General Comments

In this manuscript the authors present a new method for creating NO3 radicals for atmospheric chemistry experiments by photolysis of aerated cerium ammonium nitrate solutions. The photolysis apparatus is described and the results of an evaluation of the effects of different experimental parameters on the modeled concentrations of species in solution and species in the gas phase based on mass spectrometer measurements is presented. The apparatus is also used to generate NO3 radicals for a reaction with beta-pinene in an oxidation flow reactor (OFR) for comparison of mass spectra of gas- and particle-phase products with the literature.

The experiments and modeling were well done, and the results are thoroughly discussed and interpreted with an excellent use of information on this solution photochemistry obtained from the literature. This method has some advantages over those currently used to create NO3 radicals and so may see considerable use by the atmospheric chemistry community. I think the manuscript is appropriate for publication after the following minor comments are addressed.

Specific Comments

1. Line 329–332: The IC10H16N2O7– ion is a hydroxy dinitrate not a dinitrate, so it could not have been formed directly from an RO2• + NO3 –> RONO2 + O2 reaction of the beta-nitrooxyperoxy radical. It would require a nitrooxyhydroxyperoxy radical, for which I am not aware of a formation pathway. Also, I did not find anywhere in the Orlando and Tyndall (2012) paper that suggests that nitrates can be formed by a RO2• + NO3 reaction, nor do any others I am aware of. The only reaction they describe is RO2• + NO3 –> RO• + NO2 + O2. I suggest you just say you don’t know how this product forms.
2. Line 367–369: Competing NO3 and O3 reactions are only a problem if one tries to synthesize NO3 radicals online. It is not an issue when one synthesizes N2O5 and then stores it in a freezer until needed. The synthesis is simple and one can easily make enough in a couple hours to last for months or years.
3. Since most people using this method would be interested in knowing what RO2• reaction regime they are in, I suggest supplying a more detailed discussion of how the different synthesis conditions affect the relative concentrations of NO3, NO2, and HO2 radicals, and noting that the rate constant for RO2• + HO2 is about 10x greater than for RO2 + NO3 or NO2. It would be especially useful to say how to run the source if one wants to be in a RO2• reaction regime dominated by reactions with HO2, NO2, or NO3.
4. How rapidly do gas-phase products collide with the walls in an OFR? If ROONO2 products formed from RO2• + NO2 –> ROONO2 are significant with this method then the collisions will likely be a RO2• radical sink and a source of R=O products via loss of HNO3 from ROONO2.

Technical Comments

1. None

I reviewed this paper to fill in for missing second review to avoid further delays in the open discussion process.

This is well written manuscript that proposes a new way of producing flows containing NO3 for atmospheric experiments. The system is based on (complex) photochemistry of Ce(IV) nitrate, and the bulk of the manuscript describes tests in which concentrations, irradiation wavelengths, and light fluxes are varied to find the optimum setting for making NO3. The system is then tested by making SOA from NO3 produced by Ce(IV) nitrate and by conventional N2O5 thermal decomposition. I only have minor comments

CONTENT

I think section 3.5, especially the discussion of possible N2O6 formation, may distract the readers from the main message of the manuscript. This discussion is more pertinent to ionization chemistry in I- CIMS than to the topic of characterizing the NO3 source. I would suggest shortening this section (or maybe even removing it and developing it into a stand-alone paper). But keeping it there is OK also, as I- CIMS if fairly common, and the discussion will be user to I- CIMS users who work with this NO3 source.

EDITORIAL

L82: I would mention that flux at 421 nm was not quantified

L85, L170, etc.: I would suggest replacing “I-values” with a more explicit name

L94: Table S2 is referred to before Table S1 is mentioned on line 118. Probably best to fix the order.

L102: do you have an estimate of the effect of RO2 reactions, perhaps from OFR modelling?

Figure 3: instead of saying “Additional figure notes” you can say “The black dot corresponds to data from Wine et al. (1988)”

L216 and Figure 5: I presume the data are for 369 nm only. I would mention it here and in the figure caption.

Figure 7 caption: a NO2 -> (a) NO2

Figure 7: would it make sense to also include NO3 mixing ratio in this figure measured under the same conditions, similar to the one in Figure 2?

L330: was the order -> was of the order

Figure S2: captions mentions acetonitrile but it is not clear what role in plays in the data as normalization is done with respect to thiophene. Was everything first normalized to acetonitrile as the text states? I would mention this in the caption more explicitly.

Table S1: It appears the authors list absorption cross section (or perhaps a product of that with the quantum yield) instead of the photolysis rate constants for photolysis processes. I would note this to avoid confusion.

Table S1: when multiple values for the same process are listed, such as the Martin and Stevens (1978) values for Ce(III) + NO3 reaction, which one is being used in the model?