There is no unified equation that would correlate gas chromatographic retention indices of any organic compounds with their normal boiling points (at atmospheric pressure). Nevertheless, RI estimation with acceptable accuracy is possible using principal physicochemical properties, namely, boiling points, relative densities, and indices of refraction. The main feature of the algorithms proposed is the application of a logical criteria for comparing so-called indices of boiling points I(Tb), molecular weights I(M), and molar refractions I(MRD). If the comparison of these indices confirms the approximate equality of any two or all three of them (that is typical for nonpolar organic compounds), then as the best estimate of GC retention indices (RIcalc) we can accept I(Tb) values without any additional calculations. Such comparison is formally equivalent to the application of logical criteria for I(Tb), I(M), and I(MRD) values. For polar analytes, the relation I(Tb) > I(MRD) > I(M) is valid, and in these cases the inequality I(Tb) > RIcalc > I(M) appeared to be correct. Instead of large intervals between I(Tb) and I(M) for estimating RIcalc, the use of the relation RIcalc ≈ (1 - a)I(Tb) + aI(MRD) + b is proposed. The indices of molecular weights are not used for the calculations directly, but they are needed just for application of the logical criteria. The average accuracy of RI estimation by the method proposed is 9-16 index units for nonpolar compounds of any chemical nature and up to 27 for polar analytes. This high accuracy is comparable to that of the more sophisticated modern algorithms. The restriction of this approach is the impossibility of applying it to poly- and perfluoroorganic compounds.

Confirming the stability of analytes during gas chromatographic (GC) analysis is an important criterion, especially for previously uncharacterized compounds. However, the variations of absolute peak areas at different injector temperatures usually do not allow us to reveal the thermal instability of analytes during GC analysis. Such variations may be caused by peak area known discrimination typical for using capillary columns, especially at low split injection. The thermal instability can be revealed only when using relative peak areas. The dependences of the relative peak areas of unstable analytes on the injector temperature (descending), as well as similar dependences for products of their decomposition (ascending), are characterized by the existence of two limits. Low-temperature limits correspond to the initial quantities of unstable analytes, and hightemperature limits, to their complete transformations. Such dependences can be approximated with an equation of logistic regression (synonymous: sigmoid or Boltzmann approximation). The relationships and features of gas chromatographic analysis of thermally unstable compounds are considered with products of free-radical isopropylbenzene (cumene) chlorination and with solutions of ethyl diazoacetate in different solvents as examples. The major cumene chlorination product, (1-chloro-1-methylethyl) benzene, undergoes dehydrochlorination at injector temperatures above 200 degrees & Scy; to form a single product, alpha-methylstyrene. The analysis of the ethyl diazoacetate solutions is accompanied by the formation of ethyl alkoxyacetates, products of the insertion of intermediate ethoxycarbonylcarbene into O-H bonds of alcohols, if they are used as solvents. Comparing the temperatures of half-conversion of the initial ester, T(50 %), and halfformation of the products shows that they are equal to each other. This confirms both the process mechanism and the correctness of the data approximation.

The potential of a new algorithm for comparing experimental and reference values of gas chromatographic retention indices (RIs) is discussed. This algorithm is designed to minimize significant elements of uncertainty typical of numerous contemporary recommendations, primarily, the fixed limiting values of permissible deviations between experimental and reference RI-values, ΔRI = (RIref − RIexp). The algorithm proposed implies the calculation of deviations, ΔRI, for the most reliably identified constituents of multicomponent mixtures in different parts of chromatograms with known reference RI values, followed by calculation of coefficients of regression equations ΔRI = (RIref − RIexp) = aRIexp + b for both of the reduced sets of analytes. This equation allows for the recalculation of experimentally determined RIs into corrected values RIcorr = RIexp + ΔRI, which means replacing the fixed “global” limits with data-dependent adaptive thresholds for different constituents of multicomponent samples. Such an algorithm makes it possible to use reference RI values for semi-standard nonpolar polydimethylsiloxane phases (with 5% phenyl groups and others) for the comparison with data determined with standard nonpolar polydimethylsiloxanes and vice versa, as well as to minimize the influence of possible erroneous reference RI data. It is applicable both to statistically processed reference data and to results of single measurements. Both of these kinds of reference data are known and presented in contemporary RI databases, e.g., in the NIST RI database.

