Time-Resolved Measurements of Nitric Oxide, Nitrogen Dioxide, and

Subscriber access provided by TUFTS UNIV

Environmental Processes

Time-Resolved Measurements of Nitric Oxide, Nitrogen Dioxide, and Nitrous Acid in an Occupied New York Home Shan Zhou, Cora J Young, Trevor Casey VandenBoer, Shawn Finley Kowal, and Tara F Kahan Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b01792 • Publication Date (Web): 05 Jul 2018 Downloaded from http://pubs.acs.org on July 7, 2018

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 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 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.

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 35

Environmental Science & Technology

1

Time-Resolved Measurements of Nitric Oxide, Ni-

2

trogen Dioxide, and Nitrous Acid in an Occupied

3

New York Home

4 5

Shan Zhou1, Cora J. Young 2, Trevor C. VandenBoer 2, Shawn F. Kowal1, Tara F. Kahan1,*

6

1

Department of Chemistry, Syracuse University, Syracuse, NY, USA

7

2

Department of Chemistry, York University, Toronto, ON, Canada

8

*

Corresponding Author: Tara Kahan, [email protected], (315) 443-3285, Department of Chemis-

9

try, Syracuse University, Syracuse, NY 13244

10 11

KEYWORDS

12

Photochemistry, air exchange rate (AER), indoor air quality, oxidants, combustion, photon flux

13 14

ABSTRACT

15

Indoor oxidizing capacity in occupied residences is poorly understood. We made simultaneous

16

continuous time-resolved measurements of ozone (O3), nitric oxide (NO), nitrogen dioxide

17

(NO2), and nitrous acid (HONO) for two months in an occupied detached home with gas appli-

18

ances in Syracuse, New York. Indoor NO and HONO mixing ratios were higher than those out-

19

doors, whereas O3 was much lower (sub-ppbv) indoors. Cooking led to peak NO, NO2, and

20

HONO levels 20 - 100 times greater than background levels; HONO mixing ratios of up to 50

21

ppbv were measured. Our results suggest that many reported NO2 levels may have a large posi1 ACS Paragon Plus Environment


22

tive bias due to HONO interference. Nitrous acid, NO2, and NO were removed from indoor air

23

more rapidly than CO2, indicative of reactive removal processes or surface uptake. We measured

24

spectral irradiance from sunlight entering the residence through glass doors; hydroxyl radical

25

(OH) production rates of (0.8 – 10) × 107 molec cm-3 s-1 were calculated in sunlit areas due to

26

HONO photolysis, in some cases exceeding rates expected from ozone-alkene reactions. Steady

27

state nitrate radical (NO3) mixing ratios indoors were predicted to be lower than 1.65 × 104 molec

28

cm-3. This work will help constrain the temporal nature of oxidant concentrations in occupied

29

residences and will improve indoor chemistry models.

30 31

TOC Art

32 33

INTRODUCTION

34

Considering that people typically spend 90% of their time indoors (and 80% of that time in

35

residences), air quality in residences is of major concern for human health.1 In addition to direct

36

indoor sources and infiltration from the ambient atmosphere, harmful pollutants can be formed

37

indoors through in-situ secondary chemistry initiated by oxidants.2-5 Yet our understanding of

38

indoor oxidation processes, especially in occupied residences, lags substantially behind that for

39

outdoors.

