Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2021 Nov 24;55(23):15969–15979. doi: 10.1021/acs.est.1c01368

Chemical Investigation of Household Solid Fuel Use and Outdoor Air Pollution Contributions to Personal PM2.5 Exposures

Alexandra Lai †, Martha Lee ‡, Ellison Carter §, Queenie Chan ∥, Paul Elliott ∥, Majid Ezzati ∥, Frank Kelly ⊥, Li Yan ⊥, Yangfeng Wu #, Xudong Yang , Liancheng Zhao ∇, Jill Baumgartner ‡,○, James J Schauer †,⧫,*
PMCID: PMC8655976  PMID: 34817986

Abstract

graphic file with name es1c01368_0005.jpg

In communities with household solid fuel use, transitioning to clean stoves/fuels often results in only moderate reductions in fine particulate matter (PM2.5) exposures; the chemical composition of those exposures may help explain why. We collected personal exposure (men and women) and outdoor PM2.5 samples in villages in three Chinese provinces (Shanxi, Beijing, and Guangxi) and measured chemical components, including water-soluble organic carbon (WSOC), ions, elements, and organic tracers. Source contributions from chemical mass balance modeling (biomass burning, coal combustion, vehicles, dust, and secondary inorganic aerosol) were similar between outdoor and personal PM2.5 samples. Principal component analysis of organic and inorganic components identified analogous sources, including a regional ambient source. Chemical components of PM2.5 exposures did not differ significantly by gender. Participants using coal had higher personal/outdoor (P/O) ratios of coal combustion tracers (picene, sulfate, As, and Pb) than those not using coal, but no such trend was observed for biomass burning tracers (levoglucosan, K+, WSOC). Picene and most levoglucosan P/O ratios exceeded 1 even among participants not using coal and biomass, respectively, indicating substantial indirect exposure to solid fuel emissions from other homes. Contributions of community-level emissions to exposures suggest that meaningful exposure reductions will likely require extensive fuel use changes within communities.

Keywords: personal exposure, household air pollution, solid fuels, molecular tracers, personal/outdoor ratio, PM2.5, source apportionment

Short abstract

Exposure to household air pollution depends not only on an individual’s own fuel use but also on outdoor air pollution and their neighbors’ fuel use.

Introduction

Approximately 3.6 billion people worldwide and 500 million people in China primarily cook with stoves that burn solid fuels such as coal and biomass.1 Household solid fuel burning emits high levels of pollutants including fine particulate matter (PM2.5) into homes and communities, contributing to high levels of exposure to household air pollution (HAP). Cooking area concentrations of PM2.5 can exceed the World Health Organization’s 24 h guideline level of 25 μg/m3 by an order of magnitude or more, contributing to personal PM2.5 exposures both directly within the household and indirectly via contributions to ambient PM2.5.2 Exposure to PM2.5 is associated with an increased risk of adverse health outcomes across the life course, including cardio-respiratory diseases,3,4 low birth weight,5 and diabetes,6 thereby motivating the development of diverse strategies to reduce PM2.5 and improve public health.1

Identifying the sources contributing to PM2.5 exposures is essential for developing more effective and appropriately targeted mitigation strategies. Transitioning from solid fuels to less-polluting household energy systems can significantly lower HAP;7 however, field studies show that even with use of cleaner fuels, PM2.5 exposures often remain above the WHO guideline.2,7−9 Achieving and sustaining low exposures to PM2.5 among populations using solid fuels requires both household-level changes in energy practices and community- and regional-level mitigation of external PM sources, including industry, traffic, agricultural burning, and HAP from neighboring households.9

Personal monitoring of airborne PM is intended to account for variability in PM exposures that are attributable to variability in individual behaviors and living conditions.7 Technological developments in the size, weight, noise, and cost of personal air samplers contributed to an increase in direct sampling of personal PM2.5 exposures in settings with HAP over the last decade.10−29 These studies showed high within-subject variability in exposures and relatively small differences between clean fuel and solid fuel users, which could be driven by diverse factors including outdoor air pollution, unidentified sources, variability in emissions based on fuel characteristics or stove performance, and solid fuel use in nearby homes.2,9 Chemical analysis of PM2.5 samples could provide information about the relative impacts of these factors on exposure reductions or lack thereof; however, only a small handful of these studies included chemical analysis10−12,21−24,30 and even fewer measured source-specific tracers (e.g., levoglucosan).10,23,24,30

Using data from the INTERMAP China Prospective (ICP) study, a geographically diverse cohort of men and women with different fuel use patterns and outdoor source contributors, we investigated outdoor and personal exposures to PM2.5, its chemical components, and source contributions. Our specific objectives were to (1) measure and compare the chemical composition and sources of PM2.5 exposures by region, gender, and predominant fuel use patterns and (2) differentiate between direct and indirect (not from that individual’s home) exposures to PM2.5 emitted by solid fuel combustion. In this study, we employed chemical analysis and source apportionment techniques paralleling our previous analyses of PM2.5 exposures in rural China24 but with additional dimensions: diversity of fuel types (i.e., coal, biomass, and clean fuels such as gas and electricity) and use, inclusion of both men and women, and multiple study sites. Understanding the composition and sources of PM2.5 exposures is important for refining exposure assessments, which in turn are essential for designing effective interventions and mitigation strategies.7,9,31

Methods

Study Design

The PM2.5 samples analyzed in this study were collected for the ICP study, a longitudinal study in three provinces in northern (Beijing and Shanxi) and southern China (Guangxi) established to identify environmental and dietary risk factors for cardiovascular disease. The ICP study design, participants, and data collection have been described elsewhere.32−34 Briefly, 787 individuals aged 40–79 were enrolled in the ICP study (258 in Beijing, 290 in Shanxi, and 239 in Guangxi) in 2015 and 2016. All three study sites were rural but becoming more peri-urban, and households used biomass and/or liquefied petroleum gas (LPG) for cooking. At the Shanxi and Beijing sites, coal was also used for cooking and/or indoor space heating in winter. Through a visual questionnaire, participants recorded their use of different stove and fuel types for cooking and, where applicable, heating. Study sites are described in more detail in Section S1 and fuel use data collection and trends in Section S2 and Figure S2. Participants provided written informed consent. Study protocols were approved by ethical review boards at all investigator institutions.

