Atmospheric Simulation Chamber Studies of the Gas-Phase

Oct 16, 2017 - Reed Harris, Pajunoja, Cazaunau, Gratien, Pangui, Monod, Griffith, Virtanen, Doussin, and Vaida. 2017 121 (18), pp 3327–3339. Abstrac...
2 downloads 0 Views 3MB Size
Subscriber access provided by READING UNIV

Article

Atmospheric Simulation Chamber Studies of the Gas-Phase Photolysis of Pyruvic Acid Allison Early Reed Harris, Mathieu Cazaunau, Aline Gratien, Edouard Pangui, Jean-Francois Doussin, and Veronica Vaida J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b05139 • Publication Date (Web): 16 Oct 2017 Downloaded from http://pubs.acs.org on October 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Atmospheric Simulation Chamber Studies of the Gas-Phase Photolysis of Pyruvic Acid Allison E. Reed Harris,1 Mathieu Cazaunau,2 Aline Gratien,2 Edouard Pangui,2 Jean-Francois Doussin,2* and Veronica Vaida1*

1.

Department of Chemistry and Biochemistry, CIRES, University of Colorado, Boulder, Colorado, USA

2.

LISA, UMR-CNRS 7583, Université Paris Est Créteil (UPEC), Université Paris Diderot (UPD), Institut Pierre Simon Laplace (IPSL), Créteil, France

*Corresponding Author Email: [email protected]







and [email protected]

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 47

Abstract Pyruvic acid is an atmospherically-abundant a-keto-acid that degrades efficiently from the troposphere via gas-phase photolysis. In order to explore conditions relevant to the environment, 2-12 ppm pyruvic acid is irradiated by a solar simulator in the environmental simulation chamber, CESAM. The combination of the long path length available in the chamber and its low surface area to volume ratio allow us to quantitatively examine the quantum yield and photochemical products of pyruvic acid. Such details are new to the literature for the low initial concentrations of pyruvic acid employed here. We determined photolysis quantum yields %

'() & of !"#$ = 0.84 ± 0.1 in nitrogen and !"#$ = 3.2 ± 0.5 in air, which are higher than those reported

by previous studies that used higher partial pressures of pyruvic acid. The quantum yield greater than unity in air is due to secondary chemistry, driven by O2, that emerges under the conditions in these experiments. The low concentration of pyruvic acid and the resulting oxygen effect also alter the product distribution such that acetic acid, rather than acetaldehyde, is the primary product in air. These results indicate that tropospheric pyruvic acid may degrade in part via photo-induced mechanisms that are different than previously expected.





ACS Paragon Plus Environment

2

Page 3 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

I.

Introduction Organic molecules and their reactivity have a significant impact on Earth’s atmospheric composition and chemistry, especially regarding the processing of pollutants, aerosols, and trace gas concentrations.1-5 Volatile organic compounds (VOCs) make up a substantial component of the troposphere, with an estimated worldwide flux on the order of 1000 Tg C/yr.6-8 Biogenically emitted isoprene comprises a large portion of the total VOC budget (~500 Tg C/yr),7, 8 and, when oxidized, leads to an extensive series of cascading reactions that has been linked to the formation of secondary organic aerosol.9-19 Pyruvic acid, a keto-acid intermediate within the network of isoprene oxidation pathways,1719

is abundant in the atmosphere, with gas-phase mixing ratios up to 100 ppt and particle

concentrations up to 140 ng/m3.20-30 Unlike many tropospheric VOCs that degrade predominantly via attack by the hydroxyl radical (OH), gas-phase pyruvic acid undergoes efficient direct photolysis that dominates over OH oxidation by several orders of magnitude under atmospheric conditions.31-33 Pyruvic acid is subject to direct photolysis in the troposphere largely because its carbonyl functionality has an n to

* electronic transition with a

max near

350 nm,34 allowing most of its S1 absorption spectrum to be accessed with radiation readily available near Earth’s surface.35 The pyruvic acid system therefore provides an interesting case study for photochemical and photosensitized reactions that may be important in atmospheric processes.36-47 Since photolysis is an out-of-equilibrium process reliant on the energy from impinging radiation to break chemical bonds, the kinetics and mechanisms of photochemical reactions are inherently dependent on the wavelength and irradiance of such radiation.48 First order