With the use of the products of free-radical chlorination of indane as an example (data for tetralin are mentioned for comparison), it was shown that only their molecular formulas can be determined as a result of gas chromatography–mass spectrometry analysis, and the reaction mechanism was detailed on this basis. The structure refinement of the components requires an interpretation of their gas-chromatographic retention indices in the form of local additive schemes created for solving a specific problem under consideration. These schemes are sufficiently reliable if they are formed on the basis of data for objects containing all the structural elements of the molecules of characterized compounds. If this condition is not met (as in the case under consideration), the possibility of interpreting retention indices is fundamentally preserved; however, the accuracy of their estimates is significantly reduced and the volume of preliminary calculations increases considerably. The key fragment of the structure of chlorinated indane derivatives, which was almost not represented among the compounds characterized to date, is the chlorine atom in the alicyclic structural fragment in the α-position to the aromatic system. Secondary chlorine derivatives are formed as a result of dehydrochlorination of the primary reaction products with the subsequent addition of chlorine at the C=C double bonds.

The gas-chromatographic retention indices (RIs) of trimethylsilyl (TMS) derivatives of the simplest amino acids on standard nonpolar polydimethylsiloxane stationary phases were systematized. This processing of data included combining them for derivatives of the same amino acids depending on the number of TMS groups (from one to four) and calculating average RI values together with their standard deviations based on data from various sources of information. This form of presenting the results made it possible to identify the best characterized derivatives and evaluate the reliability of the retention indices known for them. The simplest additive scheme for calculating retention indices based on even limited data for the most common amino acids was formed to estimate their unknown values, control previously determined values, and identify erroneous data. The increment ΔRI = RI(bis) – RI(mono) for the transformation −CO2Si(CH3)3 + −NH2 → −CO2Si(CH3)3 + −NHSi(CH3)3 was well reproducible (118 ± 9). The other increments ΔRI = RI(tris) – RI(bis) were different for the transformations −NHSi(CH3)3 + XH → −N[Si(CH3)3]2 + XH (238 ± 35) and −NHSi(CH3)3 + XH → −NHSi(CH3)3 + −XSi(CH3)3 (111 ± 16). A method for monitoring the correctness of the obtained values of ΔRI was proposed.

The unique potential and different areas of applications of recurrent (synonymous: recursive) relationships in chemistry and chromatography are considered. Recurrent relations can be used in two forms: as functions of integer arguments, y(x+ 1) = ay(x) + b, and as functions of equidistant argument values, A(x+ Δx) = aA(x)+ b, Δx= const. The first form applies to all physicochemical properties of homologs in organic chemistry, because the number of carbon (and other) atoms in a molecule can be integer only. The second one applies to chemical variables depending on temperature, pressure, concentrations, etc., when the chemists should provide equal “steps” of their variations. Recurrent relations combine the properties of arithmetic and geometric progressions, which accounts for their unique approximation abilities. This was illustrated by approximating the number of isomers of alkanes, the boiling points of homologs (nonlinear dependencies), the melting points of homologs (alternation effects), the temperature dependence of the solubility of inorganic salts in water, and by revealing the anomalies of gas chromatographic retention indices and retention times in reversed-phase high performance liquid chromatography.

Background: Recurrent approximation of retention parameters in reversed-phase HPLC is effective for revealing anomalies that are otherwise difficult to detect, namely, the reversible hydration of analytes. This was demonstrated previously for restricted sets of analytes with acetonitrile–water eluents. Expanding the number of analytes and eluents seems to be a topical problem. Two kinds of derivatives of aromatic carbonyl compounds were characterized: unsubstituted hydrazones and oximes. Methods: If analyte demonstrates no anomalies in dependences of retention times vs. concentration of organic modifier, the recurrent approximations of these dependences are linear. To explain the features of recurrent approximations, the numerical experiments were proposed and considered. The artificial shifting of one, two, or more points allows for the modeling the different kinds of deviations of approximations from linearity. Results: It was shown that hydrazones are the class of analytes having no anomalies of retention parameters. On the contrary, several anomalies were detected for oximes. Downward deviations of points in the plots of recurrent approximations of retention times are the signs of reversible hydration. This effect for methanol–water eluents was detected for the first time. Some of oximes underwent hydrolysis. Conclusions: Recurrent approximation of retention times allows detecting chemical transformations of analytes during RP HPLC analysis.

