-matrix experiment cannot be replaced by observing a color in an unrelated chemical reaction.
Neither identifying a substance in collected material nor detecting a cannabinoid in oral fluid automatically reconstructs an individual’s impairment during an earlier event. The specimen, analyte, procedure and actual decision criterion must accompany the result.
What to request when reviewing a color-test result
Obtain the exact procedure and kit identity, the recorded color and timing, whether an extraction or layer-specific observation was required, and the laboratory’s independent identification results. Review what the operator actually observed against the specified endpoint before describing a reaction as positive.
Other color reactions have their own validation questions. Our discussion of CBD oil color testing and matrix effects addresses a different assay and should not be treated as interchangeable with Duquenois-Levine testing.
Primary research
Rubiano C, St Firmin J, Coticone SR. The effect of benzene ring substituents on the mechanism of Duquenois Levine (DL) test for cannabinoid detection. Journal of Chemical and Pharmaceutical Research. 2014;6(6):1261–1264.
Newmeyer MN, Desrosiers NA, Lee D, et al. Cannabinoid disposition in oral fluid after controlled cannabis smoking in frequent and occasional smokers. Drug Testing and Analysis. 2014;6:1002–1010. doi:10.1002/dta.1632.