Air Pollution Sampling

Personal exposure PM2.5 samples were collected for 24-h sampling periods on 37 mm polytetrafluoroethylene (PTFE) membrane filters (Zefluor, Pall Corporation, USA), using Harvard personal exposure monitors (PEMs) (Mesa Laboratories, USA).35 PEMs, which include impactors for PM2.5 size selection, were attached to personal sampling pumps (Apex Pro and TUFF, Casella Inc.; USA) operating at 1.8 L/min. Participants carried PEMs with them in a waistpack except while bathing or sleeping, when they kept the PEM nearby and off the ground. Pedometers were also included in waistpacks to monitor compliance, where samples were flagged as potentially noncompliant and excluded from these analyses if they had step counts of <500 steps.32

Outdoor PM2.5 samples were also collected for 24-h intervals on 37 mm Zefluor PTFE filters, placed inside either PEMs or cassettes paired with cyclones (Mesa Laboratories, USA) attached downstream from sampling pumps operating at 1.8 or 3.5 L/min, respectively. Outdoor monitors were positioned at least 4 m from the ground in a location that was central to each study village (Shanxi; n = 6 villages) or the group of study villages (Beijing and Guangxi), at least 30 m from a household chimney, and at least 100 m from other known PM2.5 sources including local industry and major roadways. We collected 10% field blanks. Pump flow rates were checked at the start and end of each sampling period against a rotameter, which was calibrated at the start and end of each field campaign using a primary gas flow meter as described in Lee et al. (2021).32

Filters were weighed before and after sampling using an automatic weighing system (Mettler-Toledo, USA) in a temperature- and humidity-controlled environment.24,36 The weighing room maintained a 24-h average temperature of 20–23 °C (standard deviation ≤ 2 °C) and a 24-h average relative humidity of 30–40% (standard deviation ≤ 5%). Filters were equilibrated for 24 hours before being weighed, and the reported weights were the average of duplicate measurements with differences <5 μg. Detailed information on air pollution measurement and related quality control procedures are published elsewhere.32

Sample Compositing for Chemical Analysis

For chemical analyses, we selected a subset of 221 PM2.5 exposure samples and 26 outdoor PM2.5 samples (out of 2073 personal and 48 outdoor PM2.5 samples). Within this subset, each exposure sample had a corresponding outdoor sample collected on the same day at the same study site (Figure S1). Fewer outdoor samples were required because personal sampling was conducted with multiple participants in parallel: for each outdoor sample, several personal exposure samples were collected concurrently at that site. All selected samples were collected in winter (November/December 2015 and 2016), and personal exposure samples were from current nonsmokers. Each filter was divided into sections for different chemical analyses. To obtain sufficient mass, filter sections were combined into 43 personal exposure and 5 outdoor composites (Figure S1, Table 1). Correspondence between outdoor and personal was preserved in the compositing scheme. For example, the Guangxi outdoor composite had samples collected on seven sampling dates, and each of the Guangxi personal exposure composites had samples from those same seven dates. Composite fuel use was defined based on whether each fuel was used in any of the individual samples; samples with similar fuel use were grouped together in composites when possible. Sample selection, compositing, and fuel use are described in further detail in Section S2 and Figure S2.

Table 1. Summary of Samples Selected for Chemical Analysis Composites [Composites (Individual Samples)]a.

  sample type
study site personal exposure, women personal exposure, men outdoor
Guangxi 5 (35) 2 (14) 1 (7)
Beijing 11 (55) 9 (45) 2 (10)
Shanxi 11 (49) 5 (23) 2 (9)
Open in a new tab
a

Data are presented as N (M) where M is the number of individual samples that were combined to form N composites for chemical analysis.

Eleven field blank composites (of 58 individual field blank filters) were also analyzed, corresponding to sample sites and the number of individual filters per composite. The majority of the field blanks were also from winter sampling seasons, with a small number of summer field blanks included when winter field blanks were not available. Seasonal differences in field blank filter masses were not statistically significant and were comparable to between-site differences, which were accounted for in the blank compositing scheme (Figure S3).

Chemical Analyses

The following chemical species were measured: water-soluble organic carbon (WSOC), water-soluble ions, trace and major elements, and organic molecular markers. Samples were divided into one half and two quarter sections before compositing for separate chemical analyses (Figure S1). Chemical analyses were conducted according to previously published methods24,37−39 and summarized below.

Composites of quarter filter sections for WSOC/ion analysis were extracted in 15 mL of ultrapure water on a shaker table for 6 hours and then filtered through a polypropylene syringe filter (0.45 μm pore; Whatman). A total organic carbon analyzer (M9 TOC Analyzer, Sievers/GE) was used to measure WSOC, and seven water-soluble ions (SO42–, NO3–, Cl–, Na+, NH4+, K+, and Ca2+) were measured using ion chromatography (IC) (Dionex ICS 1100/2100, Thermo Fisher Scientific).

After microwave digestion of elemental analysis composites (quarter sections) in ultrapure nitric acid, total concentrations of 51 elements (Table S1) were quantified using magnetic sector ICPMS (Thermo-Finnigan Element 2, Thermo Fisher Scientific).

Organic molecular marker composites (half sections) were extracted by Soxhlet in 50:50 dichloromethane/acetone and then concentrated with a rotary evaporator and ultrapure nitrogen gas. Organic compound standards were added to each sample,38 and each batch of ∼10–12 samples included a laboratory blank and a blank spiked with either standards or 10 mg of standard reference material (NIST SRM 1649, Urban Dust), extracted using the same procedure as the samples. Compound classes analyzed included polycyclic aromatic hydrocarbons (PAHs), hopanes and steranes (markers of fossil fuel combustion), n-alkanes, n-alkanoic acids, aliphatic and aromatic acids, and levoglucosan.24

Data Analysis

Source contributions to PM2.5 were estimated using chemical mass balance (CMB) to apportion primary source contributions and mass reconstruction for secondary inorganic aerosol (SIA) and dust. The CMB model calculates source contributions to measured concentrations of chemical species in PM2.5 by solving a system of linear equations using an effective variance-weighted least-squares method.40 We input the following source tracers: levoglucosan, picene, 17α(H)-21β(H)-30-norhopane, 17α(H)-21β(H)-hopane, sulfate, nitrate, and ammonium. The sources considered were residential wood burning and bituminous coal combustion, sampled in typical Chinese cooking and heating stoves,41,42 and mixed diesel/gasoline vehicle emissions sampled from Zhujiang Tunnel in Guangzhou, China.43 Chemical profiles of these sources have been published previously.24,41R2 values were above 0.98 and χ2 (chi-square) values were below 1.9, in agreement with recommended parameters for the model (R2 > 0.8 and χ2 < 4).44 SIA was defined as the sum of sulfate, nitrate, and ammonium after correcting for primary source contributions to each.24 Dust mass was calculated as the sum of oxides of major crustal elements: 2.14*[Si] + 1.89*[Al] + 1.43*[Fe] + 1.40*[Ca] + 1.66*[Mg] + 1.67*[Ti] + 1.58*[Mn].45,46 Si concentrations were estimated using Al concentrations and typical crustal ratios.47

Principal component analysis (PCA) was conducted using the “psych” package (version 2.0.7) in the statistical software R (version 4.0.2) (https://www.r-project.org). Varimax rotation was applied to maximize factor loadings, and five principal components were selected on the basis of eigenvalues (all eigenvalues >1), variance explained (each >10% of total variance), and interpretability.