Anomalies in the chromatographic retention of sorbates in reversed phase HPLC are often attributed to variations in their mechanisms of retention. However, an equally important reason seems to be a change in the chemical nature of sorbates due to interaction with components of the eluent. Chromatographic properties of several oximes of aromatic carbonyl compounds in high-performance reversed phase liquid chromatography are characterized, including their retention indices and results from recurrent approximation of the dependences of sorbate retention on the concentration of the organic component of the eluent. Such approximation allows us to identify considerably more anomalies in the retention times than other means. Chromatographic information is supplemented by spectral parameters, specifically relative optical densities Аrel = А(λ1)/А(λ2). Compounds that are stable under conditions of separation are found in the series of oximes, along with examples of reversible hydration (oximes of 2-methoxy- and 3,4-dimethoxybenzaldehydes) and irreversible hydrolysis (oximes of 2- and 4-hydroxybenzaldehydes, acetophenone) with the formation of the corresponding aldehydes. It is shown that coefficients of the dependence of retention indices on the concentration of the organic component of the eluent for aldehydes mainly satisfy inequality dRI/dC > 0, and they are usually negative for their oximes. The differences between indices of retention of retention ΔRI = RI(oxime) − RI(aldehyde) in high-performance reversed phase (RP) liquid chromatography are consequently not constant. Instead, they fall as the concentration of methanol in the eluent rises.

The oxidation of hydroquinone with hydrogen peroxide in the presence of catalytic amounts of FeSO4 results in complex mixtures of oligomers. The average composition of these products is determined by the molar ratio of reagents and varies from (C6H4O4)n at the hydroquinone to hydrogen peroxide ratio 1 : 3 to (C6H4O6)n at the ratio 1 : 5. A characteristic feature of MALDI mass spectra for polymers is the periodicity of signals. However, no such periodicity is observed for the products of hydroquinone oxidation within the m/z range <1000 Da, indicating the absence of fixed sequences of structural fragments. This unique feature arises from the variable ratios and irregular positions of polyhydroxyphenylene and polyhydroxybenzoquinone fragments. Additionally, several factors contribute to the lack of mass-spectral periodicity, namely, the potential formation of peroxides, the presence of stable hydrates, various types of linkages, and possible interactions between hydroxylated phenylene and benzoquinone fragments. In contrast, for m/z values >1000 Da, MALDI mass spectra exhibit periodicity with a mass difference of Δ(m/z) = 74. This value likely corresponds to the fragment C2H2O3, which is neither a hydroquinone nor a benzoquinone structural unit. Taking into account that an equal Δ(m/z) value is observed for fragment ions in the EI mass spectrum of tetrahydroxy-p-benzoquinone, this gives indirect evidence for the presence of polyhydroxybenzoquinone fragments within the oligomers. A key issue regarding the structure of hydroquinone oxidation products is the type of linkages between the polyhydroxyphenylene and/or polyhydroxybenzoquinone units, which could involve C–C or C–O–C bonds (or both). The available spectral data do not resolve this issue conclusively. However, the presence of a sample of the composition (C6H4O6)n, in which each unit contains six oxygen atoms, suggests that at least one oxygen atom forms a bond between the units, indicating a C–O–C connection. Overall, the unique nature of oligomeric products of hydroquinone oxidation explains their high structural variability and redox properties, which likely contribute to their unique pharmacological characteristics.