2 ACS Paragon Plus Environment

Page 2 of 35

Page 3 of 35


40

The hydroxyl radical (OH) is the most important oxidant in the atmosphere. In indoor envi-

41

ronments, OH levels are expected to be much lower than those outdoors because the primary pro-

42

duction mechanism (ozone photolysis) does not occur.6 The primary indoor OH source is thought

43

to be non-photochemical reactions between ozone and alkenes.7-8 Recent studies suggest that pho-

44

tolysis of nitrous acid (HONO), formaldehyde, and hypochlorous acid (HOCl, which is generated

45

during cleaning events using commercial bleach solution) can also be a source of indoor OH.9-15

46

Nitrate radicals (NO3) have also been suggested to be potentially important oxidants indoors.4, 16

47

There are few indoor measurements of reactive species such as OH and NO3, and none in resi-

48

dences.17-18 Even levels of more stable oxidants are not well constrained in residences because

49

most measurements have been performed in non-residential buildings; lower air exchange rates

50

(AER) in residences may result in very different oxidant mixing ratios, as indoor levels can de-

51

pend strongly on AER. For example, ozone is thought to be the most important oxidant in indoor

52

environments, but some studies suggest that ozone levels in residences may be very low, likely

53

due to low AER, adsorption to surfaces, and titration by nitric oxide (NO).19-20

54

Nitrous acid is an emerging household oxidant that may be an important indoor photochemi-

55

cal OH source.10, 12, 14-15, 21 It can be emitted directly in indoor settings by combustion processes

56

such as burning candles, open fireplaces, gas stoves, and cigarette smoking.14, 22-23 Heterogeneous

57

reactions of nitrogen dioxide (NO2) also lead to the secondary formation of HONO.24 However,

58

there have been limited HONO measurements in residences and its indoor mixing ratio is subject

59

to high levels of uncertainty. Peak HONO mixing ratios (15 min time resolution) of up to 100

60

ppbv and 24 h averages of 40 ppbv were measured in controlled house experiments.25 Measure-

61

ments made in real homes yielded much lower mean (1 – 7 days) HONO mixing ratios ranging

62

from 0.8 to 11.3 ppbv.23, 26-28 In addition, most existing measurements of the important HONO

63

precursor NO2 have been made with passive samplers that are exposed to the indoor environment 3 ACS Paragon Plus Environment


64

for extended periods of time (1 to 14 days). Such extended sampling periods are unlikely to cap-

65

ture short-term peak levels and temporal changes in mixing ratios, which are important for under-

66

standing chemistry that occurs indoors. Furthermore, it has been shown that HONO interferes

67

with most NO2 measurement methods used indoors.29-30 This suggests that many reported indoor

68

NO2 values may be positively biased, further confounding the relative importance of NO2 and

69

HONO in residences.

70

In this study, we conduct time-resolved measurements of a suite of oxidants and oxidant pre-

71

cursors including O3, NO, NO2, and HONO in an occupied residence to better characterize their

72

mixing ratios, sources, and sinks. We also measure spectral irradiance from sunlight entering the

73

residence, and calculate OH production rates from HONO photolysis as well as steady-state NO3

74

concentrations under various conditions.

75 76

EXPERIMENTAL METHODS

77

Research House

78

From September 27 to November 20, 2017, continuous ambient sampling of indoor air was

79

made in a single-family dwelling in Syracuse, NY. Located in an urban residential area, it is a 4-

80

bedroom 1.5 story house (250 m3 volume) with an unfinished basement and a detached garage.

81

The house utilizes both natural ventilation (opening windows) and forced-air system (heating or

82

air-conditioning) using 100% recirculated air. Most floors are hardwood; the family room has a

83

mixture of carpet and bamboo flooring. Illumination is primarily provided by LED lights and

84

sunlight filtered through windows. Combustion appliances include a natural gas stove with four

85

burners and an oven. Water heater, heating, and air conditioning systems were also operated us-

86

ing commercial natural gas supplied to the house. During the sampling period, the house was oc-


Page 4 of 35

Page 5 of 35


87

cupied by two non-smoking adults, one child, and a dog. Daily activity was carried out normally

88

and a human activity log book was kept at the residential site.

