This manuscript investigated the effect of HCHO on the photocatalytic renoxification of nitrate on TiO₂ particles. The investigated system is interesting. However, the experimental design has so many defects. More experiments and verification are needed to support the conclusion. 

Major comments:

1. Methods: Line 103: 400 L chamber is usually not enough for the investigation of heterogeneous reactions. Besides, only 250 L air was injected into it, which would increase the wall effect of chamber. Line 105-106: How did the author control the chamber temperature? It is well known that the chamber temperature will increase when turn on the lights. Line 111-112: the light intensities for the tube and LED lamp were different, so how to compare their results? Why was only the results in 3.1 obtained under the irradiation of tube lamps? What was the meaning for introducing two kinds of lamps in the smog chamber?
2. Important defect of this article is the composition of the mixture in the part of “2.2 nitrate-TiO₂ composite samples”: “TiO₂ was simply mixed in nitrate solutions at the desired mass mixing ratio to obtain a mash. The mash was dried at 90 ^oC and then ground carefully to ensure a uniform composite of particles.” How did the author ensure that the particles are uniform composite of nitrate and TiO₂? Did the author do some experiments to confirm these? For example, in the reference of Ma et al (EST, 8604-8612, 2021), the nitrate and TiO₂ mixture was dripped onto a quartz tube inner all, then images and Raman spectra of single composition and mixture were analyzed, and mixture were confirmed to form. However, in this work, the generation method of mixture particles is different from that of Ma’s work, and these mixture particles are sprayed by synthetic air into PVF bag. No experiments have been given to confirm the composition of the mixture particles in the chamber. In my opinion, this method can’t generate a uniform composite of nitrate and TiO₂!!! The composition and the nitrate content are the most important quantitative method factors of all the experiments in the article. If the composition and nitrate content can’t be control, how to compare the NO_x concentration in different experiments? Then, all the results are not convincing!!!
3. Another important defect of this article is the quantitative method of NO_x concentration. As shown in Ma’s work, they used the normalized concentration (ppb/mg) to quantify NO_x. However, this work just used the NO_x concentration (ppb) to compare different experiments, which meant that if more reactants were added in the chamber, the generated NO_x concentration would be higher. The initial mass concentration of particles was 300 mg/m³ (75mg/250L), and the concentration of HCHO was 10 ppm, which were much higher than that in the real environment and resulted in that the obtained results could not be directly used for an analogy to real environment. The results with ppb as unit are meaningless to reflect their influence in the real atmosphere. Were the particles kept the same in different experiments during the reaction? The author mentioned that the wall loss of particle in the smog chamber was very high at the beginning. And the wall loss for different kinds of particles and for the same kind of particles in different experiment (maybe affected by the conditions of the smog chamber wall) should be different. How did the author ensure that the particle distributions were the same in different experiments when turned on the light? Besides, the surface area, as an important factor in heterogeneous reactions, has not been detected in the experiments. Different surface areas directly affect the irradiation surface of TiO₂, the uptake of HCHO and the release of NO_x. The missing information of surface area would result in the large uncertainties in the experiments. At least, the authors should give a normalized NO_x concentration, then different experiments can compare with each other and give the reasonable results and reflect the influence in the real environment.
4. Gas HCHO and mixture particles of TiO₂ and nitrate were contained in the system. Although some controlled experiments were conducted, the role of TiO₂ and HCHO still could not be isolated. A series of important experiments such as HCHO and single nitrate particles under irradiations are needed.
5. All the proposed mechanisms couldn’t be well supported only by the changes of NO_x concentration. This work and Ma’s work indicate HONO, HNO₃, NO₃ radical, NO_x could form in these reaction systems. However, HONO, HNO₃, NO₃ radical could lead the overestimation of NO₂ concentrations by chemiluminescence method. How did the authors exclude the effect of these species? Besides, most important products such as NO₃, HNO₃, HONO were not detected in the experiments except OH radical. How did the authors make sure that the reaction pathway followed the proposed mechanisms? It is well known that TiO₂ can photocatalysis HCHO, can this reaction affect the formation of NO_x?
6. The mixture of HNO₃ and TiO₂ was used to support that HNO₃ was an important intermediate to form NO_x. However, this logic is not right. If it is right, then any N-contained components mixed with TiO₂ that enhanced the generation of NO_x could be thought as the intermediates of NO_x formation. The direct way to identify the intermediates is to measure them such as FTIR/DRIFTS to measure the adsorption products.

Minor comments:

