Please see attached review.

The paper “Characterization of the new BATCH Teflon chamber and on-line

analysis of isomeric multifunctional photooxidation products” by Finja Löher and coauthors

presents in detailed manner their new teflon batch reactor, using toluene oxidation by OH,

produced from photolysis of H2O2 via a sunlight-simulator for near-atmospheric oxidation

conditions, for characterization.

The group uses well-understood and suitable measurement techniques for characterization and

great care has been taken to ensure quantitative detection of major oxidation products and

distinguishing isomers by characterizing the analytical instruments (PTR-MS and SPME-GC-MS)

with authentic standards, added via a permeation tube.

From the described experiments, yields in agreement with literature data were calculated,

supporting that the reactor’s nonidealities (such as wall-losses) is well-understood after

characterizing. This paper is thus laying out the base for future interesting experiment at this teflon

batch reactor.

In view of the thoughtful experimental approach and clear description of the setup, I think that this

paper is well-suited for publication in AMT.

In the results section, some small parts could improve in clarity. I hope that addressing my minor

comments and questions will help improve the final revised publication in these points.

Some larger suggestions

General

I think, this paper and your (and your colleagues’) future work would benefit from stating, which of

the characterizations need to be performed regularly, before every experiment and which data can

be re-used from your publication as-is. I suggest to describe a “best-practice procedure” in this

paper, that you can refer to in the future.

UV part of the bandpass-filtered solar simulator spectrum (section 4.1.1)

As you measure ozone, you could use the photolysis of ozone alone in the chamber to double-

check the scaling of the old UV-spectrum. Currently, the scaling of the old dataset is motivated by

the comparison with the new dataset at higher wavelength, but it is not clearly tested, if the simple

scaling of the old data is correct.

Wall losses (section 4.1.2):

Since your chamber can be temperature-controlled over a large temperature-range, I suggest to

test wall losses at different temperatures (shift of vapor pressure – this would be nice to set into

relation with existing parametrizations), but also at different humidities (effect of hygroscopicity can

play a role here)

Also, since your measured wall losses seem to not level off, I would also suggest to test the

maximal wall loss rate with sulfuric acid.

Please find below some small comments, mainly to improve clarity:

ll. 120- … : Please give a temperature range for which given toluene + OH reaction rate and

abstraction/addition channel ratio are valid

Table S1:

table S1 also needs the name of the species (at least in the table description), not just the short

form, as these short forms were not introduced yet. Not in the main text until this point, nor in the SI

l. 230: here I wondered, how long is one experiment roughly? Information on this comes much later

in the manuscript, but would be interesting already at a much earlier stage, e.g. here.

l. 265: “SIM” has not been explained yetll. 339/340:

does that mean, the PTR-MS was calibrated only once, due to the complication of focusing on the

oxygenated species?

While it makes a lot of sense to calibrate these species individually, I suggest to use gas-standards

with e.g. a set of hydrocarbons, ketones, siloxanes (…) to allow for more frequent calibrations to

monitor variations in its transmission curve.

l. 370: many teflon chambers are operated continuously under pressure to avoid contamination

from outside. Did you determine the effectiveness of cleaning your chamber both with just flushing

under pressure vs with evacuation? (same: L 400 → emptying it to 30% of its original volume). I

wonder, what the impact of small leaks would be here. Did you perform leak tests, e.g. by spraying

acetone outside the teflon chamber and monitoring it inside?

L 379 ff. – not completely clear to me. Is the toluene flushed with the N2 (15mins, 5slpm) into the

larger teflon batch chamber or was the N2 flushing used to clean the toluene flask before toluene

was injected? In the latter case, how is the toluene entering the batch chamber?

fig. 3 – I suggest to add long names with the short forms in brackets in the figure description

l. 480 – You mention that the reactions with NO3 and O3 are negligible due to their small reaction

rates. This is slightly confusing, as I thought the experiments were Nox-free and you did not

mention that you actively added O3, so are the experiments not O3 and NO3-free?

Fig. 4: the legend could be improved for more clarity. There are e.g. multiple grey lines in different

shades of gray, but only 1 line in the legend. Alternatively, decreasing the visibility of the error-

minima and maxima traces and just using shading might help

l. 523: how did you find the OH concentration? Was it measured via the toluene decay?

L. 525: that’s good! But I believe, this needs to be checked before future experiments after NOx

has been added, again.

Fig. 7: “relative response” is not clear from the figure or figure description (only after reading the

text). It is not described, in relation to what the signals are shown. Please add units for more clarity.

Fig. 8: please add the shortforms (e.g. PHE-d6) in the figure caption

Fig. 11: instead of “final concentration”, maybe call it total produced concentration or similar?

And mark it in the plot as a horizontal bar with uncertainties (cause the loss rates also have

uncertainties)

Fig. 12: please just write mixing ratio instead of the new shortform “MR”

fig. S6.4: is m your calibration factor? Y-axis unit missing…