89

Real-Time Measurements

90

A custom-built mobile analytical laboratory (Mobile Indoor Laboratory for Oxidative Spe-

91

cies; MILOS) was deployed for time-resolved measurements of O3, NO, NO2, HONO, carbon

92

dioxide (CO2), relative humidity (RH), and temperature (T). Ozone mixing ratios were measured

93

by an Ecotech Serinus 10 UV photometric analyser (accuracy of 0.5 ppbv or 2% of reading). Ox-

94

ides of nitrogen (NOx: NO and NO2) were measured using an Ecotech Serinus 40 O3-based chem-

95

iluminescence analyser (accuracy of 0.4 ppbv or 0.5% of reading). The NOx analyser was modi-

96

fied to allow simultaneous quantitative measurements of HONO by a difference technique.30 A

97

HONO inlet consisting of a gas denuder (GD) channel and a bypass (BP) channel was inserted

98

upstream of the NOx analyser. The GD removes gaseous acids including HONO using a sodium

99

carbonate-coated annular denuder (URG Corp.), but does not remove NO2. The denuder was pre-

100

pared by coating the interior with a saturated aqueous Na2CO3 solution (Fisher Scientific), shak-

101

ing gently, pouring out the excess solution, and drying with zero air for ~5 min. Air sampled

102

through the BP contains NO2 and HONO; both molecules are converted to NO by the molyb-

103

denum-catalyst.25, 30-32 A Teflon solenoid three-way valve enabled the sampling of GD (NO2 on-

104

ly) and BP (sum of NO2 and HONO, Σ(NO2+HONO)) air alternating at 5 min intervals. Distinct

105

measurements of NO2 and HONO (rather than the sum of the two species) were made after

106

October 31. CO2, RH, and temperature indoors were monitored with a TSI IAQ7545. Data were

107

logged at 30 s intervals for all measurements. For NO2 and HONO, during each 5-min interval,

108

the first 1 minute of data was excluded to reject data influenced by the instrument response time.

109

Data acquired over the remaining 4 minutes was averaged. MILOS was stationed in the living

110

room while its Teflon sampling inlet was positioned in the kitchen or living room at head height 5 ACS Paragon Plus Environment


111

(approximately 1.5 m above the floor). Outdoor mixing ratios were measured intermittently

112

throughout the sampling period by moving the inlet to the patio or front stoop, generally during

113

the day. Outdoor temperature (Tout) and relative humidity (RHout) at Syracuse Hancock Interna-

114

tional Airport, ~7.3 km north to the sampling site, were retrieved from the NOAA National Cen-

115

ters for Environmental Information’s Integrated Surface Hourly Database.33

116

Calibration and Quality Control

117

Given low nitric acid (HNO3) levels previously reported in residences,25, 34 its interference to

118

NO2 and HONO measurements is considered negligible in this study. The conversion efficiency

119

of HONO by the molybdenum catalyst was determined by introducing zero air containing

120

HONO, generated using a custom-built apparatus,35 into the NOx analyser and comparing the

121

level reported by the NOx analyser to the mixing ratio measured offline using ion chromatog-

122

raphy with conductivity detection. The conversion efficiency was determined to be to be (83.5 ±

123

3)%. Previous studies have shown that small amounts of NO2 can adsorb to annular denuders.36-37

124

We flowed NO2 ranging from 13 to 270 ppbv through the GD line and NO2 losses were found to

125

be (7.0 ± 0.45)%. Reported values were corrected to account for the conversion efficiency and

126

loss.

127

Calibrations and linearity checks were performed at the beginning, middle, and end of the

128

sampling period using a dilution calibrator and ozone generator (Ecotech GasCal 1100) and NO

129

cylinder (19.8 ppmv in N2, analytical uncertainty of 5%). The 30 s limit of detection (LOD) of the

130

analytes were determined as 3 times the standard deviations (3σ) of the corresponding signals in

131

zero air. Limits of detection were 0.63, 1.1, 1.3, and 1.3 ppbv for O3, NO, NO2, and

132

Σ(NO2+HONO), respectively. Higher LODs for NO, NO2, and HONO than for O3 are due to the

133

use of different digital filters. The 5 min LOD for HONO was calculated as 0.7 ppbv, 3 times the

134

standard deviation propagated from the subtraction of 5 min NO2 from Σ(NO2+HONO). The en6 ACS Paragon Plus Environment

Page 6 of 35

Page 7 of 35


135

durance of the GD was tested and the annular denuder was recoated on average every day (6

136

hours – 4 days depending on accessibility) to minimize the influence of denuder deactivation on

137

NO2 and HONO mixing ratios. Time periods potentially influenced by denuder deactivation were

138

identified based on decreasing HONO/NO2 ratios.

