The heterogeneous conversion of SO₂ to sulfates on non-photoactive surfaces was well investigated in this paper. The authors reported that light can enhance the SO₂ uptake and sulfate formation on non-photoactive surfaces (SiO₂, Al₂O₃, kaolinite, CaO and MgO). The sulfate formation pathway involving in the participation of O₂ and H₂O was proposed. This is a novel topic since the previous studies generally focused on the sulfate formation on photoactive surfaces, such as TiO₂, Fe₂O₃ and etc.. This study highlighted a new pathway that contributed to the source of atmospheric sulfates. The paper was organized with logic, and the conclusion is convincing. I would recommend this manuscript for publication in ACP after a minor revision.

Specific Comments

Lines 38-39: The two sentences “sulfates, which is one of the most significant compositions in fine particles.” and “Sulfates can contribute greatly to the mass concentration of PM_2.5” have the same meaning. I suggest simplifying the second sentence into “the mass of sulfates in PM_2.5 is high up to 30%”.

Line 77: “and” should be modified into “with”.

Line 97: The details for measuring the light absorption spectra of samples should be given.

Line 104: “ml” should be modified into “mL”.

Lines 109-110: Rectangular flow reactor was not a conventional reactor. Was this reactor used in the previous paper? Or the feasibility of this reactor was verified before?

Lines 131, 225, 228, 279, 318 and 397: The unit of light Intensity (photons cm^-2 s^-1 or W m^-2) should be unified.

Line 167: The unit of BET surface area of SiO₂ should be m² g^-1.

The sentences in Lines 171-174 can be simplified into “The changes in the chemical compositions on mineral oxides in the SO₂ uptake process were investigated by the Fourier transform infrared (FTIR) spectrometer (Thermo Nicolet iS50) equipped with an in situ diffuse reflectance accessory and a mercury cadmium telluride (MCT) detector”.

Figure 1A: The physical adsorption of SO₂ on SiO₂ can be quantified according to the integral at the end of the reaction (t= 80-100 min).

Lines 236-237: The bands at 1074 and 1038 cm^-1 were both ascribed to sulfite. Why did only 1038 cm^-1 band appear in the dark condition (Figure 2A) and only 1074 cm^-1 band appeared in the light condition (Figure 2B)?

Figure 2: In the light condition, the band at 1038 cm^-1 decreased, while this band increased in the dark condition. Please explain this phenomenon.

Figure S7 and S9: I didn’t observe any new peaks at 974 cm^-1.

Lines 290-291: The band at 1074 cm^-1 should be marked in Figure S9.

Figure 5: The assignment of 1300 cm^-1 should be given.

Figure 5: Less sulfites were formed on kaolinite and Al₂O₃, while abundant sulfites were observed on MgO and CaO. Please explain this phenomenon.

Unexpectedly photoenhanced sulfates formation by the heterogeneous uptake of SO2 on non-photoactive mineral dust by Han et al.

Summary

This manuscript investigated heterogeneous photochemical reactions of SO2 on five mineral oxides (mainly SiO2, but also kaolinite, Al2O3, MgO, and CaO) that have no photocatalytic activity, and they found light can significantly enhanced the uptake of SO2 which converts to sulfate. Light intensity, RH, O2 contents and basicity of mineral oxides play key roles in this interfacial chemistry, especially regulates SO2 uptake coefficient. The experiments were performed under various conditions, i.e. using flowtube reactor to obtain the SO2 uptake kinetics, and DRIFTS measurements to confirm sulfate formation.

The technical part seems sound, and I enjoyed reading the manuscript as it is quite easy to follow. Overall, the whole paper is displayed in good quality, with clear writing and nice figures. I would recommend this manuscript to be published in ACP after considering the following major concerns if these comments are helpful for improving the manuscript.

Major concerns

(1) Can any possible contamination or impurity in these non-photoactive mineral dust samples be ruled out in this study?  The absorption spectra in Fig. S1 seem clean, but tiny amounts of photoactive components if existed as impurity could make it very different. This is a main worry from my side, in case other people cannot repeat the results.

(2) I am not sure whether this study is the first time to look at SO2 photochemical uptake on non-photoactive mineral surface. Is there any SO2 photochemistry reported on non-photoactive mineral oxides in previous literatures? Does it really unexpected as it seems occur on any surfaces that concluded by the authors, even onto flowtube wall in the blank experiment as shown in Fig. S4-S5?

