the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Chemical evolution of primary and secondary biomass burning aerosols during daytime and nighttime
Amir Yazdani
Satoshi Takahama
John K. Kodros
Marco Paglione
Mauro Masiol
Stefania Squizzato
Kalliopi Florou
Christos Kaltsonoudis
Spiro D. Jorga
Spyros N. Pandis
Athanasios Nenes
Abstract. Primary emissions from wood and pellet stoves were aged in an atmospheric simulation chamber under daytime and nighttime conditions. The aerosol was analyzed with the online Aerosol Mass Spectrometer (AMS) and offline Fourier transform infrared spectroscopy (FTIR). Measurements using the two techniques agreed reasonably well in terms of the organic aerosol (OA) mass concentration, OA:OC trends, and concentrations of biomass burning markers – lignin-like compounds and anhydrosugars. Based on the AMS, around 15 % of the primary organic aerosol (POA) mass underwent some form of transformation during daytime oxidation conditions after 6–10 hours of atmospheric exposure. A lesser extent of transformation was observed during the nighttime oxidation. The decay of certain semi-volatile (e.g., levoglucosan) and less volatile (e.g., lignin-like) POA components was substantial during aging, highlighting the role of heterogeneous reactions and gas-particle partitioning. Lignin-like compounds were observed to degrade under both daytime and nighttime conditions, whereas anhydrosugars degraded only under daytime conditions. Among the marker mass fragments of primary biomass burning OA (bbPOA), heavy ones (higher m/z) were relatively more stable during aging. The biomass burning secondary OA (bbSOA) became more oxidized with continued aging and resembled those of aged atmospheric organic aerosols. The bbSOA formed during daytime oxidation was dominated by acids. Organonitrates were an important product of nighttime reactions in both humid and dry conditions. Our results underline the importance of changes to both the primary and secondary biomass burning aerosols during their atmospheric aging. Heavier AMS fragments seldomly used in atmospheric chemistry can be used as more stable tracers of bbPOA and in combination with the established levoglucosan marker, can provide an indication of the extent of bbPOA aging.
- Preprint
(1524 KB) -
Supplement
(801 KB) - BibTeX
- EndNote
Amir Yazdani et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-658', Anonymous Referee #1, 06 Nov 2022
This paper examines the chemical evolution of primary and secondary biomass burning aerosols using typical daytime and nighttime oxidants in the laboratory. The combination of online AMS and offline FTIR provides a unique analysis of the chemical characteristics of the aging of bbPOA and its conversion to bbSOA. The high-quality paper addresses an important area of atmospheric chemistry and is well-suited for publication in ACP. Below are minor comments on clarification issues for the authors' consideration.
- Page 3 Line 74, The authors should give more details of the UV light source in the text here. Also, see the last comment since there might be potential photo-induced reactions worth discussing in conclusions/implications.
- Page 5 Line 133, Sres(0), is it a typo? 0 should be t?
- Page 7 Line 171, What are "the first three principle components"? A clearer discussion of the figure using consistent terms in text and legend/caption would help. Please also note a few typos of "principle" components.
- Page 7 Line 179, it is interesting to see the differences between dark and dark/humid conditions. Please explain these differences. Also, what are the directions of these changes in the dark in the f44/f43 plot? I notice the trends for WB and PB under dark and dark/humid are not the same. Please explain the differences.
- Figure 2, please provide clearer legends. WB dark and PB dark mean dark and dry condition? What is WB total OA? Why did it not show PB total OA? Need clear descriptions.
- Figure 3, please give better figure captions to avoid ambiguity.
- Why does Fig. 5 use experiments 1, 4, 6, and 8, while Fig. 7 uses experiments 1, 4, 7, and 8? Typo?
- Page 10 line 273 – 275, please consider rewriting so that the changes in the gas phase and particle phase can be distinguished.
- Table 2, why is the kOA under UV larger than under dark conditions?
- The reference experiments are helpful, but why was there no 50% RH data? Also, why were different RH used in the wood and pellet experiments for the dark and humid conditions?