139

Data Analysis

140

The decay (removal) rate constants (k) of NO, NO2, HONO, and CO2 were estimated from

141

indoor measurements during periods affected by a dominant indoor source (e.g. cooking). Only

142

time periods when all species peaked at concentrations significantly higher than the background

143

values, followed by a non-source period, were considered in the calculations. Assuming constant

144

removal rates and background mixing ratios during the decay period, and assuming an even dis-

145

tribution of indoor concentrations of these species,

146

Ct = e-kt Co + Cb

147

where Ct is the indoor mixing ratio after time t, Co is the initial (peak) mixing ratio, and Cb is the

148

background mixing ratio of the analyte gas during the decay event. The decay rate constant was

149

determined using an exponential regression of the observed mixing ratio versus time for each

150

decay period. Decay rate constant error was determined as the standard error of k obtained from

151

the regression. Time periods for CO2 decay calculations were identified based on stringent crite-

152

ria to exclude potential impacts from occupant emissions (see the Supplementary Information for

153

details).

154

Light Measurements and Photolysis Rate Calculations

[1]

155

Wavelength-resolved spectra of sunlight entering the residence through double-paned patio

156

doors (4.8 m2) leading to a deck off of the south-facing family room were acquired using a cali-

157

brated Ocean Optics USB4000 spectrometer coupled to a 1 m fiber optic cable and a cosine cor-

158

rector. Full details of the measurement procedure and data analysis are described in Kowal et al.12 7 ACS Paragon Plus Environment


Page 8 of 35

159

In brief, 3 background-subtracted spectral irradiances of sunlight indoors were measured with 1.5

160

ms integration. Average spectral irradiances were calculated, and photon fluxes (F) were estimat-

161

ed. These measurements were acquired on September 21, 2017 at approximately noon; the sky

162

was cloudless and the solar zenith angle was 41.4°.38

163

HONO photolysis rate constants (JHONO) were calculated as described previously using meas-

164

ured photon fluxes and reported absorption cross sections (σ) and photolysis quantum yields

165

(ϕ).12

166 ఒ೑

‫ܬ‬ுைேை = න ߪுைேை ሺߣሻ߶ுைேை ሺߣሻ‫ܨ‬ఒ ݀ߣ

[2]

ఒ೔

167

Rate constants were then multiplied by measured indoor HONO concentrations to calculate OH

168

production rates.

ܴܽ‫ܬ = ݁ݐ‬ுைேை ሾ‫ܱܱܰܪ‬ሿ

[3]

169 170

RESULTS and DISCUSSION

171

Outdoor temperature (Tout) ranged from -8.3 to 32.2 ºC and RHout ranged from 29 to 100%,

172

with means (± 1σ) of 9.8 ± 7.6 ºC and 78.4 ± 17.2%, respectively. Precipitation occurred 22% of

173

the time throughout the study period as rain or snow. The indoor environment was generally

174

warmer (average temperature of 20.5 ± 1.2 ºC) and drier (average RH of 44.5 ± 5.7%), with little

175

temporal variation in T and RH. Figure 1 shows a sector of the real-time temperature, RH, and

176

gas species mixing ratios measured indoors during the sampling period. The time series over the

177

entire sampling period are shown in Fig. S1. Mixing ratios of all species varied dynamically with

178

high amplitudes (Fig. S1). Ozone levels indoors were often below the detection limit (58% of the

179

observation period), whereas outdoor levels of up to 34.2 ppbv were measured. 8 ACS Paragon Plus Environment

Page 9 of 35

180


Mixing Ratios under Background Conditions

181

Indoor background oxidant mixing ratios during periods without emission sources (i.e., doors

182

closed, no cooking or heating events) were examined to obtain a baseline of mixing ratios indoors.