ACS Paragon Plus Environment

3

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 47

photolysis rate constants, J values, are then defined in terms of wavelength dependent parameters as follows: * = where F(

) is the photon flux,

(

3& ,- / 34

- 0 - !(-), Equation 1

) is the molecular cross section, and

(

) is the quantum

yield for molecular decomposition upon photon absorption. Parameters such as pressure, temperature, and concentration can also play roles in defining the rates and products from photochemical processes by influencing

(

).49-52

While there is a large body of literature regarding the gas-phase unimolecular decomposition of pyruvic acid,53-59 due to this wavelength dependence, the work conducted here is most directly comparable to prior studies of pyruvic acid in which photolysis was initiated by wavelengths of light contained in the actinic flux (l > 300 nm). Literature of pyruvic acid photolysis using such radiation typically can be separated into two categories. The first investigates the low-pressure (0-150 Torr of buffer gas) photochemistry of approximately 1 Torr of pyruvic acid, a concentration several orders of magnitude higher than would be expected in the lower atmosphere.60-62 These studies, bolstered by computational work,63-71 conclude that, following photon absorption, gas-phase pyruvic acid undergoes simultaneous concerted hydrogen atom transfer and decarboxylation with a quantum yield of unity.60-62 The immediate products are carbon dioxide and a reactive intermediate, methylhydroxycarbene (CH3COH), which is observed as acetaldehyde (CH3CHO) following isomerization.60-62 The second class of studies regarding this photolysis examines lower initial concentrations of pyruvic acid (0.33-200 ppm) in air at atmospheric pressure, although these concentrations remain above atmospheric levels.72-76 Quantum yields detected from these higher-pressure

ACS Paragon Plus Environment

4

Page 5 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

experiments range between

= 0.21 and

= 0.85,72-75 and additional minor products, such as

acetic acid, are typically identified.72-76 A recent laboratory study by Reed Harris et al.72 helped bridge the gap between the results from experiments employing high partial pressures of pyruvic acid under low buffer gas pressures with those employing low partial pressures of pyruvic acid in one atmosphere of air. To do so, they irradiated pyruvic acid with a solar simulator and examined the quantum yield and products as a function of the buffer gas pressure and the initial mixing ratio of pyruvic acid. They documented a decrease in quantum yield with increasing pressure of buffer gas and increasing concentration of pyruvic acid.72 Nevertheless, the short optical path length of the laboratory set up constrained the mixing ratios of pyruvic acid to be three to four orders of magnitude higher than what has been detected in the troposphere. Further, pyruvic acid is a highly oxidized compound and, therefore, tends to partition to surfaces. The high surface area to volume ratio of the laboratory cell limited the pressures of buffer gas to those less than 600 Torr, as pyruvic acid losses to the walls competed with photolysis at higher pressures. Therefore, the results from Reed Harris et al.72 necessitate further investigation of the photolysis of pyruvic acid using lower initial mixing ratios and higher total pressures in order to extend the understanding of this chemistry to conditions approaching those typically found in the atmosphere. In this paper, we discuss the photolysis of gas-phase pyruvic acid in the environmental simulation chamber, CESAM (French acronym for Experimental Multiphasic Atmospheric Simulation Chamber).77 Here, the photolysis of pyruvic acid is investigated at initial concentrations two orders of magnitude lower than those in the Reed Harris et al.



ACS Paragon Plus Environment

5

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 47

experiments.77 Because of its long path length and low surface area to volume ratio, CESAM provides the unique ability to examine compounds with very low mixing ratios. Furthermore, the results are atmospherically relevant as the chamber is operated at atmospheric pressure and temperature, and employs a solar simulator to initiate photolysis.77 The results indicate significant modification to the quantum and product yields from the photolysis of pyruvic acid as its initial concentration approaches atmospheric levels.77 II.

