Atmospheric Simulation Chamber Studies of the ... - ACS Publications

Article pubs.acs.org/JPCA

Cite This: J. Phys. Chem. A 2017, 121, 8348-8358

Atmospheric Simulation Chamber Studies of the Gas-Phase Photolysis of Pyruvic Acid Allison E. Reed Harris,† Mathieu Cazaunau,‡ Aline Gratien,‡ Edouard Pangui,‡ Jean-François Doussin,*,‡ and Veronica Vaida*,† †

Department of Chemistry and Biochemistry, CIRES, University of Colorado, Boulder, Colorado 80309, United States LISA, UMR-CNRS 7583, Université Paris Est Créteil (UPEC), Université Paris Diderot (UPD), Institut Pierre Simon Laplace (IPSL), Créteil, France

‡

S Supporting Information *

ABSTRACT: Pyruvic acid is an atmospherically abundant α-keto-acid that degrades efficiently from the troposphere via gas-phase photolysis. 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 allows 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 ϕNobs2 = 0.84 ± 0.1 in nitrogen and ϕAir obs = 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 photoinduced mechanisms that are different than previously expected.

I. INTRODUCTION Organic molecules and their reactivity have a significant impact on Earth’s atmospheric composition and chemistry, especially with regards to 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 have been linked to the formation of secondary organic aerosol.9−19 Pyruvic acid, a keto-acid intermediate within the network of isoprene oxidation pathways, 17−19 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 π* electronic transition with a λmax 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 © 2017 American Chemical Society

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 photolysis rate constants, J values, are then defined in terms of wavelength-dependent parameters as follows: J=

∫λ

λ2

dλ F(λ) σ(λ) ϕ(λ)

1

(1)

where F(λ) is the photon flux, σ(λ) is the molecular cross section, and ϕ(λ) 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 ϕ(λ).49−52 While there is a large body of literature regarding the gasphase 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 (λ > 300 nm). Literature of pyruvic acid photolysis Received: May 26, 2017 Revised: October 3, 2017 Published: October 16, 2017 8348

DOI: 10.1021/acs.jpca.7b05139 J. Phys. Chem. A 2017, 121, 8348−8358

Article

The Journal of Physical Chemistry A

allows for in situ tracking of chemical species in a highly regulated 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,

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 experiments range between ϕ = 0.21 and ϕ = 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 1 atm of air. To do so, they irradiated pyruvic acid with a solar simulator and examined the quantum yield and products as functions 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 2 orders of magnitude lower than those in the Reed Harris et al. 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

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 (gray 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 λmax.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.

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, 21st of June, Northern hemisphere). The chamber is continuously stirred with a 50 cm stainless-steel fan at the bottom of the chamber, resulting in mixing times of approximately 100 s. 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 of choosing 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 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

II. METHODS To investigate the photolysis of pyruvic acid under atmospheric conditions, we employ the CESAM simulation chamber, which 8349

DOI: 10.1021/acs.jpca.7b05139 J. Phys. Chem. A 2017, 121, 8348−8358

Article

The Journal of Physical Chemistry A

processes and quantified by fitting the pyruvic acid concentrations during the 30 min dark decay to eq 2.

glass bulb when at relatively high concentrations. At least 30 min 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.

[PA] = [PA]0 e−kdt

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 eq 3.

Table 1. Summary of Key Gas-Phase Photolysis Reactions in CESAM experiment pyruvic pyruvic pyruvic pyruvic pyruvic

acid acid acid acid acid

photolysis photolysis photolysis photolysis photolysis

pyruvic acid photolysis with cyclohexane

buffer gas air N2 trace [O2] 50% N2, 50% O2 N2 with 5% O2 spikea air

[O2] (ppm)

number of trials

3.5 × 105