183

Nitric oxide mixing ratios were much higher indoors than outdoors, with an average (± 1σ) in-

184

door background mixing ratio of 4.0 ± 2.5 ppbv. Most (70%) outdoor NO levels were below the

185

LOD. Using the LOD as an upper limit for the NO mixing ratio under these conditions, we esti-

186

mated a lower limit for the background indoor/outdoor (I/O) ratio of 2.1. Comparable average

187

levels of the sum of NO2 and HONO (Σ(NO2+HONO)) were observed under indoor background

188

conditions and outdoors (5.4 vs 5.2 ppbv). The average (± 1σ) indoor background NO2 and

189

HONO mixing ratios (acquired after October 31, excluding periods with potential influence from

190

denuder passivation) were 2.0 ± 0.68 and 4.3 ± 2.2 ppbv, respectively. Note that we had very lim-

191

ited outdoor distinct measurements of NO2 and HONO; their outdoor levels are not reported. Alt-

192

hough outdoor HONO mixing ratios can be variable, they are typically on the order of 1 – 2 ppbv

193

at night and sub-ppbv at midday.39 This is suggestive of higher HONO levels indoors under

194

background conditions than outdoors, in agreement with previous studies.23,

195

above its detection limit, its average background mixing ratio was 1.0 ppbv, significantly lower

196

than the outdoor levels measured (mean of 12.6 ppbv, Fig. S2).

28

When O3 was

197

The observed complex relationship between the indoor background and outdoor oxidant lev-

198

els reflect different source and sink strengths. Direct emission from traffic is the dominant source

199

of NOx in ambient urban environments. As shown in Fig. S2, outdoor NO mixing ratios peaked at

200

6 – 8 EST (12 ppbv) and 16 – 20 EST (6 ppbv). However, during the daytime with high ambient

201

ozone levels (~20 ppbv, Fig. S2), which convert NO to NO2, ambient NO levels ranged from 2

202

ppbv to below the LOD. A similar, yet less pronounced, diurnal profile was also observed for

203

ambient Σ(NO2+HONO) (Fig. S2), likely corresponding to rush-hour traffic emissions of NOx. 9 ACS Paragon Plus Environment


204

Nitrous acid on the other hand is photolyzed rapidly by sunlight, so its ambient outdoor concen-

205

tration is quite low during the day, and measured outdoor Σ(NO2+HONO) levels during the day

206

are likely almost entirely due to NO2.

207

Significantly higher NO and HONO mixing ratios indoors compared to outdoors suggest that

208

transport of outdoor NO and HONO by air infiltration were likely not the primary indoor sources

209

of these species. This is supported by the absence of rush-hour peaks for indoor levels. Combus-

210

tion processes indoors, e.g. biomass, incense, and candle burning, gas cooking, and smoking,

211

have been shown as major direct sources of NOx and HONO in indoor environments.14, 40-42 Gas

212

cooking is likely the main indoor HONO source in the current study. Elevated concentrations of

213

NO and HONO throughout the study period (including in the absence of combustion) are likely

214

due to weaker sinks for these species indoors; low O3 in the case of NO and reduced photolysis

215

indoors in the case of HONO. Since surface area to volume ratios are generally greater indoors

216

than outdoors, heterogeneous reactions of NO2 may be more important indoors, and may have

217

contributed to higher indoor background HONO levels.43

218

In addition to indoor sources (e.g. combustion), infiltration of NO2 from outdoor air also con-

