This paper presents an examination of the temperature-dependent conversion of multi-functional nitrates from isoprene chemistry in TD-CRDS measurement systems. I applaud this group for their continued careful and critical examination of this technique that has been used for many field campaigns by a number of research groups, most of whom have stopped doing this kind of homework long ago. This paper is a very useful follow-on to previous work by this group and should be published pending the authors response to the following general and specific comments.

General Comments

            There is no C-N bond in an organic nitrate molecule, nor in a peroxynitrate molecule. This appears in several places (lines 300, 313, 371) and is incorrect. It seems the bonds the authors are referring to is the RO–NO₂ bond, which is the weakest bond in the molecule (160-175 kcal/mole), and the RO₂–NO₂, which range from ~88 to 117 kcal/mole.

            What does nitryl chloride (ClNO₂) do in your inlets?

Specific Comments

Abstract: Line 15-16. Why not just include a brief description of the solution to the problem, similar to the sentence on lines 540-542?

Line 30. I disagree, NO₃ initiates oxidation of only comparatively few VOCs in the nighttime troposphere, e.g. alkenes, DMS.

Line 43. Isoprene is not the most abundant NMVOC in the atmosphere. Isoprene has the highest total source to the atmosphere, but that is a different thing. The much less reactive small alkanes and some oxygenates are more abundant than isoprene in any but the most biogenically impacted environments in the troposphere.

Line 106. It feels like the phrase ‘to acquire complete mixing…” is a bit awkward.

Line 197. Define MPAN somewhere.

Line 231. What were the humidities and NO₂ concentrations used in these tests?

Line 269. Previously, you said the experiments were conducted by allowing the temperatures to decrease from high to low.

Line 277. More efficient compared to what?

Line 295. Monotonic increase with what? Temperature?

Line 311-312. This seems to be a partial sentence, i.e. something is missing here.

Line 315. The addition of the word unambiguously implies that there might be RO₂NO₂ compounds in some isoprene-NO₃ chamber studies, is that true?

Lines 324-325. I assume these are concentrations at the exit of the SCHARK? Otherwise, how does 40.5pptv of N₂O₅+NO₃ make several ppbv of organic nitrates?

Lines 330-335. How long did you wait for the signal to stabilize at each temperature step? How well did it stabilize?

Line 428. The word “exact” is not appropriate here.

Line 469. What is meant by “exchange rate”? Don’t you really mean residence time?

Line 475. Can you give an estimate for the O-O bond in typical peroxides?

Line 523-524. You have only circumstantial evidence that the molecule has a double bond or a peroxide group. It could be one or the other since you have only the complex mixture formed in isoprene-NO₃ chemistry to go by, you haven’t tested each separately.

Figure 3a. I have a hard time distinguishing the open and closed squares, especially in the 600-650K region of the plot.

Dewald et al. examined the conversion of isoprene derived nitrates to NO2 in heated inlets. The isoprene nitrates were generated from reaction of the nitrate radical with isoprene in a large environmental chamber . Three inlet designs were evaluated: a conventional heated quartz tube, common in such instruments, a quartz inlet containing glass beads, and a heated inlet constructed from PFA Teflon. Measurement artifacts arising from the thermal decomposition of ozone (to atomic oxygen) and of nitric and nitrous acid (to hydroxyl radical) were evaluated.

This is a well-written and thorough manuscript. The paper will be of interest to the growing community of TD-CRDS and TD-CEAS users and within the scope of AMT. I recommend publications once the authors have considered my comments below.

General comment

The abstract boldly promises a viable solution to broadened thermograms observed in the measurement of isoprene nitrates by TD-CRDS but in the end falls somewhat short of this goal. The proposed solution, a better-performing PFA inlet, is not a viable alternative to current inlet designs since it cannot be heated above ~500 K without melting. Perhaps more discussion is needed to better articulate what specific problem this paper ultimately has addressed.

Specific comments

line 1/title: Replace "detection" with "quantification".

lines 15/16. An abstract for a scientific paper should be sufficiently representative of the paper if read as a standalone document. In this spirit, please name the "viable solution to this problem" rather than teasing the reader at this point.

line 108. "by passing a fraction of the air" Is UV transparent material such as quartz used here? Please provide more experimental details.

lines 141/143. The temperatures required for full conversion of PN or AN depend on residence time and on where the temperatures are measured and thus are not universal. Please note the dependence of these conversion temperatures on inlet residence time(s) here (they are given on lines 176-177) and add a qualifier such as "In our inlets, ..."

line 142, 143 "results in conversion" and "additional conversion". Should this say: complete conversion?

line 162-177. Please comment on (any) pressure drops associated with placing glass beads in the inlet.

line 166. Wouldn't the glass bead also lead to more uniform heating of the sampled gas and thus aid in the dissociation of PN and AN?

line 180/Figure 2. Please state the temperature of the SCHARK chamber. Assuming it is 298K, one would expect with ~4 ppbv of NO2 an equilibrium N2O5:NO3 ratio of ~3:1, as was indeed observed at 11:45. However, the observed ratios at the earlier times (e.g., at 10:45) seem lower than expected from equilibrium. Under the conditions described here, the time to achieve equilibrium should be sufficiently short (minutes). Please comment as to why the N2O5:NO3 deviates initially.

line 186 "after subtraction of the N2O5 mixing ratios" - please indicate that you are subtracting N2O5 here in the figure legend and caption and note that you are also subtracting NO2 (I am guessing).

line 187 The residual signal is curious. Is it possible NO2 is lost in the low-temperature reference channel to NO2+O3 and N2O5 formation but not in the heated channel?

line 194-195. Since the amount of isoprene added (i.e., its concentration) is known (line 202), please calculate the yield of AN relative to integrated amount of NO3+Isoprene.

line 200 "Very low concentrations" Please be quantitative.

line 201. Consider adding a reaction scheme for clarity.

line 205-206. "This reaction path is a minor one" How do you know this?

line 258. Please describe the Iodide-CIMS and how it "was coupled to the experiment" in section 2 "Experimental".

Line 267. How was the mixing ratio of 22 ppbv HNO3 determined?

Line 279 "by the scavenging of O-atoms". While this is plausible explanation, please note the speculative nature of this statement. Rather than scavenging O atoms, the surface could act to catalyze 2O->O2, for instance.

line 288. "H2O drives HNO3 from the surface and thus protects it from surface reactions". How much water would you expect to be sorbed to surfaces in an inlet heated well above the boiling point of water? 

line 350-351. The instrument must have sampled a very large concentration of nitric acid for a long time to produce such a large artifact.

line 400. Replace Torr with SI units, please.

line 423. Please move the experimental details to the experimental section.