Methods To investigate the photolysis of pyruvic acid under atmospheric conditions, we employ the CESAM simulation chamber, which allows for in situ tracking of chemical species in a highlyregulated environment.77, 78 Wang et al.77 and Brégonzio-Rozier et al.78 detail a comprehensive explanation of CESAM, its experimental tools, and observational instruments; therefore, here we describe only the features that are key to this study and its atmospheric relevance. CESAM is a 4.2 m3 stainless steel chamber, irradiated by three 4000 W high-pressure Xe arc lamps. Before entering the chamber, the light is filtered through Pyrex windows (6.5 mm) to remove radiation below 300 nm and create a spectrum qualitatively similar to that of the actinic flux in the troposphere.77, 78 This is evidenced by Figure 1, which shows the irradiation spectrum in CESAM overlaid with the solar spectrum and the pyruvic acid UV-absorption cross section.79, 80 The exact flux in the chamber is routinely quantified with NO2 actinometry, following the procedure recommended by Holmes et al.79, 81 According to such calibrations, the flux in the chamber is 3 to 4 times lower in intensity than the maximum solar flux at mid latitude (i.e. noon, 21 th of June, Northern hemisphere). The chamber is continuously stirred with a 50 cm



ACS Paragon Plus Environment

6

Page 7 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

stainless-steel fan at the bottom of the chamber, resulting in mixing times of approximately 100 s.

Figure 1: CESAM UV lamp spectrum (yellow line) and solar spectrum (red line) in the region of the S1 transition of the pyruvic acid UV absorption cross section (grey line). The pyruvic acid cross section is reproduced from Sander et al., and its scale in this figure is indicated by the value given at its lmax.82 The solar spectrum was calculated with the tropospheric ultraviolet-visible model (TUV, version 4.5)80 for a solar zenith angle of 40°, overhead ozone column of 350 du, and a surface albedo of 0.2. Prior to each experiment, CESAM was pumped on overnight (P ≤ 4 × 10-4 mbar) to ensure removal of all residual compounds. Because the chamber can be fully evacuated, we maintain the extremely valuable capability to choose the desired buffer gas for each reaction and, therefore, to probe the role of oxygen in the pyruvic acid photochemistry by varying the ratio of O2 to N2. To begin a reaction, CESAM was filled with evaporated liquid nitrogen (N2 (l), purity



ACS Paragon Plus Environment

7

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 47

4.5, Messer) and gaseous O2 from a high purity cylinder (O2 (g), purity 5.0, Linde) to produce the desired mixture. Pyruvic acid (98%, Sigma-Aldrich) was distilled twice under reduced pressure and degassed with 10 to 12 freeze-pump-thaw cycles. It was then vaporized into an evacuated 2.9 L glass bulb and introduced into the chamber in a flow of nitrogen. This procedure was repeated to obtain pyruvic acid mixing ratios between 2 and 12 ppm. These concentrations are about 100 times lower than in most previous studies but remain at least 10 times greater than the anticipated amount of gas-phase pyruvic acid in the troposphere.27, 30, 72, 75 The introduction of pyruvic acid into the chamber sometimes required numerous injections due to the tendency of pyruvic acid to stick to the glass bulb when at relatively high concentrations. At least 30 minutes of data was collected preceding irradiation to ensure no dark chemistry ensued and to quantify pyruvic acid’s loss to the walls. Table 1 provides a summary of the experiments conducted for this study. Table 1: Summary of Key Gas-Phase Photolysis Reactions in CESAM Number of Trials Experiment Buffer Gas [O2] (ppm) 5 6 Pyruvic Acid Photolysis Air 3.5 x 10 3 Pyruvic Acid Photolysis N2 < 20 Pyruvic Acid Photolysis Trace [O2] 20-50 4 5 Pyruvic Acid Photolysis 50% N2, 50% O2 8.4 x 10 1 2 Pyruvic Acid Photolysis N2 with 5% O2 Spikea 8.4 x 104 Pyruvic Acid Photolysis Air 3.5 x 105 2 with Cyclohexane a For these reactions, pyruvic acid was first photolyzed in nitrogen for 1-2 hours. Subsequently, ~5% O2 was injected into the chamber to allow us to observe changes in the rates of pyruvic acid decay and product formation upon the change in buffer gas.