219

tributes to indoor NO2 levels.40, 44 Transport of ambient O3 is the dominant source of indoor O3,

220

which often is present at mixing ratios 10 – 80% of ambient outdoor values, depending on the

221

AER.45 Our observations of lower background mixing ratios of NO2 and O3 indoors are consistent

222

with outdoor-to-indoor transport of these species. Weak indoor production of NO2 from the reac-

223

tion of NO and O3 due to low levels of indoor O3 may account for low background indoor NO2

224

mixing ratios. Strong sinks – e.g. NO2 deposition to indoor surfaces and O3 titration by NO – may

225

have also contributed to lower NO2 and O3 mixing ratios indoors compared to outdoors in the

226

absence of direct sources.

227

Mixing Ratios Influenced by Cooking 10 ACS Paragon Plus Environment

Page 10 of 35

Page 11 of 35


228

Indoor mixing ratios of NO frequently exceeded 100 ppbv, and occasionally reached 340

229

ppbv, 2 orders of magnitude higher than background levels (Fig. S1c). Dramatically higher in-

230

door mixing ratios of NO2 and HONO than the background values were also observed, peaking at

231

59 and 50 ppbv (Fig. S1c), which correspond to mixing ratios approximately 30 and 12 times

232

higher than background levels, respectively. These peak episodes correspond to cooking events,

233

as suggested in Fig. 1c. The average diurnal variations of indoor NO, NO2, and HONO during

234

this study also showed three recurring spikes corresponding to meal times at ~ 8 – 9, 12 – 14, and

235

16 – 19 EST with the highest levels occurring around supper time (Fig. S3). These observations

236

together suggest that direct emission from gas cooking was the major source of NOx and HONO

237

in this residence. Reactions involving NO2 likely also contribute to HONO formation during

238

cooking events.22 Our measurements do not have the temporal resolution required to distinguish

239

between primary and secondary HONO formation.

240

Statistics of maximum 5 min peak concentrations during cooking events are presented in

241

Fig. 2 and Table 1; means (± 1σ) were 113.2 ± 72.7, 17.3 ± 12.8, and 19.5 ± 10.5 ppbv for NO,

242

NO2, and HONO, respectively. These values are significantly higher than the mixing ratios meas-

243

ured under background conditions as well as those measured outdoors. Peak NO mixing ratios are

244

in agreement with those observed in other residences (~40 – 300 ppbv) but significantly lower

245

than those measured during controlled cooking experiments (100 – 2000 ppbv).22, 46-48 The peak

246

NO2 levels in this study were generally at the low end of the concentration range previously re-

247

ported (~20 – 550 ppbv),22,

248

ppbv).22

46, 48

whereas peak HONO levels were on the high end (3 – 16

249

As shown in Fig. 1c, when cooking appliances were in use, mixing ratios of combustion by-

250

products (NO, NO2, and HONO) rapidly rose. The rise time of the signal corresponded to the

251

time the appliance was on (i.e. cooking duration) which varied between 4 and 86 minutes in this 11 ACS Paragon Plus Environment


Page 12 of 35

252

field study. After the appliance was turned off, the levels rapidly reached peak levels and slowly

253

decayed away. To investigate possible factors that affect the peak mixing ratios, 74 cooking

254

events were analyzed. Positive correlations between the measured peak mixing ratios of these

255

species and cooking duration were observed (r2 = 0.42 - 0.72; Fig. 3), suggesting that operating

256

time of cooking appliances could explain 42% – 72% of the variability in the peak levels of NOx

257

and HONO. In addition, mixing ratios measured when cooking with individual burners at differ-

258

ent positions overlaps with oven operation, which suggests that appliance type overall exerted

259

little effect on the emissions. Furthermore, cooking events using multiple appliances fall tightly

260

along the same trend, indicative of negligible effect of additional appliances on the mixing ratios

261

observed. In previous controlled experiments, peak NOx levels depended on the number of gas

262

elements used.47 The reason for the lack of influence of combustion elements in this study is un-

263

certain. There may be other variables that are not accounted for such as the power of the appli-

264

ances and the setting during individual cooking events (e.g. oven temperature, gas mark on stove).