- Page 12, Line 345, heterogeneous reactions and photolysis of bbPOA are indeed complex. Recent work on bb particles and model bb chemical compounds has suggested photosensitization can be an important process in SOA and sulfate formation (Matabo et al., 2022, Liang et al., 2022). Furthermore, I would be interested in knowing if the particles contain nitrate. Furthermore, nitrate photolysis can be an effective pathway to form sulfate (from SO2) and possibly SOA due to the formation of in particle OH, NO2 and nitrite (Zhang et al., 2022). Some discussions on these possibilities and their potential influence on the experimental results would be helpful.
References:
Liang et al., Sulfate Formation in Incense Burning Particles: A Single-Particle Mass Spectrometric Study. Environmental Science Technology Letters. 2022, 9, 9, 718–725.
Mabato et al., Aqueous secondary organic aerosol formation from the direct photosensitized oxidation of vanillin in the absence and presence of ammonium nitrate. Atmospheric Chemistry and Physics 2022, 22, (1), 273-293.
Wang et al., Atmospheric Photosensitization: A New Pathway for Sulfate Formation
Environ. Sci. Technol. 2020, 54, 6, 3114–3120
Zhang et al., Photochemical Reactions of Glyoxal during Particulate Ammonium Nitrate Photolysis: Brown Carbon Formation, Enhanced Glyoxal Decay, and Organic Phase Formation. Environ. Sci. Technol. 2022, 56, (3), 1605-1614.
Citation: https://doi.org/10.5194/acp-2022-658-RC1 -
AC1: 'Reply on RC1', Satoshi Takahama, 27 Jan 2023
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-658/acp-2022-658-AC1-supplement.pdf
-
RC2: 'Comment on acp-2022-658', Anonymous Referee #2, 06 Dec 2022
The data presented within Yazdani et al. is interesting because it shows a potentially overlooked process (heterogeneous processes or photolysis) that is changing the composition of the POA. The text, figures, and tables needs to be seriously checked and cleaned up. There are a multitude of errors or unclear aspects of the text, figures, or tables that resulted in assumptions about what the authors intended. I would argue that this text can be seriously streamlined to tell the main message.
The initial discussion was centered on differences in the FTIR measured between different burning conditions, but the discussion requires more discussion on the variability of the burning conditions and day-to-day variability to say anything of substance. The authors then talk about the different oxidation processes and their impact on the bulk chemical properties measured by the AMS and FTIR. There are minor points where the authors do not provide enough information to make an informed assessment about the presented information (see thermal denuder discussion, NH4 subtraction in the FTIR data). The interesting point of the manuscript comes from the
My biggest uncertainty regarding this manuscript comes from the assessment of levoglucosan with the FTIR. There appears to be a sizeable disconnect between the values reported in Table 2 and those shown in Figure 7 for levoglucosan. I cannot reconcile these differences. These apparent discrepancies and the errors / unclear aspects of the text do not provide confidence that they are simple to explain. I do not recommend publishing at this time and a major revision is necessary.
Major Comments:
Table 1: I believe that SO2 is supposed to be O3. The OH concentration noted in Table 1 does not correspond to the initial conditions and was determined from the PTR-MS with the d9-butanol injection. What was the error of these values in Table 1? In the text the average is 3-5 x 106 molecule cm-3, but this is purely both of the values reported in Table 1.
Why does the experiment with higher ozone have lower OH exposure?
What was the modified combustion efficiency of the burns? Were there no CO/CO2 measurements? Does this describe any of the differences / variability between the fuel types?
Without knowing the modified combustion efficiency how can these measurements be effectively translated between measurements performed here and those performed at other facilities?
Page 4 line 109-end of page: I do not believe this section is clearly worded. At first reading, without detailed understanding of the data analysis of the FT-IR it is not clear how the information about the absorption coefficient is needed. A discussion about the absorption coefficient is needed here. Also, can more information be provided about this scaling because it is not clear in Reggente et al. (2019), and about the uncertainty that this introduces to the data presented here.
Section 2.3.2: As the wall-losses can be size-dependent, which can be important for large size-distributions such as those found with combustion emissions, is a mass based wall-loss correction the more appropriate correction? Could the SMPS also be used to provide a size-dependent wall loss correction, as well as a correction that occurs during aging? The argument of well-mixed POA is not necessarily valid when considering these aging pathways.