ACS Paragon Plus Environment

8

Page 9 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

CESAM is equipped with a number of instruments to monitor the conditions and chemical evolution in the chamber.77, 78 The temperature and relative humidity are tracked by a HUMICAP probe and HMP 234 Vaisala humidity and temperature transmitter. Though neither NOx nor O3 are expected from the photolysis of pyruvic acid, their concentrations were measured throughout the experiment to ensure that no contamination could enhance pyruvic acid’s decay (NOx-Horiba APNA 370, O3-Horiba APOA 370). Pressure lost to instrumental sampling from the chamber was continuously replaced by a flow of dry synthetic air adjusted to keep the overall pressure constant. This flow was continuously recorded by the CESAM control system and the values were used to evaluate the dilution rates in the chamber prior to the calculation of any chemical rates. The concentration of pyruvic acid and its photoproducts were monitored throughout the reaction with a multipath Fourier Transform Infrared spectrometer (FTIR Bruker Tensor 37, path length of 192 ± 4 m). Spectra were collected with 0.5 cm-1 resolution and composed of 200 averaged scans taken over a 5 min time period. Lastly, a proton-transfer-reaction time-of-flight mass spectrometer (PTR-ToF-MS Series II, Kore Technology) was used to aid in chemical characterization of the gas phase. The PTR-ToF-MS continuously sampled the chamber at 0.1 LPM through 1 m of 3 mm i.d. Silcosteel tubing heated at 373 K, and used the following parameters: 1 min time resolution, m/z range 0 to 200, drift tube pressure = 1.65 mbar, drift tube temperature = 373 K, drift tube voltage = 400 V, resulting in a E/N of 132 Td (1 Td = 10−17 cm2 V−1). The determination of chemical concentrations and quantum yields follows a similar procedure to that used in Reed Harris et al.;72 therefore, unless a distinction is necessary, the



ACS Paragon Plus Environment

9

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 47

approach is only summarized here. Concentrations of compounds in CESAM were calculated from FTIR spectra using the Main Polwin MATLAB program, which uses a linear square fitting method to analyze experimental spectra through a comparison with standardized spectra (see supplemental information Figure S1-S3 for product identification). Once the concentrations were corrected for the dilution in the chamber due to instrument sampling, photolytic decay rates could be extracted. Wall losses were approximated as first-order processes and quantified by fitting the pyruvic acid concentrations during the 30 minute dark decay to Equation 2. PA = [PA]9 : ;

Equation 2

Assuming the rate constant for pyruvic acid wall loss (kd) does not change over the course of the experiment, the J value was then extracted from a first-order analysis, taking the wall losses into account by including kd in Equation 3. PA = [PA]9 : ;(



Equation 3

For the photolysis reactions in air, this requires a pseudo first-order analysis because these reactions initiate secondary chemistry that causes pyruvic acid’s decay to stray from a strictly exponential curve (see Figure 2a). To determine the best possible J values, we use only the first 1800 s of pyruvic acid irradiation, where the decay better obeys a single exponential trend. Nevertheless, in the presence of oxygen, we expect the data’s poor fit to exponential curves to be the largest source of error; therefore, we define the uncertainty of the individual J values to be the difference between the J value extracted from the fit to the first 1800 s of irradiation and that extracted by fitting an exponential curve to the entire range of data. When derived this way, the uncertainties reflect the error introduced from the pseudo first-order analysis.