265

A ventilated range hood was used during some cooking events (N = 15) during the study pe-

266

riod. Recent assessments of range hood performance in the US indicate wide variations across

267

hoods and airflow and burner positions. Several studies reported that using overhead kitchen ex-

268

haust fans was associated with reduced NOx levels indoors.42,

269

Σ(NO2+HONO) production rates during cooking events were lower when the fan was in opera-

270

tion than when it wasn’t: 2.4 vs. 4.2 ppbv min-1 for NO, and 0.97 vs. 1.32 ppbv min-1 for

271

Σ(NO2+HONO) (Fig. 3a,3b). These observations suggest that the vented exhaust fan in the kitch-

272

en reduced NO and the sum of NO2 and HONO mixing ratios by approximately 43% and 27%,

273

respectively.

274

Indoor Decay/Removal Rate Constants


46, 49-50

In this study, NO and

Page 13 of 35


275

For many of the cooking events, there was a clear exponential decay of the combustion by-

276

product mixing ratios after the appliance was shut off. The decay/removal rate constants of the

277

combustion by-products were estimated for 55 uninterrupted episodes after cooking. The decay

278

rate constant of CO2 can potentially be referred to as the indoor-outdoor air exchange rate (AER).

279

Estimated AER from 19 selected cooking events ranged from 0.1 – 1.6 hr-1 with a mean value of

280

0.62 (± 0.36) hr-1. This is consistent with reported AER for US residential homes (median/mean

281

of 0.3 – 1.12 hr-1).51-54 Decay rate constants of NO, Σ(NO2+HONO), NO2, and HONO were di-

282

rectly compared to AER; all were significantly higher (Fig. 4). This suggests that mixing ratios of

283

all these species decrease more rapidly than can be accounted for by air exchange alone and that

284

reactive processes (gas phase reactions and surface uptake) in the house make a significant con-

285

tribution to their removal.

286

The NO decay rate constant, kNO, was on average 0.92 ± 0.37 hr-1. A plot of kNO vs. kCO2

287

yielded an intercept of 0.35 h-1 (Fig. 4a), which can be considered the approximate reactive re-

288

moval rate constant of NO. The reactive NO removal rate constant can also be estimated as the

289

difference between kNO and the AER; rate constants determined by this method ranged from 0.12

290

to 0.79 hr-1 and averaged 0.33 hr-1. Although we are not certain of NO removal mechanisms other

291

than air exchange, its reaction with O3 infiltrated indoors is one possibility. As NO mixing ratios

292

indoors were high, it possibly quickly reacted with O3 and resulted in stable low levels of ozone

293

whereas high NO levels were sustained for hours. The lifetime of NO, i.e. the time it took for the

294

mixing ratio to decrease to e-1 of its initial value, varied between 30 – 228 min.

295

Reactive decay rate constants for NO2 and HONO were also determined via the difference

296

method. We note that limited distinct measurements of NO2 and HONO resulted in considerable

297

scatter in the data, especially for kNO2 (Fig. 4c, 4d). The performance of the denuder may also

298

have contributed to the scatter, as the denuder was sometimes deactivated quickly during cooking 13 ACS Paragon Plus Environment


299

events when very high HONO concentrations were generated. This denuder passivation could

300

result in overestimations of kHONO and underestimations of kNO2 in this study. However, since the

301

decay rate constants calculated during periods with potential influence from denuder deactivation

302

match the trend of those without (Fig. 4c, 4d), the uncertainty introduced by this potential artefact

303

is minimal, and we are confident that the reactive HONO loss observed in this residence is real.