Section 3.1 paragraph 1: I understand the authors are trying to make the point that there are different components coming from different types of burning species. How confidant are the authors that these are unique features and not driven from burning variability? How do the authors verify that the burning conditions from experiment to experiment are reproducible? Where do you show or discuss the variability? Otherwise, I don’t see how this part of the discussion is useful, and raises significant questions about the rest of the manuscript. If there are no metrics to say that each burn is starting roughly near the same point, how can the authors be confidant that their results are meaningful? Thankfully, in the later part of the manuscript there are no conclusions about the differences between pellet vs. wood burning.
Figure 1: Are these averages across all experiments? Are these representative spectra from specific experiments? Can these be specified. WB and PB are not defined. What is the variability between experiments? If they are from specific experiments, can you specify which experiments.
Figure 2: I guess the legend should change dark to “dark and dry” to be consistent with the “dark and humid” Maybe it would be helpful to demonstrate where the initial data point is for each experiment? As an atmospheric scientist that uses an aerosol mass spectrometer, I understand what the “beginning” of the experiment is when looking at this plot, but it can be clarified better so non-experts can understand this as well. Also, why don’t the authors put the same points used in Figure 3 on Figure 2? This would provide consistency in the discussion.
Page 7 line 172: where are atmospheric bbOA factors shown in Figure 2?
Page 7 line 178 – 182: Many of the authors on this paper are also authors on Kodros et al (2020), and they observed clear enhancements in OA due to the exposure to NO3 radicals under both dry and humid conditions. In the experiments presented here, what is the OA enhancement under dry and humid conditions? Are they the same as Kodros et al (2020)? I would guess not, because the results presented in Kodros et al (2020) show that even under dry and dark conditions there is an increase in f44. Why are there differences between these results (dark and dry / dark and humid) to the results in Kodros et al (2020) – see Figure 3A?
Because of the differences in the results presented here vs. Kodros et al (2020) it begs more questions about what drives the observations? Are single experiments ‘good enough’ to derive meaningful conclusions? For example, these authors in Kodros et al. (2020) had 4 repeats of dry experiments, and 2 repeats of humid experiments.
Figure 3: What does the legend mean? WB-P, WB-A, PB-A, PB-A?? I guess these are supposed to be P = primary and A = aged. I also suppose that there is a typo in the legend and one of the PB-A should be PB-P…. I have no idea what experiments WP nor PB correspond to… Are they dark NO3 experiments? Are the UV experiments? Are they dry? Are they humid? I am going to ignore the points corresponding to WB and PB, because I don’t know what they correspond to. I will assume that the markers for each aging point matter the most.
Page 7 lines 183 – 200: Wouldn’t it be most accurate to talk about this in terms of oxidation rather than ‘aged’? The metric that you are using to assess the mass spectra is effectively how oxygenated the molecules are, the “aging” length is known based on the time scale and the extent of the oxidation that the particles are undergoing. In these terms, you can talk about something concrete, because in these experiments you know what is happening to the particles is not some ambiguous process.
Figure 6: The difference between the average value (black) vs. the error bars (red) are difficult to observe the differences. What the error bars correspond to? The variation during the experiment? Standard deviation of the signal over the whole experiment? The standard deviation over a specific window? What time frame of the experiment was chosen for these figures? Why are the B, C, D figures not showing the aerosol constituents that are decreasing? This is an important point in those experiments. I recommend showing the full version of the figures. Also, what experiments do these correspond to? Because in Figure 7 the experimental data does not correspond to the same exact experiments as Figure 5.
Page 9 line 244: I would argue that the negative aspects of the mass spectra are at least as interesting as those increasing and would advocate to add those to the Figure 6.
Section S5: How is this subtraction actually performed? In the text the carboxylic acid portion of the is the focus of how the NH4+ obscures the data in the FT-IR. When looking at Figure S6 the CH bands are also effectively subtracted. Does this mean that the aCH are also obscured here?
Section S6: do you want to compare that to your data? It seems like that would be a reasonable thing to do.
Page 9 line 265: I am very curious about how Table 2 is formed for levoglucosan. When I look at Figure 7b I do not see an appreciable loss of levoglucosan.