ACS Paragon Plus Environment

10

Page 11 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 2: Example of the experimental decay of pyruvic acid (black dots) in CESAM in 1 atm of (a) air and (b) nitrogen. Irradiation begins at t = 0 (yellow dotted line). The grey line indicates the expected pyruvic acid loss to the walls (Equation 2), extrapolated from the decay of pyruvic acid in the dark. The red dotted line is the exponential fit to the first 1800 s of pyruvic acid loss during irradiation, and is the equation used to determine the J value for pyruvic acid (Equation 3). The blue line is an exponential fit to the loss of pyruvic acid for the duration of irradiation, and is used to define the error of the J value. To illustrate the discussion above, Figure 2a shows data from an average decay of pyruvic acid in air, along with curves fit to the dark decay (grey line), initial 1800 s of data (red dotted



ACS Paragon Plus Environment

11

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 47

line), and the entire dataset (blue line). The non-first order shape of the decay in Figure 2a clearly demonstrates there are secondary loss pathways for pyruvic acid when low concentrations are photolyzed in air. Conversely, pyruvic acid decays by a strictly exponential trend when photolyzed in nitrogen (Figure 2b). Thus, the calculation of J values in nitrogen does not require a pseudo first order analysis. Nevertheless, to maintain consistency, we determine the J value and error using the same method used in the air reactions. From the determined J values, a quantum yield for the photolysis of pyruvic acid can be determined using Equation 1. Since a constant, broadband spectrum was used for the experiments in CESAM, we can only determine an average quantum yield (given as

Avg in the

following discussion) over the illuminated portion of the S1 pyruvic acid absorption spectrum (280-400 nm). Nevertheless, the pyruvic acid quantum yields are atmospherically-relevant since we initiate the photochemistry using a solar simulator that qualitatively reproduces the actinic flux in the UV. III.

Results and Discussion III.A The Kinetics of Pyruvic Acid Photodecomposition in the CESAM Simulation Chamber When approximately 3 ppm of pyruvic acid is photolyzed in dry air in CESAM, we obtain an '() average quantum yield of !"#$ = 3.2 ± 0.5, suggesting an approximate lifetime in the

atmosphere of 30 min. This quantum yield is greater than unity and coincides with a nonexponential decay of pyruvic acid (Figure 2a), revealing secondary losses that have not been described in the literature to date. As this is the first record of such secondary chemistry, our reported quantum yield is larger than all previously published values. Among the highest reported photolysis quantum yields for pyruvic acid under atmospheric pressure of air is that



ACS Paragon Plus Environment

12

Page 13 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

published by Berges and Warneck.75 They found a quantum yield of

= 0.85 ± 0.16 when they

photolyzed 50-100 ppm pyruvic acid, concentrations at least an order of magnitude higher than those used in this study.75 They also employ a light source emitting monochromatic radiation of 350 nm,75 while we use a broadband solar simulator. Because of these experimental differences, the quantum yield found by Berges and Warneck75 is substantially smaller than that found here and not indicative of any secondary consumption of pyruvic acid. Other previously reported quantum yields are even lower, including that from Reed Harris et al.,72 who report

= 0.24 ± 0.05 from the photolysis of ~100 ppm pyruvic acid in 600 Torr of

air. Their publication, however, directly addresses the influence of the initial concentration of gas-phase pyruvic acid on the kinetics and products of the photolysis, using pyruvic acid pressures of 0.05-0.9 Torr. They documented an increase in quantum yield with decreasing concentration of pyruvic acid.72 Because CESAM allows for the accurate monitoring of species with low concentrations, the measured quantum yield in CESAM is for concentrations of pyruvic acid at least an order of magnitude lower than those in the Reed Harris et al. study.72 However, '() our detected !"#$ = 3.2 ± 0.5 is in agreement with the trend documented by Reed Harris et al.,72

and provides further evidence that concentration is a major factor in determining the photolysis quantum yield of pyruvic acid. We explore the role of oxygen in this chemistry by monitoring pyruvic acid’s decay in 1 atm %

& of nitrogen. For these experiments, we find an average quantum yield of !"#$ = 0.84 ± 0.1,

which corresponds to a lifetime of just over an hour. In contrast to the photolysis reactions in air, pyruvic acid obeys a first order decay in these experiments, indicating there is little to no secondary loss of pyruvic acid when it is photolyzed in nitrogen (Figure 2b). The exponential