1. Abstract: many sentences are confusing me! I can’t understand what the main meaning of the work. What’s the main results. The languages need to be improved.
2. “photocatalysis”, “photolysis”, “photocatalytic”, “photochemical” appeared in the manuscript everywhere, the author should make sure the exact meaning of these words and give the right usage of these words.
3. Line 232-233, the photodegradation of HCHO on TiO₂ is not zero-order reaction kinetics, the curve is not a line as shown in Figure S6, which decreased slowly and then fast. The reason for it should be the large amount of adsorption of HCHO on the particle during the long-time injection of HCHO. Besides, the continuous wall loss of particle would result in the change of kinetic coefficient. The concentration of particles and HCHO were too high, and the injection time was too long to give clear kinetic parameters. Generally, the photocatalytic process is supposed to be a first order reaction.
4. I can’t understand why the authors used KNO₃ and HNO₃ to mixture with TiO₂. In Ma’s work, they indicated the NO_x concentration formed from KNO₃ was the lowest. KNO₃ only accounts for small proportion in the atmospheric particles. HNO₃ is acid species and can react with TiO₂, which would result in the component changes in this mixture particles. I think that the components in this mixture particles were different from the discussion in the article.
5. OH radical was measured by ESR in this study. However, the role of OH radical has not been discussed. And the OH radical generated in different particles and under different conditions have not been compared and analyzed. Besides, NO₃ radical was proposed to be the important intermediates in the reaction. Why did not the authors measure NO₃ radical?
6. Weight percentage was used to quantify nitrate in the mixed particle. However, different nitrate has different molecule weight, which would result in that the molar concentrations of different nitrates with the same weight percentage were different. For example, the molar concentration of N in 4 wt % HNO₃-TiO₂ is higher than that of N in 4 wt % KNO₃-TiO₂. This effect should be considered in the formation of NO_x.

Liu et al. investigated the possible renoxification processes occurring on TiO2 particles and mineral dust particles in presence of adsorbed nitrate, or HNO₃, in presence of HCHO. They suggest that HCHO and TiO2 have a significant synergistic effect on the photocatalytic renoxification via a NO₃-NO₃-HCHO-HNO₃-NO_x pathway, in which adsorbed HCHO may react with nitrate radicals through hydrogen abstraction to form HNO3 on the surface, resulting an enhanced generation of NO_x.

Overall this is an important topic, which certainly falls within the scope of journal Atmospheric Physics and Chemistry.

The experiments presented here were performed in a simulation chamber consisting of a 400 L polyvinyl fluoride (PVF) bag filled with synthetic air. In such a small baga, the life time of particles is expected to be very short, as shown also in figure S2. Surprisingly, the authors do observe, after some induction time, a stable size distribution over hours. How can this happen? Is it an indication of some dynamic interactions with the chamber’s walls, during which particle adsorb and desorb constantly? Such process may be induced to some air turbulences around the bag, or through its deflation during the experiments (by the way, was the bag closed and its volume shrinking or was it flushed by pure air all the time during the experiments?). Anyhow, this is a strong indication that wall effects may play a significant role in the reported experiments. Therefore, a thorough discussion of these effects has to be included in the manuscript.

The main conclusion of this work is that adsorbed HCHO reacts with adsorbed nitrate radicals, promoting NOx formation. This assumes that this reaction is faster than the one of HCHO with the photochemically generated holes on the surface of the mineral. Is this justified by any means? HCHO being efficiently degraded on illuminated TiO₂, one would expect that this VOCs may compete with the nitrate anions to react with the holes, with the synergy between nitrate anions and HCHO vanishing at low surface coverage (where both compounds would react with the holes with no interactions with co-adsorbed species). Is this observed here?

Spraying mixture of SiO₂ and TiO₂, would result in an externally mixed aerosol, isn’t it? Then it should represent an experiment with the TiO₂ particles simply being diluted as compared to the pure TiO₂ experiment.

An effect of acidity is observed and explained by the enhanced photolysis of HNO₃. Could an alternative explanation arise for the chemistry of O₂^-? This superoxide would react with H⁺ inducing HO₂ chemistry that may change a series of surface reactions. Could the authors comment on that?

The author reported formaldehyde may have synergistic effect in photocatalytic renoxification of nitrate with TiO2, in order to explain the difference between field data and modeling result. The article focuses on the significant synergistic effect, i.e., HCHO and TiO2 have on photocatalytic reactions and providing one possible reaction pathway- NO3−-NO3·-HCHO-HNO3-NOx. These findings improve the understanding of the role of reactions between organic components and nitrate in the chemical and physical properties of aerosol particles in low relative humidity region. It has significant implication in the research of atmosphere and air pollution, but some issues in the article must be improved. I recommend accept this article after resolve those issues. There are the comments I have for this work:

1. There are some misdescriptions in the manuscript. Like:

We suggested that the produced NO3· contributed to the enhanced uptake of HCHO. Therefore, we suggest that NO3· production contributed to enhanced HCHO uptake. (Line 236, Page 12)

photochemical cycle of HOx radicals in the atmosphere and the formation of (Line 448, Page 25)

2. Please explain why 4 wt.% KNO3-TiO2(1 wt.%)/SiO2 underwent reaction to release NOx, while 4 wt.% KNO3-SiO2 and 4 wt.% KNO3-TiO2 not in same condition.

3. In Figure S3, the concentration of HCHO and TiO2 particles reached stable in 60 min after introduced into experimental chamber, but other experiments almost stared in -30min, did HCHO and TiO2 have been stable?

4. The BET of TiO2 nanoparticles is huge, the uptake of HCHO in TiO2 nano-particles can’t be ignored. The photodegradation of HCHO on TiO2 and 4 wt.% KNO3-TiO2 particles should start after adsorption and desorption balance.

5. In section 3.3.1, 4 wt.% KNO3-TiO2 particles release less NOx than Equal amounts of 4 wt.% NH4NO3-TiO2 particles at 293K and 0.8% relative humidity, which may be the result of the Relative molecular mass difference between KNO3 and NH4NO3.

6. In section 3.3.4. more different initial concentration of HCHO should be test to find out from which content the positive effect become weakening.