Lets start with Figure 1, there is a clear signal that is associated with levoglucosan just below 1000 cm-1. Now when I look at Figure 7b is there even levoglucosan there? Figure 1 also shows that there is negligible levoglucosan present. So are the values in Table 2 for levoglucosan reasonable and above the limits of detection?
I am at a loss to understand values reported in Table 2 for levoglucosan for the dark experiments for the FT-IR measurements. Note all values are less than 1 meaning that the values should be decreasing with the dark experimental aging times.
Now in Figure 7A, the levoglucosan peak is negative in the POA sample, and is maybe still negative (but less so) in the Aged sample. This results in a net positive residual. Based on this analysis how can the FT-IR data reported in Table 2 be even close to the reported value for experiment 1? I have similar questions about the data in Figure 7C where it looks like the levoglucosan region is increasing with aging
Page 10 lines 283-290: without solid evidence for their thermal denuder modelling, I believe the only points the authors can use is that the TD shows levoglucosan is more volatile than their lignin marker. Wouldn’t the most appropriate comparison be investigating the volatility prior to initiating the experiment?
Section S2: What experiment is shown in Figure S2?
Why is equilibrium partitioning a valid assumption with the thermal denuder? The residence time of the thermal denuder plays a massive role on the extent of evaporation, and if the TD has actually achieved equilibrium.
There are no details about the length, flow rate, temperature homogeneity of the TD, nor residence times. These parameters are required to determine any thermodynamic properties.(Saha and Grieshop, 2016; Bilde et al., 2015; Salo et al., 2010; Saleh et al., 2010) The lack of details presents a major flaw, and a concrete story about the volatility of the bbOA cannot be made without these details provided and discussed.
Losses due to thermophoreisis are not included, yet this can result in losses of ~10-15% in other thermaldenuders at elevated temperatures. Typically these losses are more important at high temperatures and longer residence times, where equilibrium can be assumed.
Why was 40 kJ mol-1 used to model the thermal denuder data? Compilations of data comparing saturation vapor concentrations and the enthalpy of vaporization suggests the enthalpy of vaporization should be in the ball park of 90-120 kJ mol-1 for molecules with a Log C* ~1 . See (Epstein et al., 2010) and (Macleod et al., 2007)
Page 10 line 291-294: Looking at Figure S7 it appears that the C10H13O3+ increases initially upon irradiation. This means there is also a source for this fragment in the AMS, this would make it a tenuous marker. (I also disagree with this point in the conclusions)
Page 10 line 301 to end of paragraph: my concerns about the quantification in Table 2, makes me doubt the authenticity of the values reported here.
Page 11 lines 310-312 (also page 12 line 345): Can’t the net carbon loss and increased oxidation also be a result of evaporation of semi volatile species + OH oxidation in the gas phase with subsequent re-condensation? Based on this experiment, this is not a tell-tale sign of heterogeneous processes.
Minor Comments:
Section 2.1 How was the NO2 and O3 concentrations assessed? Were they measured?
What type of PTR-MS was used? (I guess this should be reported)
Section 2.2: What was the typical size distributions measured during these experiments? (Were they reproducible?)
Page 2 line 93: what is the resolution? Is it 1 cm-1 or something else?
Page 4 line 111: I believe the full citation should be in paraentheses.
Page 4 line 113: The reference database needs to be cited.
Page 6 line 164: CE is not defined, but I presume this is collection efficiency. Also, where is this shown in the previous section? I read this to mean in the previous section (section 2.3.4?) that the OA concentrations were correlated with the OM:OC, I don’t understand what is meant here.
Page 7 line 171: What are the “first three principle components”? <- also I think principle is misspelled.
Page 7 line 184: Does the residual spectra refer to the chamber mass spectrum at the end of the experiment? (Or is this residual from the PCA analysis?)
Page 7 line 186: The authors should note that the mass spectrum from each PC is found in the supplement Figure S5.
Page 7 line 187: What is meant by loadings? Is it equivalent to intensity?
Page 7 line: 197-198: “Additionally, the PCA analysis indicates the functional group content of aerosols discussed in this and following sections.” What does the PCA analysis indicate?