ACS Paragon Plus Environment

13

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 47

%

& loss of pyruvic acid in nitrogen implies that our !"#$ = 0.84 ± 0.1 provides a good estimate of the

amount of pyruvic acid that directly photolyzes when ~2 ppm of pyruvic acid in one atm of dry buffer gas is irradiated with a broadband light source, excluding any pyruvic acid consumed by non-photolytic reactions. Despite being less than one, this value is also substantially higher than the previously reported quantum yield for the photolysis of pyruvic acid in 600 Torr of nitrogen (

= 0.27 ± 0.06).72 Therefore, while not as sizable as the effect in air, the lower

pyruvic acid concentrations used in CESAM in nitrogen also increase the pyruvic acid photolysis quantum yield. Although we have made a considerable effort to minimize oxygen in CESAM during these experiments by using the purest reagents available and by streamlining procedures, because of the size of CESAM, O2 may never be completely eliminated. Based on the required number of injections of pyruvic acid for each experiment (each one introduces ca. 6 ppm of air) and the purity of the liquid nitrogen used to fill the chamber, we estimate these reactions have < 20 ppm O2. Nevertheless, we could not take precise oxygen measurements in CESAM. For this reason, while we believe the experiments in N2 to be relatively free of secondary chemistry given the observed exponential decay of pyruvic acid (Figure 2b), the value we report here for pyruvic acid’s photolysis quantum yield in nitrogen should be taken as an upper bound for the true value. Because pyruvic acid’s decay in nitrogen is strictly exponential, pyruvic acid’s secondary losses in air are likely due to chemistry driven by O2. To probe whether the decay rate in air will further increase with increasing partial pressures of oxygen, we examine the photochemistry of pyruvic acid in one atmosphere of 50% O2 and 50% N2. The quantum yield for this reaction is



ACS Paragon Plus Environment

14

Page 15 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

'() well within the uncertainty of !"#$ , demonstrating there is no further modification to the

chemistry when the partial pressure of O2 is greater than ~21%. Similarly, when we inject 5% oxygen into the chamber after 1-2 hours of pyruvic acid photolysis in nitrogen, the rate of decay increases to match the rate of decay in air (Figure S4). These experiments indicate that, at a total pressure of 760 Torr and concentration of pyruvic acid on the order of parts per million, a buffer gas mixture with less than 5% oxygen can initiate the secondary chemistry observed in CESAM. While we only include the J values from the experiments with the three lowest anticipated partial pressures of O2 when calculating the quantum yield for the photolysis of pyruvic acid in nitrogen, we performed the reaction numerous times with trace concentrations of oxygen. When the estimated mixing ratios of O2 from these additional experiments are plotted against the J values obtained for every nitrogen based experiment, we find that the J value for pyruvic acid decomposition roughly increases with the predicted oxygen concentration in the chamber (Figure S5). The oxygen mixing ratios are on the order of 20-50 ppm, indicating that even trace amounts of oxygen can invoke secondary losses of pyruvic acid in CESAM. A previous analysis of the gas-phase photolysis of pyruvic acid found that it may generate OH with a percent yield of 5 ± 3%,31 suggesting hydrogen abstraction from pyruvic acid by OH is a possible secondary loss pathway. Though the reaction between pyruvic acid and OH is known to be quite slow, because of the evidence here for secondary chemistry, we irradiated pyruvic acid in the presence of a large quantity of cyclohexane (about 8 ppm for a pyruvic acid concentration of 1.5 ppm), a well-known OH scavenger.83 This experiment serves to ensure that OH oxidation is not a major contributor to the pyruvic acid decay.



