Trends linking the topological characteristics of polynuclear aromatic hydrocarbons (PAH) to their electronic properties are reported. TD-DFT electronic spectra computations, using the 6-31G* basis set and B3LYP exchange correlation functional, were calculated for a series of PAH, allowing for the HOMO–LUMO gaps to be reported. Clar structures provide an avenue to link the physical structure and the aromaticity of the molecule; which, when extended by bond length and harmonic oscillator model of aromaticity analysis, provide powerful tools to understand the link between electronic and physical structure. These results lead to the conclusion that all PAH structures show a decrease in HOMO–LUMO gap as a function of size, but the rate of that decrease is directly related to the topology of the molecules. A PAH taxonomy was developed that categorizes PAH into categories with similar topological properties, which allows for modelling of changes in the HOMO–LUMO gap with PAH size. An atom-pair minimization algorithm was used to calculate the binding energy (BE) of homogeneous dimers of the studied PAH. The BE per carbon atom increases with the overall size of the structure to an asymptotic limit, but as with the HOMO–LUMO gap, topology plays a critical secondary factor. Previously published, experimentally determined optical band gaps (OBG) from Tauc/Davis–Mott analysis of extinction spectra in various laminar, non-premixed flames produced a correlation between the HOMO–LUMO gaps of high-symmetry, nearly circular D2h symmetry molecules to molecular size. The work presented here provides a much more nuanced and predictive evaluation of how OBG depends on structure and size.

Visible light extinction was measured in a series of nitrogen-diluted, ethylene/air, non-premixed flames and this data was used to determine the optical band gap, OBG, as a function of flame position. Collimated light from a supercontinuum source is telescopically expanded and refocused to match the f – number of a dispersing monochromator. The dispersed light is split into a power metering channel and a channel that is periscoped and focused into the flame. The transmitted light is then recollimated and focussed onto a silicon photodiode detector. After tomographic reconstruction of the radial extinction field, the OBG was derived from the near-edge absorption feature using Tauc/Davis–Mott analysis. A slight evolution in OBG was observed throughout all flame systems with a consistent range of OBG observed between approximately 1.85 eV and 2.35 eV. Averaging over all positions the mean OBG was approximately 2.09 eV for all flame systems. Comparing these results to previously published computational results relating calculated HOMO–LUMO gaps for a variety of D2h PAH molecules to the number of aromatic rings in the structure, showed that the observed optical band gap is consistent with a PAH of about 14 rings or a conjugation length of 0.97 nm. This work provides experimental support to the model of soot formation where the transition from chemical to physical growth starts at a modest molecular size; about the size of circumpyrene.

The predictions of “soot” concentrations from numerical simulations for nitrogen-diluted, ethylene/air flames are compared with laser-induced incandescence and Raman spectra observed from samples thermophoretically extracted using a rapid insertion technique. In some flame regions, the Raman spectra were obscured by intense, radiation that appeared to peak in the near infrared spectral region. There is a good agreement between spatial profiles of this ex situ laser-induced incandescence (ES-LII) and the “traditional” in situ laser-induced incandescence (IS-LII). Raman signatures were observed from low in the flame and extended into the upper flame regions. The spectra consisted of overlapping bands between 1000 and 2000 cm−1 dominated by the “G” band, near ≈1580 cm−1, and the “D” band in the upper 1300 cm−1 range. Several routines are explored to deconvolve the data including 3- and 5-band models, as well as a 2-band Breit–Wigner–Fano (BWF) model. Because the Raman signals were observed at heights below those where in situ LII was observed, we postulate that these signals may be attributable to smaller particles. The results suggest that the observed Raman signals are attributable to particulate with modest (≈1 nm) crystallite sizes. This observation is discussed in the context of current models for nascent particle formation.

