Schill et al. present global-scale aircraft measurements of bromine and iodine in accumulation mode non-sea-salt aerosol carried out using the PALMS instrument. It is worth highlighting that measurements of iodine, and specially of bromine in aerosol are scarce and almost exclusively confined to near-surface campaigns. Therefore, the dataset presented in this paper is extremely valuable. The measurements show that iodine and bromine are common in nSSA, although their mass concentrations are low. Analysis of the data indicates two sources of iodine and bromine in the troposphere: a primary source related to biomass burning and a secondary source linked to uptake of gas phase compounds, where ozone is likely to play a role in converting reactive halogens into their aerosol-bound forms, and the halogen mass fraction increases as aerosols age. In the stratosphere, the iodine and bromine mass concentrations are also observed to increase with increasing ozone, but the highest concentrations are found in organic-rich aerosols of tropospheric origin, which leads the authors to conclude that organics play a role in retaining iodine and bromine in nSSA (or are a proxy of some process helping retention). This extensive dataset including measurements from the MBL to the UT-LS for different latitudes and longitudes provides a benchmark against which the performance of global models including halogen aerosol chemistry can be tested. The manuscript presents a comparison of the iodine observations with simulations from the GEOS-Chem model, even though some processes such as the primary biomass source of iodine are not implemented in the model yet. The simulations compare reasonably with the measurements in the background free troposphere, indicating that the secondary source is essentially captured. Elsewhere (MBL, tropospheric biomass burning plumes, stratosphere) the lack of agreement is expected because the relevant processes are not included in the model.

The instrument characterization is sound and the measurements are of high quality. The paper is well written and the conclusions are supported by the data. I recommend publication after minor revisions. I have compiled below a list of comments that the authors may want to consider in order to improve their manuscript.

Minor comments:

Page 2, line 31: Cuevas et al (2018) reported both things first. Please cite Cuevas et al. 2018 also for the enhanced ozone inorganic iodine source alongside Legrand et al. 2018.

Page 2, line 41: a few years after Dean et al. (1963), Duce et al. (JGR, 88, 1983) provided evidence showing that marine aerosols are enriched with iodine largely because they sorb gas-phase iodine.

Page 2, lines 42-43: it would be helpful for the non-expert to provide a brief definition of nSSA, mentioning the relevant nSSA particle size range compared to SSA. This would help to understand better the statements in lines 52-53 in page 3.

Page 3, line 58: a more recent and comprehensive compilation of near-surface measurements of total iodine can be found in Gomez Martin et al. (2021), who retrieved data from campaigns not included in the Saiz-Lopez et al. (2012) review, as well as extended datasets of campaigns included in that review (e.g. the AEROCE data in Arimoto et al. 1995 cited by Saiz-Lopez et al. refers to the period from 30 April 1989 to 29 April 1990, while the full dataset collected by Gomez Martin et al. (2021) goes from 1989 to 1997). Saiz-Lopez et al. (2012) did not discriminate between total iodine(TI) and total soluble iodine (TSI), while Gomez Martin et al. (2021) came up with an expression to convert bulk aerosol TSI into TI using field data where both quantities were measured. The range (min-max) of TI concentrations in bulk aerosol in Gomez Martin et al. (2021) is broader (0.01-530 ng m-3) than reported by Saiz-Lopez et al. 2012. Since this may include outliers, it is perhaps a better idea to provide a range of median values (0.3-42 ng m-3). Nevertheless, since the authors are studying accumulation nSSA, in think the relevant comparison is with TI in fine aerosol. According to Gomez Martín et al. (2022), about 50% of TI is contained in the fine fraction.

Page 3, line 63. For the non-expert, it would be helpful to explain briefly why the study focuses on nSSA and how is nSSA discriminated. The discrimination of nSSA particles is explained in page 5, lines 121-123 but I think this should be explained earlier in the text.

Page 4, line 111. “An inherent assumption of this technique is that the particle type fractions are constant within each bin.” Is this a reasonable assumption?

Page 4 line 110 and elsewhere: mass concentrations are given in ug m-3 here, but in the abstract are given in pmol mol-1. Then again in page 10, line 258, the median and IQR of mass concentration for iodine and bromine are given in ng m-3, and are then converted into pmol mol-1. From this point on, concentrations are given in pmol mol-1. However, this is a mixing ratio "unit" (actually dimensionless), not a mass concentration unit (mass per volume of air). This can be confusing, so I would suggest that the authors define precisely what they mean by mass concentration and how they convert it to mixing ratio, and then to keep consistence in using either concentrations or mixing ratios throughout the text. I would suggest not using concentration and mixing ratio interchangeably and modifying accordingly the axes tittles in a number of figures indicating concentrations in pmol mol-1.

Page 6, lines 151-152: in what physical magnitude is given the MCP output? What unit is C?

Page 8, section 3.1: A comparison with previous aerosol bromine and iodine measurements is absent from the discussion. For example, do the MBL ATom measurements of aerosol iodine show similar geographical trends to near-surface TI measurements?

Page 9, lines 246-249 and Figs A5 and A6: using O3 mixing ratio in ppbv rather than absolute concentration may blur the  link between halogen-containing particle fraction and ozone as a limiting reactant. Within the troposphere the pressure changes significantly, which means that the same mixing ratio at different altitudes corresponds to different concentrations. By contrast, Figs. 7 and 9 are fine because in those halogen mixing ratios are plotted vs trace gas mixing ratio, which effectively is the same as plotting concentration vs concentration.

Page 9, line 248: would it be possible as well that from a certain O3 concentration, heterogeneous O3 reactions on the aerosol surface like iodide + O3 recycle halogens back to the gas phase?

Page 15, figure 7: is pptv here not the same as pmol mol-1?

Page 18, Figure 9: : is ppbv here not the same as nmol mol-1?

Page 18, line 380: “It has also been shown from aerosols collected at the Mace Head research station that over 90% of soluble aerosol iodine is organically bound, with the rest being iodide or iodate (Gilfedder et al., 2008)”. This has been observed only at Mace Head and in a single campaign (MAP 2006), so it should be considered as a special case. It has been noted that long ultrasonication times employed in the iodine extraction process in that work may have influenced speciation. But it is remarkable that the high SOI fractions reported for MAP 2006 are concurrent with extremely high values of total soluble iodine. The ground-based campaign and the associated cruise reported the highest median values of the total iodine aerosol record (44 ng m-3)  (Gomez Martin et al., 2021). So even though this is possibly a special case, the two observations appear to support the authors’ argument that SOI helps retaining iodine in aerosol. Elsewhere on the surface, soluble organic iodine makes up between 40 and 60% of fine aerosol iodine (Gómez Martín et al. 2022).

Page 20, line 416: I would not say that HIO3 formation is a “known” process. The HIO3 molecule has been observed in the atmosphere and appears to be an important and ubiquitous iodine carrier, but its formation is not understood. Finkenzeller et al. (2023) have proposed a theoretical mechanism involving I2O2, O3 and H2O that looks unfeasible on thermochemical grounds. In any case, HIO3 is likely to form somehow from IxOy + H2O, and either HIO3 or IxOy, or both, are lost to aerosol, so it should not make a huge difference as long as an effective uptake process is included in the model.

Page 22, line 455-456: regarding the overestimation of nSSA iodine by a factor of 7, is it possible that the iodine recycling mechanism is incomplete or missing? For example, could gas phase ozone react heterogeneously with aerosol iodide to form HOI and I2?