Results and Discussion

PM2.5 Concentrations and Chemical Components

Results for all personal PM2.5 exposure and outdoor PM2.5 measurements are presented and discussed in Lee et al. (2021).32 Here, for context and comparison with the overall data set, we briefly summarize the PM2.5 mass data for the samples selected for chemical analyses. Outdoor PM2.5 levels were the highest for the Beijing samples (arithmetic mean ± standard error: 91 ± 24 μg/m3), while Shanxi and Guangxi had similar outdoor PM2.5 (48 ± 16 and 56 ± 9 μg/m3, respectively) (Figure 1). Personal PM2.5 exposures were similar between men and women at each site and were higher for Shanxi (men: 168 ± 19 and women: 158 ± 17 μg/m3) and Beijing samples (men: 157 ± 13 and women: 171 ± 13 μg/m3) than for Guangxi samples (men: 71 ± 7 and women: 68 ± 6 μg/m3) (Figure S4). There are two likely reasons why exposures were lower among the Guangxi samples: the selected samples were collected in winter, when indoor space heating was necessary at the Shanxi and Beijing study sites but not Guangxi, and there was no coal use in the Guangxi households (Figure S2). Mass concentrations are representative of the samples overall: within each study site and sample type (outdoor and personal), median PM2.5 in the chemical analysis subset was within 25% of the median of all samples (Figure S5).

Figure 1.

Figure 1

Open in a new tab

Average source contributions to ambient PM2.5 and personal PM2.5 exposures at each study site, normalized to PM2.5 mass. “Unidentified sources” refer to the difference between gravimetric PM2.5 and the sum of identified source contributions from the CMB model. Gray points and error bars represent the average and standard error, respectively, of PM2.5 concentrations for each site/sample combination and are plotted according to the gray secondary y-axis on the right.

Among the samples selected for chemical analysis, average outdoor and personal PM2.5 exposures in Beijing and personal exposures in Shanxi exceeded the China National PM2.5 standard (75 μg/m3).48 All outdoor and personal exposure composites had average PM2.5 concentrations exceeding the World Health Organization’s 24 h PM2.5 guideline (Figure S4).49 Although indoor solid fuel use is highlighted as a major contributor to PM2.5 exposures, ambient air pollution in rural areas is not negligible.50,51 Throughout China between 2001 and 2015, average PM2.5 levels in rural areas were only about 5–20 μg/m3 lower (up to ∼30%) than in the corresponding urban areas of each province.52

Concentrations of PAHs, elements, ions, and WSOC are summarized (mean ± SD) in Table S1. Concentrations of K, emitted from biomass burning as K+,53 in our study sites (mean: 857, 1267, and 677 ng/m3 in Guangxi, Beijing, and Shanxi, respectively) were similar to those measured in a recent study of chemically speciated PM2.5 exposures among rural Chinese adults using both biomass and clean fuel stoves (mean: 691 and 1678 ng/m3 in agricultural and nomadic villages, respectively).12 By comparison, exposures to elements associated with coal combustion54 in our northern study sites (Beijing: 7.9 ng/m3 As and 96 ng/m3 Pb; Shanxi: 10.3 ng/m3 As and 113 ng/m3 Pb) were much higher than those reported by Ye et al. (As was not detected in most exposures, <5 ng/m3 Pb at both agricultural and nomadic villages), consistent with no documented coal use at their study site.12

Average concentrations of benzo[a]pyrene, a carcinogenic and commonly measured PAH, were 1.2, 20.1, and 12.8 ng/m3 in Guangxi, Beijing, and Shanxi, respectively. These levels were on the lower end of reported mean benzo[a]pyrene exposures (in ng/m3) among biomass (∼18–9612,22,23,30), coal (15–8321−23), and mixed (12–2021,23) biomass/coal users in rural China but exceeded urban benzo[a]pyrene exposures measured in Hong Kong, where study participants cooked only with gas or electricity (mean: 0.02 ng/m3).55

Sources of Outdoor and Personal PM2.5

Coal combustion was the dominant primary source in Shanxi outdoor samples, contributing 16–41% of outdoor PM2.5, whereas the biomass burning contribution was smaller (5%) (Figure 1, Table S2). Coal and biomass contributions were similar in Beijing (8–11 and 6–13%, respectively). In Guangxi, biomass contributed approximately 8% of outdoor PM2.5, and no coal signature was detected. Motor vehicles were among the largest contributors to outdoor PM2.5 in Beijing (12%), which was similar to the solid fuel contribution, whereas vehicles comprised very little outdoor PM2.5 in Guangxi (2%) and were not identified as a source in Shanxi. Outdoor dust levels were the highest in Guangxi, in terms of both absolute (5.2 μg/m3) and mass-normalized concentrations (9%), and negligible in Beijing and Shanxi. SIA was the largest component of outdoor PM2.5 at all sites. Unidentified sources (mass not apportioned by the CMB model) include secondary organic aerosol (SOA) formation,56,57 which is supported by the results of the PCA discussed below. On a mass basis, outdoor PM2.5 sampled in this study was highly correlated with measurements from nearby government air-monitoring stations, indicating that these samples are representative of regional PM2.5.32

On a regional scale, source apportionment studies in China report that household solid fuel combustion is a major source of ambient PM2.5.58,59 Biomass and coal combustion have also been identified as major sources of volatile organic compounds (i.e., potential SOA precursors) in ambient PM,60−65 including rural or background sites in China.60,63−65 Solid fuel combustion contributed 8–46% of outdoor PM2.5 in this study in terms of primary emissions and thus likely also contributed to SOA.

Sources and components of personal exposure PM2.5 resembled outdoor PM2.5 in terms of mass fractions, suggesting substantial influence, and were generally higher than outdoor in terms of absolute concentrations, with the exception of SIA (Figure 1). Biomass burning and dust were the two major personal exposure sources in Guangxi, contributing 8–22 and 7–29% of PM2.5 exposures, respectively. Although picene was not detected in any of the Guangxi samples and thus coal contributions were considered to be zero, sulfate was the most abundant secondary ion in Guangxi personal exposure PM2.5 (6–9% was sulfate, compared with 3% or less nitrate or ammonium), indicating possible long-range transport of coal emissions including secondary sulfate precursors.66,67 Biomass, coal, and vehicle emissions contributed approximately equally to exposures in Beijing, with median contributions of 10–12% of PM2.5, ranging from 6 to 39% from biomass and 7–27% from coal combustion and vehicles. Dust contributed 5–17% of Beijing personal exposures, and SIA concentrations were highly variable, with median contributions of 1–6% but maxima up to ∼30%. In Shanxi, coal was the dominant source of exposures to PM2.5, contributing on average 32% (median: 29%), but as high as 86%, of PM2.5 mass. Biomass burning accounted for up to 25% of Shanxi PM2.5 exposures, and dust and secondary ions generally contributed less than ∼10%. In addition to SOA, unapportioned personal exposure PM2.5 mass also includes contributions from sources related to personal activity such as household dust.24

As another approach to identify major sources, we conducted a PCA using both organic and inorganic chemical components (Table 2). The CMB and mass reconstruction approach described above (Figure 1) is useful for quantifying primary source contributions, particularly in small data sets, but is limited because it requires input of source profiles and does not quantify secondary source contributions.40 PCA is an exploratory data analysis technique that does not require detailed knowledge of contributing factors/sources and can be applied to smaller data sets than multivariate source apportionment techniques such as positive matrix factorization (PMF) and thus is applicable in this context to supplement CMB and provide more information about PM2.5 sources.