304

On average, kNO2 was 1.54 ± 0.52 hr-1. The reactive NO2 removal rate constant determined

305

from the difference method was 0.74 (± 0.35) hr-1, larger than the mean AER. This suggests that

306

reactive removal was more important to NO2 lifetimes indoors than air exchange. The mean reac-

307

tive NO2 removal rate constant determined in this house is similar to the mean reactive rate con-

308

stants reported in controlled experiments in residences.44, 55 Laboratory experiments have demon-

309

strated that heterogeneous reactions play an important role in removing NO2 from the air in in-

310

door environments due to the elevated surface area to volume ratios.24, 56-59 The reactive NO2 re-

311

moval rate constants determined in this study are comparable to those determined through cham-

312

ber experiments in the presence of carpets under various conditions,43 further suggesting that NO2

313

surface chemistry may be the dominant reactive NO2 removal mechanism indoors. Photolysis is a

314

significant sink for NO2 outdoors, but photolysis rates indoors are much lower due to the lack of

315

high energy photons. An NO2 photolysis rate constant of 8.49 × 10-4 s-1 was determined for sunlit

316

indoor regions based on our spectral irradiance measurements (see discussion below). This results

317

in a photochemical loss rate of 1.70 × 10-3 ppb s-1 (4.57 × 107 molec cm-3 s-1) at background NO2

318

levels (2.0 ppb) and 5.01 × 10-2 ppb s-1 (1.35 × 109 molec cm-3 s-1) at peak NO2 concentrations (59

319

ppb). However, photolysis likely did not significantly affect our measured kNO2 because only

320

small volumes of air in the residence were in direct sunlight at any given time, and the sampling

321

inlet was in shaded regions of the house throughout the campaign. Indeed, NO2 removal rate con-


Page 14 of 35

Page 15 of 35


322

stants did not show dependence on time of day (or availability of indoor sunlight) in this study

323

(Fig. S4).

324

The average decay rate constant for HONO (kHONO) was 1.68 (± 0.55) hr-1, with a mean reac-

325

tive removal rate constant of 0.58 (0.16 – 1.03) hr-1. Previous experiments have reported negligi-

326

ble reactive HONO removal constants indoors.43-44 One possible HONO sink is uptake to surfac-

327

es.57 Although the specific mechanism is not clear, heterogeneous reaction of HONO with sur-

328

face-deposited nicotine from tobacco smoke highlights the possible reactive sinks on indoor inter-

329

faces involving HONO and organics.2 Thermodynamic partitioning of HONO into indoor materi-

330

als that act as HONO reservoirs could be another possible sink.43 As with NO2, photolysis is the

331

dominant sink for HONO outdoors, but this is likely not the case indoors, for the same reasons

332

outlined above with respect to NO2 (Fig. S4).

333

Photochemical OH Production

334

HONO photolysis has been predicted to be an important source of indoor OH; this has been

335

experimentally verified in sunlit classrooms.10 Indoor OH production rates from HONO in resi-

336

dences have not been reported. We measured spectral irradiance from sunlight entering the resi-

337

dence through glass patio doors and calculated OH production rates at measured HONO levels.

338

We do not present steady state OH mixing ratios because mixing ratios of OH sinks (especially

339

VOCs) are poorly constrained in residences, and were not measured in this study. Estimated

340

steady state OH mixing ratios would therefore be associated with very large uncertainty. Hydrox-

341

yl radical production rates, which are much better constrained in this study, can be incorporated

342

into indoor chemistry models to estimate steady state OH mixing ratios under a range of condi-

343

tions. Figure 5 shows the noon maximum wavelength-resolved photon fluxes inside this resi-

344

dence through the glass patio doors. A spectrum with the glass doors open is shown for compari-



Page 16 of 35

345

son. Light is completely attenuated by the glass at wavelengths shorter than ~340 nm, but a sig-

346

nificant amount of light is transmitted in the actinic region (

Time-Resolved Measurements of Nitric Oxide, Nitrogen Dioxide, and

Recommend Documents