(3) Line 52: Some recent findings on multiphase SO2 oxidation leading to sulfate formation should also be mentioned here. Please read: Liu T, Chan A W H, Abbatt J P D. Multiphase oxidation of sulfur dioxide in aerosol particles: implications for sulfate formation in polluted environments[J]. Environmental Science & Technology, 2021, 55(8): 4227-4242.

(4) Line 60-62: “Thus, investigating the heterogeneous oxidation of SO2 on mineral dust is of fundamental importance to reveal large missing sources of atmospheric sulfates in the haze periods.” I feel this is likely overstated which may not objectively reflect current understanding

(5) Line 97-99: This UV-Vis measurement (300-800 nm) does not match the results shown in Fig. S1

(6) Fig. S4-S5 shows a blank example for SO2 loss onto the flowtube wall at a specific condition with irradiation. The SO2 uptake coefficient is actually measured following a blank subtraction. Does this blank change at different conditions, i.e. different O2, RH, light intensity?

(7) Line 193-194: Here it says the measurements on non-photoactive mineral oxides are comparable (10⁻⁷ − 10⁻⁶) with those previously reported in literatures, especially dust containing photocatalytic components. Does this mean both photoactive and non-photoactive mineral oxides showing equal/comparable ability of SO2 photochemical uptake, and those photocatalytic components (such as TiO2, GDD, ATD) do not actually play much role ?

(8) Line 295-302: The dependence of γ on five different minerals is very interesting, and explained by their pH differences.  Did the authors check such pH-dependence for the same type of mineral oxides (i.e. SiO2) to really prove this pH effect, i.e. via experimentally adjusting the pH such as adding NaOH ?

(9) Fig. 4 and Fig. S10: The photo-enhanced SO2 uptake is not very significant for other three minerals, especially CaO. This suggests that the enhanced SO2 photochemical uptake at higher pH (more basic mineral oxides) is actually attributed to SO2 dark uptake, which is a bit contradict with the pH explanation. Why SO2 dark uptake is so strong under these basic mineral surfaces? Fig. 5 also shows a lot of interesting results but not yet discussed in details. I would suggest the authors to stress the SO2 dark uptake on some basic minerals as an important process, with more detailed discussion here.

(10) Line 347-358: I like these DRIFTS experiments designed by adding Ru(bpy)3(Cl)2) or NaHCO3. How are these added ? Are these ³SO2 or OH scavengers also performed in the flowtube reactor to check the SO2 photochemical uptake, which should be unchanged in the presence of these scavengers ?

(11) Line 368-369: Did you test SO2 uptake coefficient under visible light i.e adding an optical filter at 400nm? Does visible light (>400 nm) also contribute this photoenhanced SO2 uptake?

(12) Line 370-372 “It means that any surfaces, providing absorptive sites for SO2, can significantly enhance the photooxidation of SO2 to sulfates.” This could be true, but I am afraid it is not very strong yet, especially the current experiments on some basic minerals indicate SO2 dark uptake is more important under these conditions.

(13) Line 386: The lifetime of SO2 photochemical loss on minerals was calculated and compared with those from literatures. Are these conditions comparable? Otherwise should be very careful

(14) Line 393-416 & Table S1: I have greatest concerns about the last section on atmospheric implications. The importance of this SO2 photochemical chemistry on sulfate budget is not yet strictly evaluated, which needs to be done under a uniform model framework. The current calculation on sulfate production rates and comparison among these mechanisms are still very speculative, based on my opinion. Thus, it should be not extrapolated too much. I would suggest to minimize these text and reservedly conclude that this is a newly identified sulfate formation pathway that might occur in some dust-rich conditions.

Minor comments:

Line 39: change to “… with the mass fraction of sulfates … ”

Line 114: How did the Reynolds number (Re) being calculated ?

Line 142: Can you provide the detailed numbers (i.e., V, S, w, D, N_u, etc) you used for equation (2), (3) and (4) calculations ?

Line 151: “The corrected γ can be calculated by inserting the equation 3 into the equation 2”. I am a bit confused here. My understanding is that equation (3) is to give a corrected k, then it needs a separate equation to calculate corrected γ

Line 186: didn’t = did not

The light intensity in many places are presented i.e. 250W/m2 or xxx photons cm⁻² s⁻¹. I am not sure they are the same