Table 2. PCA Results.

  principal component name variance explained (%) determining species (>0.8, 0.6–0.8, 0.4–0.6)a
PC1 coal combustion 22 As, Pb, SO42–, picene, BbF, BeP, Zn
PC2 regional aerosol 19 IcdP, NO3–, NH4+, hopane, ΣBTCAs, norhopane, BbF
PC3 roadway emissions 17 Fe, Mn, Ba, Zn, norhopane, hopane
PC4 soil dust 11 Ti, ΣREEs, Ca, Ba
PC5 biomass burning 10 WSOC, wsKb, levoglucosan
Open in a new tab
a

Species for each principal component are listed in order of loading from highest to lowest; those with loadings >0.8 in bold, between 0.6 and 0.8 in italics, and between 0.4 and 0.6 in regular font. BbF: benzo[b]fluoranthene; BeP: benzo[e]pyrene; IcdP: indeno[1,2,3-cd]pyrene; ΣBTCAs: sum of benzenetricarboxylic acids (1,2,3-, 1,2,4-, and 1,3,5-); norhopane: 17α(H)-21β(H)-30-norhopane; hopane: 17α(H)-21β(H)-hopane; and ΣREEs = sum of rare-earth elements (La, Ce, Pr, Nd, Sm, Eu, Dy, Ho, Yb, and Lu).

b

Water-soluble K (wsK) was calculated by subtracting K+ (measured by IC) from total K (measured by ICPMS).

The sources identified using PCA were consistent with those used in CMB and mass reconstruction, with some differences that help to explain the “unidentified sources”. The component explaining the most variance, PC1, likely represents coal combustion: As and Pb, metals found in emissions from coal combustion as well as coal fly ash,68 had the highest loadings. Other determining species in PC1 are also known tracers of coal combustion: sulfate,66,67 zinc,68,69 and particularly picene, a PAH specific to coal combustion emissions.41,54

Nitrate, ammonium, hopanes, benzenetricarboxylic acids, and PAHs were the determining species for PC2, suggesting a mixed regional aerosol source (including secondary aerosols). Ammonium nitrate is a major component of SIA, and the presence of hopanes and PAHs suggests precursor sources of vehicles/coal and other combustion, respectively.70 Benzenetricarboxylic acids are thought to be derived from aromatic compounds such as PAHs and have been correlated with SOA in ambient aerosols.71,72 In source apportionment studies utilizing both CMB and a multivariate method such as PMF, PMF secondary aerosol source contributions have corresponded well with mass not apportioned by CMB,56,57 supporting our hypothesis that “unidentified sources” include secondary aerosol.

The elements with higher loadings in PC3—Fe, Mn, Ba, and Zn—have as common sources brake wear, tire wear, and/or road dust.73−75 Along with the presence of hopanes, this indicates that PC3 likely represents a roadway source comprising both tailpipe and non-tailpipe emissions.75 We included vehicle tailpipe emissions in the CMB model, but non-tailpipe emissions are derived from multiple sources (e.g., brake wear and tire wear) with variable inorganic and organic content,75,76 making them ill-suited to include in the CMB model. Predominance of the crustal elements Ti, Ca, and summed rare-earth elements in PC4 suggests that PC4 represents resuspended soil dust.73 PC5 is dominated by levoglucosan, water-soluble K, and WSOC, all of which are tracers of biomass burning emissions.77

Personal/Outdoor Ratios of Chemical Components

We calculated personal-to-outdoor concentration ratios, using each personal exposure composite’s corresponding outdoor composite from the same sampling dates and site, in order to investigate the sources of personal exposures in relation to outdoor air pollution.

Primary tracers of solid fuel combustion (levoglucosan, picene, and water-soluble K)53,54 and dust (Ti)73 were enriched in the majority of personal PM2.5 exposure samples (Figure 2a–d). Levoglucosan was up to 20 times higher in personal exposures than in outdoor PM, with median P/O ratios of 2.3, 2.1, and 6.6 in Guangxi, Beijing, and Shanxi, respectively. Outdoor levoglucosan concentrations were the lowest in Shanxi (Table S1), driving the higher P/O ratios observed. For comparison, P/O ratios were much lower (0.5–1.5) among individuals without reported household use of solid fuels in Guangzhou.78 Picene, a marker of coal combustion, was not detected in personal or outdoor samples in Guangxi and was elevated in all Beijing and Shanxi personal exposures. Fuel use data indicate that the Guangxi personal exposure composites did not include any participants using coal, while all composites in Shanxi and the majority in Beijing did include coal use (Figure S2), which corresponds well to the observed picene trend. A nationwide household energy survey reported that, at the province level in rural communities, Guangxi had lower coal use (2.8%) than Beijing (5.8%) or Shanxi (58.6%).79 Water-soluble K was most enriched in personal exposures in Shanxi, slightly enriched in Beijing exposures, and lower in personal PM than in outdoor PM in Guangxi. This trend is consistent with major outdoor contributions to exposure: biomass was used in the majority of Guangxi composites but only half of the Shanxi composites, and relative biomass contributions to outdoor PM were higher in Guangxi than in Shanxi (Figure 1).

Figure 2.

Figure 2

Open in a new tab

Personal/outdoor ratios of selected chemical components of PM2.5. The dashed line is plotted at y = 1 for reference. 1,2,3-BTCA = 1,2,3-benzenetricarboxylic acid. Box midlines indicate median ratios, the box represents the interquartile range (IQR; top = 75th percentile and bottom = 25th percentile), and whiskers extending above and below the box mark the 90th and 10th percentiles, respectively. Individual points represent values that were above the 90th percentile or below the 10th percentile. No data are plotted for picene in Guangxi because it was not detected in outdoor or personal exposure PM2.5 samples there. Gray asterisks denote statistical significance (one-sample Wilcoxon test; *** represents p < 0.001, ** represents p < 0.01, and * represents p < 0.05).

