|The authors have given so many responses to my comments. However, there are still some responses couldn’t solve the defects of the work. |
1. The authors stated that “DRIFTS measurements were made to test their homogeneity”. The particles were laid and aggregated on the holder in DRIFTS experiment, not the real particles suspended by sprayed into the chamber. I suggested the authors to collect the particles from the chamber to do SEM or TEM, which can give the clear composition of the suspended particles in the chamber, to see whether the particle is physical mixing or chemical dopped.
2. The quantity and the wall loss of particles still have large uncertainty, and this defect wasn’t well solved. The size of those symbols in Figure 3-5 is too large to clearly see the difference of particle size distribution at different time. Why did the authors not put all the plots in one figure? I feel that these three figures have some differences. The particle number is very important for the quantification and kinetics study, but we can see “about” in front of most particle numbers description. As I pointed before, large wall loss of particles must exist after the particles sprayed in the chamber. The value of 75mg/250L particles were injected into the chamber, only 8000 particle/cm3 (dN/dLog(dp)) or 4000 particle/cm3 were detected for the most abundant diameter particle, where were the other particles? were they still in the chamber? Heterogeneous reactions can also occur on the wall! Larger particle can easily deposit on the wall, which also can be found from the difference of the particle distributions between 0 min and hundreds of minutes later in Figure 3-5. Besides, the detected particle size was only up to 250 nm, how about those larger particles? These results were not enough to predict the wall effect.
3. The proposed mechanisms still couldn’t convince me. What’s the main significance of this article? The quantity of NOx has great uncertainties, and most of the mechanisms were deduced from so many hypotheses. Though some phenomena and products were observed and detected in previous studies, those proposed key intermediates and products were not directly measured in your experiment. If all those phenomena could be deduced and expected from previous studies, there is no need to do such experiments.
Some supplementary experiments are needed, not in the future but in the present work!
4. The zero-order reaction kinetics still not right. The y-axis is wrong for the first-order reaction. I feel that the correlation is better in red line than black line in Figure 10. why did the R2 shows opposite results? I still can’t accept the zero-order reaction kinetics, the red dots are still not in a line but a curve line with little difference from a line. The authors misunderstand my points, I mean the “60 min for HCHO equilibrium after particles injection” are maintained too long time, that a lot of HCHO (3 ppm HCHO as shown in Figure 12) will be adsorbed on the particles. Then two aspect effects will appear, the first is that HCHO occupy the active sites of particles which inhibit the photoreactions, the other is that both the adsorbed HCHO and gas-phase HCHO can attend the photoreactions. These effects lead the results of Figure 11. The authors should carefully consider the kinetic analysis part, in case mislead the readers.
5. The authors misunderstand my question about the mixture of HNO3 and TiO2. HNO3 has small affect on TiO2, but this treatment has great impact on HNO3. The mixture can’t be HNO3-TiO2. What’s the state of HNO3 on TiO2 surface? gas adsorbed? nitrate? liquid? definitely not HNO3 solid!!!