Content
Citation(s) & Footnote(s)
Book(s)Citation and full coverage
Irina Abramova, Marina Medrish, Alexandra Romanova, Vladimir Ovchinnikov and Darya Gavrilova (2021), “The quality control system of distilled spirits,” BIO Web of Conferences 36, 05006, FSRAABA 2021. DOI: https://doi.org/10.1051/bioconf/20213605006. Seven supplied pages fully read, including every cell and footnote of Tables 1–4 and all nine reference entries. Pages 2–7 inspected as rendered pages; no figures or supplied appendices. Reading references here does not mean reading the cited works. CC BY 4.0 stated on the article. Original mirror preserved; same-name/same-size Proton copy remains a provisional match until cloud bytes can be hashed.
What the study actually does
The authors propose combining GC-FID volatile analysis, HPLC-UV phenolic/furan measurements, and ion chromatography to support quality control and authentication. Table 1 (pp. 2–3) contains 16 samples: ten rum samples INCLUDING two labeled adulterated, three tequilas and three whiskeys. The whiskeys are one three-year Scottish blend, one six-year US bourbon and one eight-year Irish malt. This is a small exploratory cross-product comparison, not a validation cohort.
GC-FID uses a CP-Wax 57 CB column after preliminary distillation and dilution to 40% alcohol. HPLC uses a C18 column, 280 nm detection and an acid/acetonitrile gradient; ion chromatography measures five cations and six anions after filtration/dilution (pp. 3–4). The three methods interrogate different chemical fractions. No sensory panel establishes whether these concentrations improve flavor or liking.
Evidence by table
Table 1: product style, age and manufacturer country are confounded. Tequila entries include age ranges rather than individual exact ages. They must not be republished as a current legal definition.
Table 2 (p. 4): R9 has 144.6 mg/L vanillin and 0.8 mg/L syringaldehyde, yielding the reported very high ratio of about 180. This supports suspicion of added vanilla, not independent proof of an adulteration mechanism. R10 has vanillin and HMF above the listed reporting boundary while other TARGET analytes are below it; it does not contain only two chemicals. Wood-derived compounds and furans can have more than one source, including cask treatment, distillation and caramel. The stated rum vanillin/syringaldehyde range 0.2–0.4 is narrower than the table: R5 is 0.81 and other ordinary rums include 0.45 and 0.43. It is not a validated authenticity cutoff.
Displayed rounded concentrations do not reproduce every printed ratio (e.g. R2 syringic acid/syringaldehyde 0.9/1.1 versus listed 0.91; R6 vanillin/syringaldehyde 0.1/0.3 versus 0.21). Underlying unrounded values are unavailable, so do not automatically call these analytical errors. W2 bourbon ellagic acid is 22.8 mg/L versus W3 Irish malt 7.3 and W1 Scotch blend 1.2. The younger bourbon's higher concentration illustrates why this table cannot establish a universal maturation clock.
Table 3 (pp. 5–6): concentrations are mg/dm3 of ABSOLUTE ALCOHOL, except methanol, whose values are percentages under the p. 6 footnote. Table 2 and Table 4 use mg/dm3 without this absolute-alcohol qualification. Preserve denominators; methanol percentage basis is not fully specified for independent compliance conversion. The ratio in Table 3 is 2-methylbutanol/3-methylbutanol, whereas introductory prose discusses the reciprocal. Neither a flipped ratio nor a concentration copied without units is acceptable evidence.
The claim that fake samples have lower volatile impurities is not universal even within these data: R10 ethyl acetate is 694.6, the highest listed across all 16 samples, and R9 isobutanol 316.4 exceeds several ordinary rums. A single 'more' or 'less' rule would misrepresent the multicomponent argument. No detection performance is reported.
Table 4 (p. 6): ions vary substantially. For Cuban R1–R4 aged 3, 5, 7 and 10 years, sodium is 1.4, 22.7, 21.2 and 19.1 mg/L; potassium and oxalate also fail a monotonic age progression. These are different products, not longitudinal barrel samples. Background claims about water treatment or mineral precipitation are not tested through source-water measurements or a sedimentation experiment. R8 and W1 share the same last three ion values (13.5, 54.0, 12.2); repetition alone does not prove copying. Values below a reporting boundary are censored measurements, not zero concentrations or proof of absence.
Critical assessment
Useful as an introductory example of complementary chemical measurement and a table-reading exercise. Weak as an authentication rulebook: only three whiskeys, confounded categories, no clearly reported replicate-bottle design, uncertainty distributions, recovery/LOD/LOQ validation detail, independent adulteration ground truth, blinded decision threshold, external test set, or sensitivity/specificity. The conclusion's 'universal and specific criteria' language is stronger than demonstrated diagnostic performance. Standards and legal statements are historical context, not current compliance guidance. No health, safety or superior-taste conclusion follows from the measurements.
Linked synthesis and proposed Academy uses
Pair with Jack and Steele (2002), whose adulterated sensory challenge was confidently classified as Scotch, to explain why complementary measurements need independent validation rather than a confident score. Pair with Kew et al. (2017), whose age model failed validation, and Mosedale's oak reviews to distinguish a wood signature from an age clock. Pair with Pawlaczyk et al. (2019) on elemental profiles to separate geographic association from source-water causation.
Proposed internal teaching exercise: give learners Tables 2–4 with product labels hidden and require them to state units, ratio orientation, reporting limits and at least two plausible alternative causes before proposing an authenticity conclusion. A second exercise compares the bourbon's ellagic acid and the nonmonotonic rum ion series with stated ages. These are proposed Academy applications; no public course page was edited.
Verification boundary
Full examination of this seven-page article is complete. This does not certify the paper's claims, reproduce its analyses, verify the Proton copy by hash, or verify browser rendering. Native Notion attachment and saved records are checked separately.