Titanium, a crustal element representative of total dust, had the lowest P/O ratios in Guangxi (1.2), slightly elevated P/O ratios in Beijing (1.7), and the highest in Shanxi (3.7). This may be driven by outdoor PM: outdoor dust concentrations were higher in Guangxi than at the other two sites, and average personal exposure dust concentrations were similar across sites (7–15%). Dust exposures exceeding the corresponding indoor or outdoor dust concentrations have previously been observed in rural settings,11,24 and dust is a plausible component of the “personal cloud” effect, where individuals’ behavior acts as an additional source of exposure.80,81

The P/O ratios of species associated with secondary aerosol were below or approximately one, indicating that these tracers are generally not present above outdoor levels in personal PM2.5 exposures. Median P/O ratios of SO42–, NO3–, and NH4+, which comprise SIA, were 0.4–0.8 in Guangxi, 1.1–1.2 in Beijing, and 0.7–1.2 in Shanxi (Figure 2e–g). As noted earlier in discussion of the PCA, benzenetricarboxylic and other aromatic acids are candidate tracers for anthropogenic SOA. A representative aromatic acid, 1,2,3-benzenetricarboxylic acid, was not enriched in personal exposures: as with the secondary ions, median P/O ratios were 0.9–1.3 at all three sites (Figure 2h). In contrast to solid fuel combustion, secondary aerosol exposures are only due to outdoor air pollution, and P/O ratios close to or less than 1 for secondary aerosol tracers are consistent with that.

Trends in the 17α(H)-21β(H)-hopane and WSOC (Figure 2i,j) were more complex. Hopane concentrations were somewhat elevated in Beijing and Shanxi personal exposures but not in Guangxi (median P/O: 1.9, 1.7, and 0.8, respectively). In an urban context, hopanes are predominantly tracers of vehicle combustion, but with prevalent household coal combustion in the Beijing and Shanxi study areas, coal combustion is also a likely source.41 Hopane/picene ratios and hopane concentrations in general were higher in Beijing than in Shanxi (Figure S6). Given the specificity of picene as a tracer of coal combustion,54 hopane levels are likely more influenced by coal in Shanxi than in Beijing. WSOC was the most elevated in Shanxi (P/O = 2.7), slightly enriched in Beijing exposures (P/O = 1.4), and similar to outdoor in Guangxi personal exposures (P/O = 1.1). Biomass burning and SOA are typically considered the main sources of atmospheric WSOC. However, levoglucosan emission rates and OC solubility depend on combustion conditions,53 which can be highly variable in household biomass stoves, making it difficult to distinguish between WSOC sources in these samples.

Gender Differences in Personal Exposures

Across all three study sites, differences in P/O ratios by gender were generally small: men’s median P/O ratios for each species were 74–123% of women’s median P/O ratios among all but three of the chemical species included in the PCA analysis (Figure S7). Women had higher median P/O ratios than men for levoglucosan, picene, and WSOC (men’s median P/O ratios were 51, 57, and 67% of women’s, respectively). However, women’s and men’s median P/O ratios did not differ significantly for any of these species (Wilcoxon p > 0.05), indicating that variability in the P/O ratios within male and female subgroups was greater than between subgroups.

This overall similarity of chemical composition of men’s and women’s personal exposures in our subsample analysis supports our earlier finding from the full study population that men’s and women’s PM2.5 exposures were nearly identical after accounting for tobacco smoking.32 It is often assumed in HAP studies that women’s exposures exceed men’s due to more time spent cooking and inside the home.7 However, in our study of rural Chinese adults, although women were usually the primary cooks in their households, men were involved with heating stove use and other household energy tasks, and the majority of participants (>70%) at each study site had the same occupation.32 Additionally, in gender-specific mixed-effect models, use of an outdoor heating stove was associated with a smaller decrease in exposure for men than for women.32

Our results are similar to the few previous HAP studies comparing PM2.5 exposure and sources by gender, where women’s and men’s exposures did not differ significantly. In a multicountry study of rural communities using solid fuel stoves, exposures to PM2.5 were not different for men and women in China, India, Chile, Colombia, Tanzania, and Zimbabwe, despite women spending more time per day in the kitchen,15 although women in Bangladesh and Pakistan did have higher PM2.5 exposures than men.15 Men and women in households using biomass in peri-urban India had very similar PM2.5 exposures (means: 55.1 and 58.5 μg/m3, respectively), even though most men reported spending no time on cooking with biomass.13 In contrast, a small exposure study in rural Uganda and Ethiopia reported that women’s PM2.5 exposures were 5–6 times higher than men in the same villages.29 The spectrum of differences in gender-specific exposures highlights the importance of understanding the distribution of domestic responsibilities when assessing exposures, particularly if solid fuels are also used for heating in addition to cooking.

To our knowledge, no studies have assessed potential gender differences in the specific chemical components and sources of exposure in a setting of household solid fuel use. Chemical analysis of even a small subset of PM2.5 exposure samples, as was conducted in this study, can show whether a different set of sources contributes to women’s exposures compared to men’s exposures and thus explain trends observed in men’s and women’s overall PM2.5 exposure levels.

Fuel Use and Exposure to Solid Fuel Tracers

To evaluate how PM2.5 exposures and fuel use were related, we compared P/O ratios of coal and biomass burning tracers in composites with versus without use of that fuel. Conceptually, P/O ratios of tracers of a given fuel should be much higher for users of that fuel than for those who do not use that fuel. Using P/O ratios rather than concentrations normalized to PM2.5 mass or air volume accounts for background ambient PM2.5 contributions, which also enables comparison across all study sites together.

Coal users had higher P/O ratios of coal combustion tracers than participants not using coal (Figure 3a). For As, Pb, and sulfate, P/O ratios were greater than 1 for coal users and at or below 1 for nonusers, and differences were statistically significant (median coal and non-coal As: 1.4 and 0.9, p = 0.044; Pb: 1.6 and 1.0, p = 0.0008; and sulfate: 1.2 and 0.8, p = 0.0008). The same trend was observed for picene P/O ratios, although not statistically significant. Interestingly, all picene P/O ratios were greater than one, averaging 3.3 with coal use and 2.6 without coal use, likely due to coal use in nearby households which is not captured by ambient samples representing regional PM2.5.

Figure 3.

Figure 3

Open in a new tab

Personal/outdoor ratios of selected species emitted by (a) coal and (b) biomass burning. For each fuel, composites with use of that fuel are plotted as the darker box (left), while the lighter box (right) represents composites without use of that fuel. The dashed line is plotted at y = 1 for reference. Medians were compared using the unpaired Wilcoxon test (ns = not significant [p > 0.05]; * = p ≤ 0.05; and *** = p ≤ 0.001). Box-and-whisker statistics are calculated as described for Figure 2.