При окислении гидрохинона пероксидом водорода в присутствии каталитических коли­честв суль­фата желе­за(II) образуются сложные смеси олигомеров, усредненный сос­тав которых опреде­ляется соот­ноше­ни­ем реагентов и для рассматриваемых об­разцов варь­ирует от (С6Н4О4)n при мольном соотношении гидрохи­но­на и Н2О2 равном 1:3 до (С6Н4О6)n при со­отношении 1:5. Масс-спектры MALDI полимеров обычно отличают­ся пе­рио­ди­ч­но­стью сигналов в масс-спектрах, но для продуктов окисления гидрохинона в об­ла­­­сти m/z < 1000 Да ка­кие-ли­бо эле­мен­ты периодичности массовых чисел не вы­явле­ны. Это свиде­те­ль­ствует об отсутствии по­следо­ва­тельностей од­но­ти­п­ных зве­нь­ев в таких оли­гомерах, что следует считать их уникальной особенностью, обус­лов­ленной, прежде все­го, переменным соотно­шением, нере­гу­­ляр­­ным че­ре­дованием и легкостью взаимного превра­ще­ния по­ли­гидрок­си­фениленовых и полигидро­к­си­­бен­зо­хино­но­вых структурных фра­г­мен­тов. Дополнительно к отсут­ствию пе­рио­дичнос­ти массовых чи­сел сигналов в этой области m/z приводит дейст­вие еще нес­ко­лько факто­ров, а именно существо­ва­ние достаточ­но ус­тойчи­вых ги­д­ратов, воз­мо­жность образования пер­окси­дов, различ­ные способы сочлене­ния эле­ментарных звеньев и, кроме того, возмо­ж­ность взаимо­дей­ст­вия фени­лено­вых и бен­­зо­хиноно­вых фрагментов с образованием комплексов с переносом заряда. Тем не менее, в области масс-спектров MALDI с m/z > 1000 Да продуктов взаимо­дей­­­ст­вия гид­ро­хинона с пе­р­ок­си­дом водорода удалось выявить периодичность сигна­лов, со­ответствующую раз­нос­ти массовых чисел D(m/z) = 74. Такая разность может быть обус­ло­в­лена частицами только состава С2Н2О3, что непосредственно не отвечает ни гидрохи­но­­но­­во­му, ни бен­зо­хино­новому фрагментам структу­ры. Поскольку такое значение D(m/z) наблюдается для осколочных ионов в масс-спектре ионизации элек­т­ронами тетрагид­рок­си-п-бензохинона, то его можно рассматривать как косвенное подтве­р­ж­дение присутствия в структуре олиго­меров полигидроксибензохиноновых фрагментов. Важной проблемой строения олигомерных продуктов окисления гидрохинона яв­ля­ется тип сочленения элементарных звеньев между по­ли­гидроксифениле­новыми и полиги­д­роксибензохиноновыми фрагмен­тами, а именно за счет связей С-С или С-О-С (или и тех, и других). Доступная спектральная информация не позволяет непосредственно от­ве­тить на этот вопрос. Однако сущест­во­ва­ние образца состава (С6Н4О6)n, содержащего шесть ато­мов кислорода в элементар­ном зве­не, рационально можно объяснить, только ес­ли один из атомов кислорода образует связь между звеньями, т.е. их соединением по типу С-О-С. В целом же уникальность олигомерных продуктов окисления гид­ро­хинона состоит в высо­кой вариабельности их структур и редокс-характеристик, что, по-ви­димому, и опре­де­ляет как сложности установления их струк­ту­ры, так и уникальные фармакологические свой­ст­ва.

Regularities and peculiarities of gas chromatographic analysis of thermally unstable compounds were considered on the example of mixture of the reaction products of isopropylbenzene (cumene) free-radical chlorination. The principal constituent in this mixture is (1-chloro-1-methylethyl)benzene, which has the lowest thermal stability, and is partially converted to a-methylstyrene, the only product of its thermal destruction at the chromatograph injector temperatures up to 300 °С. Nevertheless, the results of the study confirms that the gas chromatographic analysis of chloroalkylarenes is possible without their decomposition with the injector temperatures up to 200 °С, even if the analytes contain chlorine atoms at the tertiary carbon atoms and in the “benzylic” positions relative to the aromatic fragment. Similar control of thermal stability of analytes can be recommended for other samples contained potentially unstable constituents. It is shown that thermal decomposition of thermally unstable constituents of samples cannot be revealed from the results of gas chromatographic analysis with capillary columns using variations of their absolute peak areas. Such task can be solved only by using relative peak areas calculated in respect to thermally stable compounds. The dependencies of relative peak areas of unstable constituents vs. temperature (descending), as well as those of their decomposition products (ascending) are characterized by presence of two limits. Low temperature limits correspond to the real content of unstable constituents or their decomposition products is the samples, while the upper limits – to the composition of such samples at their hypothetically complete destruction. Such dependencies can be approximated by logistic regression equation if sampling into capillary columns is carried out at relatively high split ratios (approx. not less than 10 : 1). At lower split ratios the temperature dependencies of peak areas of unstable constituents and products of their transformation are strongly distorted by so-called sample’s composition discrimination effects that make impossible data approximation using logistic regression.