Page 7 line 199 – end of page: Does Figure 3 show this?
Page 8 line 210: Doesn’t the agreement between the FTIR and AMS just mean that they agree…? it doesn’t say anything independent about if the wall loss correction was performed correctly.
Page 9 line 243: “… formation of new oxidized species” I would add “as expected based on the mass formation observed.” But in order to say this you would need to include what the OA mass increase is.
Page 9 line 258: It looks to me in Figure 6b there is one other species that decreases more than levoglucosan (but less than C9H11O3+), what is that?
Page 9/10 lines 269 – 277: This conversation is muddled. I think the authors want to convey that different processes could be happening, but can’t single out a specific pathway that is definitively occurring.
Page 9 line 269: oxidation reactions
Page 9 line 270: Bertrand et al. also includes OH reactions in their assessment of levoglucosan losses. I don’t see what is added in the discussion about the results presented here, maybe the authors could say this is consistent with these results.
Page 10 line 275: Maybe it would be clearer to say “fragments from less volatile…” having diminution and produced in the same sentence is a bit confusing.
Supplemental:
Experiment S1: Why is the SMPS data not used at all in this paper? The size distribution shown in Figure S1 shows a peak at 200nm, but what is the cut-off of the lens used for the measurements?
References:
Bilde, M., Barsanti, K., Booth, M., Cappa, C. D., Donahue, N. M., Emanuelsson, E. U., McFiggans, G., Krieger, U. K., Marcolli, C., Topping, D., Ziemann, P., Barley, M., Clegg, S., Dennis-Smither, B., Hallquist, M., Hallquist, Å. M., Khlystov, A., Kulmala, M., Mogensen, D., Percival, C. J., Pope, F., Reid, J. P., Ribeiro da Silva, M. A. V., Rosenoern, T., Salo, K., Soonsin, V. P., Yli-Juuti, T., Prisle, N. L., Pagels, J., Rarey, J., Zardini, A. A., and Riipinen, I.: Saturation Vapor Pressures and Transition Enthalpies of Low-Volatility Organic Molecules of Atmospheric Relevance: From Dicarboxylic Acids to Complex Mixtures, Chemical Reviews, 115, 4115-4156, 10.1021/cr5005502, 2015.
Epstein, S. A., Riipinen, I., and Donahue, N. M.: A Semiempirical Correlation between Enthalpy of Vaporization and Saturation Concentration for Organic Aerosol, Environmental Science & Technology, 44, 743-748, 10.1021/es902497z, 2010.
MacLeod, M., Scheringer, M., and Hungerbühler, K.: Estimating Enthalpy of Vaporization from Vapor Pressure Using Trouton's Rule, Environmental Science & Technology, 41, 2827-2832, 10.1021/es0608186, 2007.
Saha, P. K. and Grieshop, A. P.: Exploring Divergent Volatility Properties from Yield and Thermodenuder Measurements of Secondary Organic Aerosol from α-Pinene Ozonolysis, Environmental Science & Technology, 50, 5740-5749, 10.1021/acs.est.6b00303, 2016.
Saleh, R., Khlystov, A., and Shihadeh, A.: Effect of Aerosol Generation Method on Measured Saturation Pressure and Enthalpy of Vaporization for Dicarboxylic Acid Aerosols, Aerosol Science and Technology, 44, 302-307, 10.1080/02786821003591810, 2010.
Salo, K., Jonsson, Å. M., Andersson, P. U., and Hallquist, M.: Aerosol Volatility and Enthalpy of Sublimation of Carboxylic Acids, The Journal of Physical Chemistry A, 114, 4586-4594, 10.1021/jp910105h, 2010.
Citation: https://doi.org/10.5194/acp-2022-658-RC2 -
AC2: 'Reply on RC2', Satoshi Takahama, 27 Jan 2023
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-658/acp-2022-658-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Satoshi Takahama, 27 Jan 2023
Amir Yazdani et al.
Amir Yazdani et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 475 | 198 | 12 | 685 | 50 | 3 | 5 |
- HTML: 475
- PDF: 198
- XML: 12
- Total: 685
- Supplement: 50
- BibTeX: 3
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1