Biomass burning tracers did not exhibit the expected trend: personal exposure composites with and without biomass users did not have significant differences in P/O ratios of levoglucosan, water-soluble K, and WSOC (Figure 3b). This may be partially attributable to these species’ selectivity as tracers. Water-soluble K and WSOC are emitted by biomass burning but are also components of dust and secondary aerosol, respectively.53 However, levoglucosan is a pyrolysis product of cellulose and thus considered specific to biomass burning,82 although it has also been detected at lower levels in emissions from coal combustion.83,84 Moreover, median P/O ratios of biomass tracers were greater than one regardless of biomass use (biomass and non-biomass levoglucosan: 3.6 and 5.9; wsK: 1.4 and 1.7; and WSOC: 1.5 and 2.4). These high levoglucosan P/O ratios, particularly for exposures among participants not using biomass, are evidence of exposure to biomass burning emissions that are neither from regional ambient PM2.5 nor participants’ own household fuel use. Most likely, as noted above for picene, it is proximity to other nearby homes using biomass stoves that elevates their exposures.

These discrepancies between self-reported fuel use and chemical components of actual exposures parallel discrepancies between fuel use and PM2.5 exposures or area concentrations, in this study and others. In our earlier study of all personal PM2.5 exposures, exclusive clean fuel users had less than 15% lower PM2.5 exposure compared with solid fuel users, even after adjusting for outdoor PM2.5 and other sociodemographic variables.32 A different source apportionment study of personal PM2.5 exposures in Yunnan province (China) found that, although cooking and wood combustion typically accounted for over 75% of women’s PM2.5 exposures, self-reported cooking frequency was not strongly correlated with contributions from either source.30 Duration of wood-chimney stove use was also not a predictor of kitchen PM2.5 in a study in rural households in Sichuan, China.31

The importance of community-level change has been noted for some of the many complex factors in household energy transitions (such as adoption and continued use) and increasingly also for achieving meaningful reductions in air pollution exposures in order to improve health outcomes.85,86 We found that chemical tracers of solid fuel combustion were present in exposures at levels not entirely explained by household fuel use or background levels in outdoor PM2.5, indicating that fuel use in neighboring homes may impact personal exposures. Notably, exposures of participants using biomass were not higher in levoglucosan than those of participants not using biomass, even after accounting for background ambient levels. Chemical composition of PM2.5 exposures did not differ between men and women, despite differences in domestic activities. Our findings contribute to growing evidence that reducing exposures to HAP will require both household- and community-level changes in stove and fuel use and that outdoor air pollution is an important consideration when evaluating or predicting the efficacy of household-level air pollution controls.

Acknowledgments

The authors thank the field staff and participants involved in the ICP study. We also thank Angela Albrecht, Brandon Shelton, and Martin Shafer at the Wisconsin State Laboratory of Hygiene for chemical analysis of samples and Sierra Clark and Joy Tseng for sample and data management assistance. This research was supported by the Wellcome Trust, UK (grant 103906/Z/14/Z); National Natural Science Foundation of China, China (grant 81473044 and Innovative Research Groups grant 51521005); and the Canadian Institutes for Health Research (grant 137535).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.1c01368.

  • Additional details about study sites, sample selection, and fuel use data; compositing scheme, fuel use trends, filter and blank PM2.5 masses, picene/hopane comparison, and personal/outdoor ratios of chemical components; and summary statistics for chemical species and source contributions (PDF)

The authors declare no competing financial interest.

Supplementary Material

es1c01368_si_001.pdf (1MB, pdf)

