Comments on manuscript egusphere-2026-3391, “Molecular representation of benzene and phenol secondary organic Aerosols” (by Aurélien Le Bayon et al.)

This manuscript updated the recent studies on oxidation mechanism of benzene and phenol, particularly the mechanism of HOMs formation including formation of hydroxybenzenes by successive OH addition, the geminal diol pathways in phenols, and structure rearrangement of bicyclic peroxyl radicals. The mechanism was assessed by the experimental observations under different conditions. The study here identified the dominant role of geminal diol BPRs and their structural rearrangement. The mechanism was also reduced to a small set of 28 species and 30 reactions for greatly reduced computational demanding in 3D modeling while keeping satisfied accuracy. This may be useful in 3D modeling impact on global scale.

Major Comments:

1. Geminal diol BPR: It should be noted that geminal diol from OH addition to ipso-OH-cites often represent a minor channel in reaction of hydroxybenzenes. However, geminal diols contribute higher fractions of SOA in all simulations. Do this suggest SOA formation via Phenol-1OH is higher than that via Phenol-2OH?

2. Line 194: If F12 energies were calculated for the lowest-energy conformers only, what’s the basis for MC-TST calculations?

3.  Line 221-223: “Phenol has only one BPR isomer due to the symmetry of the molecule”. This statement is only true for Phenol-1OH adduct, which, as stated above, is the minor channel in reaction of phenol and OH. Instead, Phenol-2OH should lead to two different BPRs as Phenol-2OH-1,3-OOs-4OOa and Phenol-2OH-1,3-OOs-6OOa.

4. # Figure 4a: The C₆H₇O₈ radical is a carbonyl peroxyl radical, which might have a fast H-atom shift from the -OOH group according to the result from the Demark group.

5. Supplement S2 contains only the gas-phase reactions, while modeling study contains gas-particle partitioning and multiphase chemical processes. I would suppose the authors adopted these processes from SSH-aerosol v2.0 model. What about the ‘new’ species added in this study?

6. The authors stated “H-abstraction from the aromatic ring” (Line 80), “7%/30% abstraction branching ratio for reactions of benzene and phenols with OH” (Line 343), and “for OH addition and H abstraction in reactions of phenols with OH” (title of Table 2). I would expect all these H-abstractions were for H-abstraction from -OH group of phenol. Direct H-abstraction from aromatic ring is usually ignored under the atmospheric conditions.

7. Mechanism Evaluation as Shown in Figure 5: Mech1 and Mech2 perform obviously better than MCM. In Mech1 and Mech2, adding the overlooked HOMs formation results in new mass balance for SOA mass. The difference in this bar chart simply reflects the different branching ratios in the mechanisms. Agreement of SOA mass with the experimental data may not be able to discriminate Mech1 and Mech2.

8. Line 445: “Stronger photochemical activity and higher OH concentration in summer favor SOA formation via geminal diol pathway”. The fractions of different reaction pathways would vary with temperature, and photochemical activity and OH concentration would affect SOA formation via secondary reactions of intermediates. Importance of autoxidation would also vary with temperature and NO/RO₂ concentrations.

In winter, “reduced photochemical activity” should be “slow intramolecular H shift” in RO2 radicals, therefore self-reactions of RO₂ become more important.

9. Reduced Mechanism: Reactions in Supplement S4 do not match the reaction routes in Figure 10.

Minor Comments:

Line 86: “the formation of BPRs through H abstraction and O₂ addition”. Should it be “through OH addition and O₂ addition”?

Line 87: “RO₂ + NO” should be “RO₂ + HO₂” in order to form alkyl peroxides.

Figure 1: (1) Citation to Fu et al. should be added to figure caption, also proper citation to Figure 2. (2) B-RONO₂ and B-ROOH were used in Figure 1, but T-RONO₂ and T-ROOH were used in the text.

P1 of Supplement S1: “the correct value” (after Table S1) should be “the corrected value”.

Figure 4b: Two structures shown as diradical are incorrect.

Line 380: “C6H603” should be “C6H6O3”.

General Comments

In this manuscript the authors describe a study in which they compared results of a computational model with experiments conducted on the oxidation of benzene and phenol by OH radicals. The model is an updated version of the Master Chemical Mechanism (MCM) and the experimental results are mass concentrations of secondary organic aerosol (SOA) formed in a laboratory study reported in the literature. The model includes a mechanism that has been expanded to include new gas-phase reactions developed using computational chemistry and literature methods and data, gas-particle and gas-wall partitioning, photolysis, and multiphase chemistry. The authors present a very thorough evaluation of the model by comparing results of the unmodified MCM and multiple versions of the modified mechanisms with each other and with SOA measurements conducted under a range of conditions (NOx, seed particles, etc.). Sensitivity analyses are also performed and a reduced version of the model is compared to the more complete version.

This study is very well done. The modifications to the MCM are extensive and carefully and thoroughly justified. The results are clearly explained and interpreted, and the paper is well-written and well-organized. The study represents a significant improvement to our understanding of the complex chemistry involved in the oxidation of aromatic compounds, a topic which has challenged scientists for decades. I recommend publication in ACP after the following very minor comments are addressed.

Specific Comments

1. How would this model be used in the atmosphere under conditions where RO2 + RO2 radicals are important, since RO2 radicals from other VOC reactions will be present?
2. Figure 7. It appears that the proper treatment of glyoxal partitioning and oligomer formation can have a large effect on predicted SOA formation. What about the role of other oxidation products in oligomer formation? Are multiphase reactions likely to be the major model uncertainty for SOA predictions?
3. I understand that the model results for SOA mass are compared to experiments because that is the best data available. In the future the authors might consider presenting model predictions of the functional group composition of the SOA. I suppose one could calculate this from data in Figure 11 but presenting it for a range of laboratory conditions would be better. This could inspire experimentalists to measure those quantities and would provide a more rigorous evaluation of the model than the SOA mass comparison, which can be good or bad for many different reasons.

Technical Comments

1. Line 133: Should be 11.8%.