General comments

This paper by Avery et al. attempts to understand the contribution that pyrolysis makes to the emissions from biomass burning in the form of aerosol production form different wood types. This is important because pyrolysis underpins the processes of combustion and understanding the yield and composition of aerosols from is important for assessing impacts on climate and human health.   This paper examines typical fragments in mass spectra generated by an aerosol mass spectrometer to detail the patterns in terms of yields and emission factors, and tries to explain the processes occurring that make up the aerosols that are measured.  It builds upon the experimental work conducted by Fawaz et al. (2021) who characterise the pyrolysis chamber used in the experiment to answer questions as those posed in this paper.

Although the paper attempts to address relevant questions, and produces some interesting and new insights, I find its conclusions weak based up important information missing from the manuscript.  Most importantly, is the number of replicate runs done in this study.  In lines, 169-171 it mentions running one wood type in duplicate only.  There is no mention as to how many replicates of the other two woods were examined at the different temperatures and sizes.  Indeed, in Table 1, which the authors state in L179 is a full list of experimental conditions, the duplicate results are shown for the one wood type but the other woods and conditions are only shown as one replicate.  Also, several of the graphs only show one line for some of the experimental conditions but two for the material that was stated to be run in duplicate.  If only one replicate was run, then the lack of replication in the experiment does need not give me confidence in the findings of this study and undermines statements made by the authors about high reproducibility.  

Specific comments

L71-76:  This is your aim of the study.  I suggest moving this to the end of the introduction

L101-103:  What is the relevance of mentioning PMF?  It’s not a technique you use in this study.

L104-109:  This reads as justification for methodology.  I’d suggest moving this to the methods section.

L127-128:  What is the relevance of this statement?  You don’t do modelling in this paper.

L134-137:  Like L71-76, this is part of your aims of the study.  I’d suggest moving it to the end of the introduction.

L151-152: Missing reference for the pyrolysis temperatures of cellulose and hemi-cellulose.

Section 2.1: The title for this section needs to include material characterisation.  All, I would expect some basic characterisation of the wood to have been presented.  For example, density, proximate or ultimate analysis, cellulose, hemi-cellulose, and lignin contents.  The compositional data would be important for explaining f60 and f73 results, for example.  

L167-168: The Genus (in Genus species) should start with a capital letter

L176-179: You talk about changing the dilution ratio but don’t describe how (either here or in the supplemental material). This is important to know in order for this experiment to be replicated.

L191:  Who is the developer of the Squirrel and Pika?  Please insert details.  

Table 1:  Description of abbreviations is not given in the legend.  Also, the asterisk seems to have been used for two different purposes.  These needs changing for one use and the other use described with the appropriate symbol.  Also, what do the plus/minus values represent?  Standard deviations, ranges?

L274:  Reproducibility should be repeatability – you are using the same equipment, materials etc., all within the same timeframe.  However, what metric are you using to demonstrate this?  You don’t present one.

L278-279:  How did you calculate your uncertainties?  A comment in the methods section is needed to explain this.

L321 – 323: You need a reference to the FIREX campaign.

L443 – 446:  How did you make this connection?  Having characteristics of the wood would substantiate this comment.

L520 – 522:  What is the implication of this finding?

L523 – 528: If this is important to show why is the graph in the supplementary materials?

L553 – 556:  What you are suggesting is coupling the AMS to thermogravimetric analysis (TGA) which is a common technique for pyrolysis studies.  There are an abundance of studies on TGA and gas analysis which you could cite, but not on aerosols and their composition.

Figure S3: There is no explanation in the legend of the dashed line at 0.3% being the background f60 as described by Cubison et al. (2011).  Nor is there an explanation that the solid lines are the expected relevant ratios for the atmosphere.

Technical comments:

All tables:  Does not match the formatting requirements for the Journal i.e your use of vertical lines.

All tables and graph:  What are the number of independent samples?

The paper titled “Chemically distinct particle-phase emissions from highly controlled pyrolysis of three wood types”, by Avery and Co-Authors presents the results of 18 experiments in which the authors investigate the chemical composition of aerosols generated during the pyrolysis of wood typical of the western US. The authors investigate 3 variables: wood type, wood size, and temperature. The results are corroborated by the comparison with the chemical composition of aerosolized cellulose carried out in the lab. One of the major strengths of the results presented here is that the experimental setup guarantees that the aerosols are generated by pyrolysis only, with no combustion happening, thanks to the use of a heated nitrogen flow. The study is of interest to the scientific community and addresses relevant scientific questions, such as the chemical composition of aerosol generated during pyrolysis. This topic is well within the scope of ACP.

The main weaknesses of the paper are the lack of a schematic picture of the pyrolysis reactor, the relatively small number of repeated experiments, the lack of explanation of the uncertainties, and the lack of measured dilution for the maple samples.

The data presented are novel and the conclusions reached are substantial and will be of use for the interpretation of future data (e.g., the oxidation state of emitted OA that remains largely constant over the course of a pyrolysis experiment, and that CO+/CO2+ greater than five is identified as a marker for pyrolysis). The methods are valid, in particular, the repeatability of the temporal profiles is remarkable.

Title: I recommend either adding a hyphen to “particle-phase” or just using “particles”

40 biofuel burning, I recommend adding a reference

84 “Previous work has shown …”, citation needed

135 please add which different conditions are explored

160 Section 2.1, I recommend adding a schematic picture of the reactor it would help the reader follow the text without having to dig up the previous paper cited.

166 please add that the wood was cut along the grain lengthwise as it’s important later on in the paper

173 m3/s correct the format for the units

175 "The aerosol sampling line was further diluted" Please add the range of dilutions used

187 "The size distribution of particles was well within the AMS standard lens size range" I recommend corroborating this statement with a picture in the SI

278 "The uncertainty for the ER ..." add here or in the method section a line/citation on how they are calculated 

288 "The uncertainty for the EI ..." add here or in the method section a line/citation on how they are calculated