References

  1. Health Effects Institute . State of Global Air 2020; Special Report; Health Effects Institute: Boston, MA, 2020; p 28.
  2. Du W.; Li X.; Chen Y.; Shen G. Household Air Pollution and Personal Exposure to Air Pollutants in Rural China – A Review. Environ. Pollut. 2018, 237, 625–638. 10.1016/j.envpol.2018.02.054. [DOI] [PubMed] [Google Scholar]
  3. Kelly F. J.; Fussell J. C. Air Pollution and Airway Disease. Clin. Exp. Allergy 2011, 41, 1059–1071. 10.1111/j.1365-2222.2011.03776.x. [DOI] [PubMed] [Google Scholar]
  4. Miller M. R. Oxidative Stress and the Cardiovascular Effects of Air Pollution. Free Radic. Biol. Med. 2020, 151, 69–87. 10.1016/j.freeradbiomed.2020.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Smith R. B.; Fecht D.; Gulliver J.; Beevers S. D.; Dajnak D.; Blangiardo M.; Ghosh R. E.; Hansell A. L.; Kelly F. J.; Anderson H. R.; Toledano M. B. Impact of London’s Road Traffic Air and Noise Pollution on Birth Weight: Retrospective Population Based Cohort Study. Br. Med. J. 2017, 359, j5299. 10.1136/bmj.j5299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brook R. D.; Newby D. E.; Rajagopalan S. Air Pollution and Cardiometabolic Disease: An Update and Call for Clinical Trials. Am. J. Hypertens. 2017, 31, 1–10. 10.1093/ajh/hpx109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Clark M. L.; Peel J. L.; Balakrishnan K.; Breysse P. N.; Chillrud S. N.; Naeher L. P.; Rodes C. E.; Vette A. F.; Balbus J. M. Health and Household Air Pollution from Solid Fuel Use: The Need for Improved Exposure Assessment. Environ. Health Perspect. 2013, 121, 1120–1128. 10.1289/ehp.1206429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Delapena S.; Piedrahita R.; Pillarisetti A.; Garland C.; Rossanese M. E.; Johnson M.; Pennise D. Using Personal Exposure Measurements of Particulate Matter to Estimate Health Impacts Associated with Cooking in Peri-Urban Accra, Ghana. Energy Sustainable Dev. 2018, 45, 190–197. 10.1016/j.esd.2018.05.013. [DOI] [Google Scholar]
  9. Bruce N.; Pope D.; Rehfuess E.; Balakrishnan K.; Adair-Rohani H.; Dora C. WHO Indoor Air Quality Guidelines on Household Fuel Combustion: Strategy Implications of New Evidence on Interventions and Exposure–Risk Functions. Atmos. Environ. 2015, 106, 451–457. 10.1016/j.atmosenv.2014.08.064. [DOI] [Google Scholar]
  10. Piedrahita R.; Kanyomse E.; Coffey E.; Xie M.; Hagar Y.; Alirigia R.; Agyei F.; Wiedinmyer C.; Dickinson K. L.; Oduro A.; Hannigan M. Exposures to and Origins of Carbonaceous PM2.5 in a Cookstove Intervention in Northern Ghana. Sci. Total Environ. 2017, 576, 178–192. 10.1016/j.scitotenv.2016.10.069. [DOI] [PubMed] [Google Scholar]
  11. Xu H.; Li Y.; Guinot B.; Wang J.; He K.; Ho K. F.; Cao J.; Shen Z.; Sun J.; Lei Y.; Gong X.; Zhang T. Personal Exposure of PM2.5 Emitted from Solid Fuels Combustion for Household Heating and Cooking in Rural Guanzhong Plain, Northwestern China. Atmos. Environ. 2018, 185, 196–206. 10.1016/j.atmosenv.2018.05.018. [DOI] [Google Scholar]
  12. Ye W.; Saikawa E.; Avramov A.; Cho S.-H.; Chartier R. Household Air Pollution and Personal Exposure from Burning Firewood and Yak Dung in Summer in the Eastern Tibetan Plateau. Environ. Pollut. 2020, 263, 114531. 10.1016/j.envpol.2020.114531. [DOI] [PubMed] [Google Scholar]
  13. Sanchez M.; Milà C.; Sreekanth V.; Balakrishnan K.; Sambandam S.; Nieuwenhuijsen M.; Kinra S.; Marshall J. D.; Tonne C. Personal Exposure to Particulate Matter in Peri-Urban India: Predictors and Association with Ambient Concentration at Residence. J. Expo. Sci. Environ. Epidemiol. 2020, 30, 596–605. 10.1038/s41370-019-0150-5. [DOI] [PubMed] [Google Scholar]
  14. Chen Y.; Fei J.; Sun Z.; Shen G.; Du W.; Zang L.; Yang L.; Wang Y.; Wu R.; Chen A.; Zhao M. Household Air Pollution from Cooking and Heating and Its Impacts on Blood Pressure in Residents Living in Rural Cave Dwellings in Loess Plateau of China. Environ. Sci. Pollut. Res. 2020, 27, 36677–36687. 10.1007/s11356-020-09677-1. [DOI] [PubMed] [Google Scholar]
  15. Shupler M.; Hystad P.; Birch A.; Miller-Lionberg D.; Jeronimo M.; Arku R. E.; Chu Y. L.; Mushtaha M.; Heenan L.; Rangarajan S.; Seron P.; Lanas F.; Cazor F.; Lopez-Jaramillo P.; Camacho P. A.; Perez M.; Yeates K.; West N.; Ncube T.; Ncube B.; Chifamba J.; Yusuf R.; Khan A.; Hu B.; Liu X.; Wei L.; Tse L. A.; Mohan D.; Kumar P.; Gupta R.; Mohan I.; Jayachitra K. G.; Mony P. K.; Rammohan K.; Nair S.; Lakshmi P. V. M.; Sagar V.; Khawaja R.; Iqbal R.; Kazmi K.; Yusuf S.; Brauer M. Household and Personal Air Pollution Exposure Measurements from 120 Communities in Eight Countries: Results from the PURE-AIR Study. Lancet Planet. Health 2020, 4, e451–e462. 10.1016/s2542-5196(20)30197-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fandiño-Del-Rio M.; Kephart J. L.; Williams K. N.; Moulton L. H.; Steenland K.; Checkley W.; Koehler K. Household Air Pollution Exposure and Associations with Household Characteristics among Biomass Cookstove Users in Puno, Peru. Environ. Res. 2020, 191, 110028. 10.1016/j.envres.2020.110028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Chan K. H.; Xia X.; Ho K.-f.; Guo Y.; Kurmi O. P.; Du H.; Bennett D. A.; Bian Z.; Kan H.; McDonnell J.; Schmidt D.; Kerosi R.; Li L.; Lam K. B. H.; Chen Z. Regional and Seasonal Variations in Household and Personal Exposures to Air Pollution in One Urban and Two Rural Chinese Communities: A Pilot Study to Collect Time-Resolved Data Using Static and Wearable Devices. Environ. Int. 2021, 146, 106217. 10.1016/j.envint.2020.106217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Milà C.; Salmon M.; Sanchez M.; Ambrós A.; Bhogadi S.; Sreekanth V.; Nieuwenhuijsen M.; Kinra S.; Marshall J. D.; Tonne C. When, Where, and What? Characterizing Personal PM2.5 Exposure in Periurban India by Integrating GPS, Wearable Camera, and Ambient and Personal Monitoring Data. Environ. Sci. Technol. 2018, 52, 13481–13490. 10.1021/acs.est.8b03075. [DOI] [PubMed] [Google Scholar]