Using three different synthetic methods, 48 monoalkyl esters of 6 aliphatic dicarboxylic acids and 8 aliphatic alcohols were synthesized. Despite the method of synthesis, monoalkyl alkanedioates occurred in reaction mixtures along with corresponding dialkyl esters. This was caused both by a reaction mechanism and by possible secondary bimolecular disproportionation in solutions. However, for example, unlike monoalkyl phthalates, monoalkyl esters of aliphatic dicarboxylic acids appeared to be stable during their gas chromatographic separation. In no cases, characteristic chromatographic profiles indicating the decomposition of analytes in the chromatographic column were observed. The simultaneous presence of mono- and dialkyl esters in the samples allowed us to use the correlation of retention indices of mono- and dialkyl esters for their identification because the latter were characterized in more detail, namely, with RI(monoester) = aRI(diester) + b and some similar relations. The chromatographic peaks of monoalkyl alkanedioates (like for all carboxylic acids) on columns with nonpolar stationary phases were highly asymmetric. Their electron ionization mass spectra naturally indicated the presence of no signals of molecular ions. Mass spectra of monomethyl and monoethyl esters differed markedly from those of corresponding dialkyl esters, but they were similar for alcohols with R ≥ C3H7.

Unsubstituted hydrazones RR′C=NNH 2 are unstable during gas chromatographic separation. Testing of their resistance to hydrolysis under reversed-phase HPLC showed that aromatic ketone hydrazones are stable. In contrast, aldehyde hydrazones are only stable in neutral methanol–water systems (in the absence of acidic modifiers). In acetonitrile–water systems containing 0.1% of formic acid, only aromatic ketone hydrazones are stable, while aldehyde derivatives are completely hydrolyzed. This difference in stability must be taken into account in determining other compounds of these classes. To detect the hydrolysis of analytes, we compared the retention indices of the initial carbonyl compounds and hydrazones at different volume ratios of organic modifiers and aqueous phases and different pH values of the eluent, the relative absorbance values of the characterized components A (254/220) = A (254)/ A (220), and the chromatography–mass spectrometric data.

Characterizing the dependence of the retention indices (RIs) of sorbates of different chemical origins on the content (С) of organic solvents (methanol, acetonitrile) in eluents used in reversed-phase HPLC shows that the coefficients dRI/dC of the same compounds for eluents of different compositions differ considerably. The range of variations in coefficients dRI/dC for water–methanol eluents (–5.6 to +4.6) is approximately twice that of water–acetonitrile eluents (−4.5 to +0.8). It is also found that for nonpolar compounds when using water–acetonitrile eluents, the dRI/dC values are typically lower than those for water–methanol eluents. The opposite is observed for polar compounds. The signs and absolute values of coefficients dRI/dC are largely determined by the polarity of the sorbates. dRI/dC > 0 for nonpolar sorbates, and dRI/dC < 0 for sorbates of higher polarity. Determination even of the sign of this coefficient therefore provides important information about the chemical nature of sorbates. It is shown that coefficients dRI/dC do not correlate with either the RI values or hydrophobicity log P . However, they do correlate with homologous increments of hydrophobicity ilogP and the homologous increments of RI iRI for eluents of any composition. Quantities ilogP and iRI characterize the polarity of homologous series of sorbates, rather than the polarity of the particular sorbates.

Retention parameters of sorbates in reversed-phase high-performance liquid chromatography tR depend to the greatest extent on concentration of the organic component C in the eluent. To find weaker pronounced effects against such a “background”, approximation of the retention parameters with the use of recurrent first-order correlations tR(C + ΔC) = atR(C) + b is required. For the sorbates that do not form hydrates (Khydr $$ \ll $$ 1) or exist in a chromatographic column in the hydrated form only (Khydr $$ \gg $$ 1), the corresponding dependences are linear (the correlation coefficients generally exceed 0.999). If, however, hydration constants of the sorbates Khydr ≈ 1, deviations from linearity are observed. Such abnormalities are the most pronounced in the case of the use of acetonitrile–water eluents, while they are less noticeable in the case of methanol. The reason for this most likely consists in the fact that methanol forms more stable hydrates in comparison with other compounds.