ACS Paragon Plus Environment

15

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 47

Cyclohexane is known to react with OH with a rate constant of 6.97 × 10-12 cm3 molecule-1 s-1 (Atkinson, 2003)84 at 298 K, while the corresponding rate constant for the reaction between pyruvic acid and OH is 1.25 × 10-13 cm3 molecule-1 s-1 (Mellouki and Mu, 2003).31 Thus, if radical chemistry were occurring as a result of OH formed during the photolysis of pyruvic acid, cyclohexane would react with it approximately 300 times faster than pyruvic acid. This would essentially prevent the secondary decay of pyruvic acid and cause significant loss of cyclohexane. The rate of cyclohexane loss did not deviate from its dark decay at any point after irradiation began, confirming [OH] < 2 × 105 molecules/cm3 during the photolysis of pyruvic acid in CESAM (Figure S6). Even this maximum OH concentration would correspond to a loss of pyruvic acid through reaction with OH that is four orders of magnitude slower than the observed decay rate in the chamber, indicating this process cannot be responsible for the high quantum yield we observe in air. Taken together, the results from the experiments discussed above reveal that 2-12 ppm pyruvic acid in 1 atm of buffer gas has a direct photolysis quantum yield of f= 0.84 ± 0.1, when irradiated with broad band radiation similar to the solar spectrum. However, at the very low concentrations of pyruvic acid made observable by CESAM, concentrations of O2 greater than 20 ppm initiate secondary chemistry that rapidly consumes pyruvic acid, leading to a quantum yield greater than unity in air. We have ruled out a reaction with OH causing this additional pyruvic acid loss, and will present other possibilities for such chemistry after discussing the products observed from these reactions.

ACS Paragon Plus Environment

16

Page 17 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

III.B. Products from the Photolysis of Pyruvic Acid in CESAM Products identified here from the photolysis of pyruvic acid, in both air and nitrogen, include CO2, CO, acetic acid, and acetaldehyde. Formaldehyde and methanol are also observed, but only when the photolysis of pyruvic acid is conducted in the presence of oxygen. All of these products have been documented by previous studies,61, 62, 72, 74-76 although only Reed Harris et al. detected them all.72 Table 2 provides the average percent yields, calculated from at least three individual experimental runs, for the photolysis of pyruvic acid in air and nitrogen. The displayed yields are ratios of the total final product concentrations to the pyruvic acid concentrations lost to chemical processes throughout irradiation, indicating the number of molecules of each product created for every 100 pyruvic acid molecules that have chemically reacted. As photolysis is the only significant reaction pathway for pyruvic acid in nitrogen, the product yields for these reactions specifically indicate the molecules of each product created per 100 pyruvic acid molecules lost exclusively to photolysis. Conversely, because the photolysis of pyruvic acid in air includes significant secondary chemistry, the yields for this reaction indicate the ratio of the number of molecules of each product created per 100 pyruvic acid molecules lost to both photolysis and secondary processes. The percent yields are calculated from the final product concentrations, so they not only include primary photolysis products, but they may also contain compounds formed through any other processes, such as photolysis of photochemical intermediates and products, or from radical reactions. The reported errors in the table are the standard deviations of the product yields from the individual experimental runs used to determine the averages. PTRMS data



ACS Paragon Plus Environment

17

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 47

confirms these trends and detects an unidentified additional product at m/z 87, which appears to be inhibited in the presence of oxygen (Figure S7-S8).



ACS Paragon Plus Environment

18

Page 19 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

The Journal of Physical Chemistry

Table 2: Product Percent Yields from the Photolysis of Gas-Phase Pyruvic Acid in CESAMa Experimental Conditions Product Percent Yields data source light source buffer gas [PA]0 (ppm) b CO2 c CO CH3COOH CH3CH(O) HCH(O) CH3OH CESAM Solar Simulator Nitrogen 3-12 ppm 16±9 0.6±0.3 5±1 10±1 Reed Harris Solar Simulator Nitrogen 50-400 ppm 108±18 4±0.5 6.4±0.7 53±10 6.0±0.8 4.2±0.6 et al.72 CESAM Solar Simulator Air 3-12 ppm 28±5 0.8±0.4 16±3 0.8±0.8 2.2±0.6 1.4±0.4 CESAM, Solar Simulator Air 3-12 ppm 108±19 3±2 63±12 5±2 8±2 5±2 effective d Reed Harris Solar Simulator Air 50-400 ppm 137±26 5±2 9±2 51±9 4.1±0.2 2.3±0.7 et al.72 Berges and Monochromatic Air 50-100 ppm 127±18