Polycyclic aromatic hydrocarbons identified in soot extracts from domestic coal-burning stoves of Henan Province, China.

Using high-pressure liquid chromatography with ultraviolet-visible diode-array detection, we have analyzed polycyclic aromatic hydrocarbons (PAH) in the dichloromethane extracts of soot deposits from coal-burning stoves in several homes of Henan Province, China--including Linxian County, where esophageal cancer rates are some of the highest in the world. Thirty-two individual polycyclic aromatic compounds, ranging in size from three to eight fused aromatic rings, have been unequivocally identified among the soot extract components--including 20 benzenoid PAH, 6 fluoranthene benzologues, 1 cyclopentafused PAH, 1 indene benzologue, 3 oxygenated PAH, and 1 ring-sulfur-containing aromatic. Most of the identified compounds have been observed before among the products of laboratory coal pyrolysis experiments, but two of the components, the six-ring C24H14 napthol[1,2-b]fluoranthene and the eight-ring C30H16 tribenzo[e,ghi,k]perylene, have never before been documented as products of coal in any system. All of the Henan coal soot extracts are remarkably similar qualitatively in that they contain the same set of identified PAH, but absolute levels of individual species vary by up to 5 orders of magnitude, from sample to sample. The bulk of the identified component mass in all of these soot extracts lies in the five- and six-ring PAH--the largest single class being the family of five-ring C20H12 isomers, to which the samples' most abundant components, benzo[b]fluoranthene and benzo[e]pyrene, belong. The five- and six-ring PAH also account for the majority of the samples' known mutagens. The three strong mutagens identified in these soot samples are the C20H12 benzo[a]-pyrene and two C24H14 PAH, dibenzo[a,e]pyrene and naphtho-[2,1-a]pyrene. Seven moderate mutagens are found among the C20H12, C22H12, C22H14, and C24H14 PAH. A major class of mutagens, the cyclopenta-fused PAH, appears to be absent from these samples, but our detection of an oxidation product of the major mutagen cyclopenta[cd]- pyrene--itself mutagenic--suggests that these soot deposits may contain additional mutagenic cyclopentafused PAH oxidation products as well.

QSRR prediction of chromatographic retention of ethynyl-substituted PAH from semiempirically computed solute descriptors.

Retention prediction of 12 ethynyl-substituted polycyclic aromatic hydrocarbons (PAH) and their six unsubstituted parent compounds has been elucidated by the application of quantitative structure-retention relationship (QSRR) analysis. Retention data of the PAH were obtained from reversed-phase high-pressure liquid chromatography (HPLC) utilizing an octadecylsilica stationary phase operated under linear-gradient elution (60:40 water/acetonitrile to pure acetonitrile in 40 min). Six solute descriptors (moment of inertia, total energy, polarizability, ionization potential, dipole moment, subpolarity), computed from the optimized semiempirical AM1, MNDO, and PM3 solute geometries, were examined. Results from one-parameter QSRR analysis showed that retention of solutes was best predicted with solute polarizability as the parameter, computed from the AM1-(r = 0.969), MNDO-(r = 0.970), or PM3 (r = 0.967)-optimized solute geometries. From two-parameter QSRR analysis involving a size-specific parameter accompanied by a polarity parameter, it was found that solute retention was best reproduced by using solute polarizability and subpolarity as the parameters calculated from the AM1-(r = 0.983), MNDO-(r = 0.983), or PM3 (r = 0.984)-optimized solute geometries. On the basis of the results from both one-parameter and two-parameter regression analysis, the two-parameter QSRR equation with polarizability and subpolarity as parameters was found to be the best relation in relating solute molecular structure to retention under the HPLC conditions investigated. The results obtained in this study are of significance to predicting the identify of unknown product components based solely on parameters derived from solute structure.