Some polar analytes (X) can reversibly form hydrates in water-containing eluents under the conditions of reversed-phase HPLC analysis, X + H2O ⇄ X × H2O. One of the methods to detect their formation is the recurrent approximation of the net retention times of such analytes, tR(C + ΔC) = atR(C) + b, where ΔC = const is the constant step in the variation of the organic modifier content of an eluent. These dependencies are linear if hydrates are not formed, but in the case of hydrate formation, they deviate from linearity under high water content. It has been shown that UV spectroscopic parameters, namely, relative optical densities: Arel = A(λ1)/A(λ2), depend on eluent composition for some organic compounds, but their variations cannot be used as indicators for hydrate formation. The coefficients that characterize the dependence of the analyte retention indices on the organic component concentration of an eluent, dRI/dC, appeared to be the most informative additional criterion for hydration. The values of these coefficients for most polar analytes are largely negative (dRI/dC < 0), whereas, for nonpolar compounds, they are largely positive (dRI/dC > 0).

Gas chromatographic analysis of a long-stored sample of tetraethoxysilane (tetraethyl orthosilicate) showed that it lacked the main component due to the hydrolysis of this compound by traces of water. Instead, ethanol and three other components with retention indices (RIs) of 537 ± 2 (no. 1), 608 ± 1 (no. 2, the most abundant component), and 727 ± 3 (no. 3) on a column with a HP-5 stationary phase were detected. These components are unstable, and they cannot be isolated preparatively; as a result, they have not been characterized previously. To identify them, the chemical properties of this sample were characterized, and the recurrent approximations of the RIs of the detected components and their correlation with the retention indices of structural analogs were considered. It was established that they were congeners of the initial tetraethoxysilane, namely, exotic products of its partial hydrolysis—triethoxysilanol (C 2 H 5 O) 3 SiOH, diethoxysilanediol (C 2 H 5 O) 2 Si(OH) 2 , and ethoxysilanetriol (C 2 H 5 O)Si(OH) 3 . In accordance with published data, some silanediols R 2 Si(OH) 2 and silanetriols RSi(OH) 3 are stable, especially, compounds containing substituents capable of conjugation with vacant d orbitals of silicon atoms. Among them are phenyl- (π– d conjugation systems) and alkoxy-substituted ( p–d systems) silanediols and silanetriols. The identified products of partial hydrolysis of tetraethoxysilane belong to the latter type.

Chromatographic analysis of a series of two-component samples (solutions of a target analyte and a standard) allows simultaneous comparison of the possibilities and features of six variants of quantitative analysis using the external and internal standards techniques according to the criteria of results precision and repeatability (random and systematic errors values). These variants include the simplest version of the external standard method (I); its modified version (II) that implies the application of an additional standard and averaging not the absolute, but relative peak areas; the commonly used version of the internal standard method (III); and its version modified in a similar manner (IV). Besides, two variants of using a homologue of the target analyte as the internal standard without determining the calibration coefficients are considered (V) and (VI). This topic is of interest for optimization of practical works on chromatography and teaching the subject in general. The requirements to the additional and internal standards are not identical. Any compound (both present in the samples or added to them) can be selected as the additional standards, and precise setting of their concentrations is not required. It is only necessary to ensure their equal concentrations in the analyzed and reference solutions. It is shown that the modified versions of the external and internal standard methods are characterized by the equal relative standard deviations of the results. The minimal relative standard deviations of the results are typical for the variants implying the use of an additional (II) or an internal (IV) standard and the averaging the ratio of peak areas of target analytes and such standards. The systematic errors of determinations appeared to be minimal for the same variants. Controlling the values of the calibration coefficients is informative for revealing the possible distortions of the composition of samples due to the partial evaporation of volatile constituents in the course of handling such samples.

Despite the constant improvement of complex computer algorithms for the calculation of gas chromatographic retention indices (RIs), the simplest methods for their evaluation based on linear correlations between the indices of structural analogs from different taxonomic groups, RI1 ≈ aRI2 + b, remain important. It is shown that symbatic variations of the first numerical differences of retention indices, Δ _RI^1 = RIn + 1 – RIn (equivalent to the first derivatives of the retention indices with respect to the structural parameter varied in the group), are the conditions of correctness for such correlations in the simplest groups (substituted methanes). A monotonic variation of Δ _RI^1 in one of the groups with the presence of extrema in the other group is an unequivocal sign of the absence of a linear correlation between retention indices. If the values of Δ _RI^1 in one of the groups increase and decrease in the other, the ranking order of compounds in any one of them should be reversed. It is shown that the simplest relationship RI1 ≈ aRI2 + b is also applicable to more complex taxonomic groups (substituted ethanes, benzenes, and naphthalenes), and it allows one not only to estimate the RIs of compounds not yet characterized but also to refine known reference data.