A technique for the determination of species’ concentrations from the molecular growth regions of flames is presented. Samples are obtained by microprobe extraction from a nitrogen-diluted, methane/air, nonpremixed laminar flame supported on a coannular burner. Quantification of measurements is accomplished by doping the flame’s fuel flow with argon at a level to match that in the laboratory’s air. A library of 70 eV fragmentation patterns for several flame species is used in conjunction with a simplex algorithm to analyze mass spectra obtained at each flame location. Each fragmentation pattern is normalized for its integrated intensity and its ionization cross-section relative to argon. This technique provided sub-part-per-million sensitivity of a large range of major and minor carbon-containing species ranging in size from C2 to C12 hydrocarbons. This flame can be acoustically forced to oscillate at a frequency emulating natural flame flickering behavior. Time-resolved measurements are obtained using a modified quartz microprobe synchronized to open and close with the flame oscillations. The near real-time sampling and analysis time and the relatively high sensitivity make this technique preferable to other extraction-based flame measurements.

Calculations of intermolecular potentials are presented for homo-molecular and hetero-molecular clusters of 24 peri-condensed PAH spanning monomer masses ranging from 78 to 1830 Da. Binding energies of homo-molecular dimers rise rapidly with molecular size and asymptotically approach the experimentally established exfoliation energy for graphite of 5.0 kJ mol−1 (carbon atom)−1. Binding energies of hetero-molecular dimers correlate well with the reduced mass of the pair. From calculations of homo-molecular stacks, binding energies were observed to increase with each added molecule and rise asymptotically, approaching a limit which scales linearly with monomer molecular mass. These results are reviewed in the context of molecular growth in flames and in the context of astrophysical observations.

A technique for linking the physical composition of polynuclear aromatic hydrocarbon (PAH) stacks and clusters to their electronic properties is reported. Kohn–Sham HOMO–LUMO gaps are reported for a series of monomers, stacks, and clusters of six, high-symmetry PAHs (pyrene, coronene, ovalene, circumpyrene, circumcoronene, and circumovalene) generated by DFT calculations with the 6-31G* basis set and a B3LYP exchange correlation functional in NWChem. A previously published, atom-pair minimization algorithm was used to optimize the geometries of the PAH stacks and clusters. HOMO–LUMO gaps decrease with an increase in monomer size; homogeneous stacks and clusters indicate substantial lowering of the HOMO–LUMO gap because of agglomeration effects with the formation of dimers and formation of clusters (two or more stacks) being the most dominant contributions. Heteromolecular particulates had HOMO–LUMO gaps that were strongly influenced by the larger components in the system. The HOMO–LUMO gaps of homogeneous clusters approached a maximum agglomeration effect because of the localization of electronic interactions among adjacent stacks. Previously published, experimentally determined optical band gaps (OBG) from Tauc/Davis–Mott analysis of extinction spectra in various laminar, non-premixed flames had an average OBG of 2.1 eV. Based on the computations presented here, this work suggests that clusters with this OBG are comprised of modest molecular size PAH, about the size of ovalene.

Visible light extinction was measured in a series of nitrogen-diluted, ethylene/air, non-premixed flames and this data was used to determine the optical band gap, OBG, as a function of flame position. Collimated light from a supercontinuum source is telescopically expanded and refocused to match the f – number of a dispersing monochromator. The dispersed light is split into a power metering channel and a channel that is periscoped and focused into the flame. The transmitted light is then recollimated and focussed onto a silicon photodiode detector. After tomographic reconstruction of the radial extinction field, the OBG was derived from the near-edge absorption feature using Tauc/Davis–Mott analysis. A slight evolution in OBG was observed throughout all flame systems with a consistent range of OBG observed between approximately 1.85 eV and 2.35 eV. Averaging over all positions the mean OBG was approximately 2.09 eV for all flame systems. Comparing these results to previously published computational results relating calculated HOMO–LUMO gaps for a variety of D2h PAH molecules to the number of aromatic rings in the structure, showed that the observed optical band gap is consistent with a PAH of about 14 rings or a conjugation length of 0.97 nm. This work provides experimental support to the model of soot formation where the transition from chemical to physical growth starts at a modest molecular size; about the size of circumpyrene.