  19. Sambandam S.; Mukhopadhyay K.; Sendhil S.; Ye W.; Pillarisetti A.; Thangavel G.; Natesan D.; Ramasamy R.; Natarajan A.; Aravindalochanan V.; Vinayagamoorthi A.; Sivavadivel S.; Uma Maheswari R.; Balakrishnan L.; Gayatri S.; Nargunanathan S.; Madhavan S.; Puttaswamy N.; Garg S. S.; Quinn A.; Rosenthal J.; Johnson M.; Liao J.; Steenland K.; Piedhrahita R.; Peel J.; Checkley W.; Clasen T.; Balakrishnan K. Exposure Contrasts Associated with a Liquefied Petroleum Gas (LPG) Intervention at Potential Field Sites for the Multi-Country Household Air Pollution Intervention Network (HAPIN) Trial in India: Results from Pilot Phase Activities in Rural Tamil Nadu. BMC Publ. Health 2020, 20, 1799. 10.1186/s12889-020-09865-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Grajeda L. M.; Thompson L. M.; Arriaga W.; Canuz E.; Omer S. B.; Sage M.; Azziz-Baumgartner E.; Bryan J. P.; McCracken J. P. Effectiveness of Gas and Chimney Biomass Stoves for Reducing Household Air Pollution Pregnancy Exposure in Guatemala: Sociodemographic Effect Modifiers. Int. J. Environ. Res. Publ. Health 2020, 17, 7723. 10.3390/ijerph17217723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. He K.; Xu H.; Feng R.; Shen Z.; Li Y.; Zhang Y.; Sun J.; Zhang Q.; Zhang T.; Yang L.; Liu P.; Ho K. F.; Cao J. Characteristics of Indoor and Personal Exposure to Particulate Organic Compounds Emitted from Domestic Solid Fuel Combustion in Rural Areas of Northwest China. Atmos. Res. 2021, 248, 105181. 10.1016/j.atmosres.2020.105181. [DOI] [Google Scholar]
  22. Downward G. S.; Hu W.; Rothman N.; Reiss B.; Wu G.; Wei F.; Chapman R. S.; Portengen L.; Qing L.; Vermeulen R. Polycyclic Aromatic Hydrocarbon Exposure in Household Air Pollution from Solid Fuel Combustion among the Female Population of Xuanwei and Fuyuan Counties, China. Environ. Sci. Technol. 2014, 48, 14632–14641. 10.1021/es504102z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li Y.; Xu H.; Wang J.; Ho S. S. H.; He K.; Shen Z.; Ning Z.; Sun J.; Li L.; Lei R.; Zhang T.; Lei Y.; Yang L.; Cao Y.; Cao J. Personal Exposure to PM2.5-Bound Organic Species from Domestic Solid Fuel Combustion in Rural Guanzhong Basin, China: Characteristics and Health Implication. Chemosphere 2019, 227, 53–62. 10.1016/j.chemosphere.2019.04.010. [DOI] [PubMed] [Google Scholar]
  24. Lai A. M.; Carter E.; Shan M.; Ni K.; Clark S.; Ezzati M.; Wiedinmyer C.; Yang X.; Baumgartner J.; Schauer J. J. Chemical Composition and Source Apportionment of Ambient, Household, and Personal Exposures to PM 2.5 in Communities Using Biomass Stoves in Rural China. Sci. Total Environ. 2019, 646, 309–319. 10.1016/j.scitotenv.2018.07.322. [DOI] [PubMed] [Google Scholar]
  25. Hu R.; Wang S.; Aunan K.; Zhao M.; Chen L.; Liu Z.; Hansen M. H. Personal Exposure to PM2.5 in Chinese Rural Households in the Yangtze River Delta. Indoor Air 2019, 29, 403–412. 10.1111/ina.12537. [DOI] [PubMed] [Google Scholar]
  26. Pillarisetti A.; Carter E.; Rajkumar S.; Young B. N.; Benka-Coker M. L.; Peel J. L.; Johnson M.; Clark M. L. Measuring Personal Exposure to Fine Particulate Matter (PM2.5) among Rural Honduran Women: A Field Evaluation of the Ultrasonic Personal Aerosol Sampler (UPAS). Environ. Int. 2019, 123, 50–53. 10.1016/j.envint.2018.11.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Wu J.; Xiao X.; Li Y.; Yang F.; Yang S.; Sun L.; Ma R.; Wang M. C. Personal Exposure to Fine Particulate Matter (PM2.5) of Pregnant Women during Three Trimesters in Rural Yunnan of China. Environ. Pollut. 2020, 256, 113055. 10.1016/j.envpol.2019.113055. [DOI] [PubMed] [Google Scholar]
  28. Huang Y.; Du W.; Chen Y.; Shen G.; Su S.; Lin N.; Shen H.; Zhu D.; Yuan C.; Duan Y.; Liu J.; Li B.; Tao S. Household Air Pollution and Personal Inhalation Exposure to Particles (TSP/PM2.5/PM1.0/PM0.25) in Rural Shanxi, North China. Environ. Pollut. 2017, 231, 635–643. 10.1016/j.envpol.2017.08.063. [DOI] [PubMed] [Google Scholar]
  29. Okello G.; Devereux G.; Semple S. Women and Girls in Resource Poor Countries Experience Much Greater Exposure to Household Air Pollutants than Men: Results from Uganda and Ethiopia. Environ. Int. 2018, 119, 429–437. 10.1016/j.envint.2018.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Huang W.; Baumgartner J.; Zhang Y.; Wang Y.; Schauer J. J. Source Apportionment of Air Pollution Exposures of Rural Chinese Women Cooking with Biomass Fuels. Atmos. Environ. 2015, 104, 79–87. 10.1016/j.atmosenv.2014.12.066. [DOI] [Google Scholar]
  31. Snider G.; Carter E.; Clark S.; Tseng J.; Yang X.; Ezzati M.; Schauer J. J.; Wiedinmyer C.; Baumgartner J. Impacts of Stove Use Patterns and Outdoor Air Quality on Household Air Pollution and Cardiovascular Mortality in Southwestern China. Environ. Int. 2018, 117, 116–124. 10.1016/j.envint.2018.04.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lee M.; Carter E.; Yan L.; Chan Q.; Elliott P.; Ezzati M.; Kelly F.; Schauer J. J.; Wu Y.; Yang X.; Zhao L.; Baumgartner J. Determinants of Personal Exposure to PM2.5 and Black Carbon in Chinese Adults: A Repeated-Measures Study in Villages Using Solid Fuel Energy. Environ. Int. 2021, 146, 106297. 10.1016/j.envint.2020.106297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Carter E.; Yan L.; Fu Y.; Robinson B.; Kelly F.; Elliott P.; Wu Y.; Zhao L.; Ezzati M.; Yang X.; Chan Q.; Baumgartner J. Household Transitions to Clean Energy in a Multiprovincial Cohort Study in China. Nat. Sustain. 2020, 3, 42–50. 10.1038/s41893-019-0432-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yan L.; Carter E.; Fu Y.; Guo D.; Huang P.; Xie G.; Xie W.; Zhu Y.; Kelly F.; Elliott P.; Zhao L.; Yang X.; Ezzati M.; Wu Y.; Baumgartner J.; Chan Q. Study Protocol: The INTERMAP China Prospective (ICP) Study. Wellcome Open Res. 2020, 4, 154. 10.12688/wellcomeopenres.15470.2. [DOI] [Google Scholar]
  35. Demokritou P.; Kavouras I. G.; Ferguson S. T.; Koutrakis P. Development and Laboratory Performance Evaluation of a Personal Multipollutant Sampler for Simultaneous Measurements of Particulate and Gaseous Pollutants. Aerosol Sci. Technol. 2001, 35, 741–752. 10.1080/02786820152546789. [DOI] [Google Scholar]
  36. Ni K.; Carter E.; Schauer J. J.; Ezzati M.; Zhang Y.; Niu H.; Lai A. M.; Shan M.; Wang Y.; Yang X.; Baumgartner J. Seasonal Variation in Outdoor, Indoor, and Personal Air Pollution Exposures of Women Using Wood Stoves in the Tibetan Plateau: Baseline Assessment for an Energy Intervention Study. Environ. Int. 2016, 94, 449–457. 10.1016/j.envint.2016.05.029. [DOI] [PubMed] [Google Scholar]
  37. Okuda T.; Schauer J. J.; Shafer M. M. Improved Methods for Elemental Analysis of Atmospheric Aerosols for Evaluating Human Health Impacts of Aerosols in East Asia. Atmos. Environ. 2014, 97, 552–555. 10.1016/j.atmosenv.2014.01.043. [DOI] [Google Scholar]
  38. Stone E. A.; Snyder D. C.; Sheesley R. J.; Sullivan A. P.; Weber R. J.; Schauer J. J. Source Apportionment of Fine Organic Aerosol in Mexico City during the MILAGRO Experiment 2006. Atmos. Chem. Phys. 2008, 8, 1249–1259. 10.5194/acp-8-1249-2008. [DOI] [Google Scholar]
  39. Villalobos A. M.; Barraza F.; Jorquera H.; Schauer J. J. Chemical Speciation and Source Apportionment of Fine Particulate Matter in Santiago, Chile, 2013. Sci. Total Environ. 2015, 512–513, 133–142. 10.1016/j.scitotenv.2015.01.006. [DOI] [PubMed] [Google Scholar]
  40. Hopke P. K. Review of Receptor Modeling Methods for Source Apportionment. J. Air Waste Manage. Assoc. 2016, 66, 237–259. 10.1080/10962247.2016.1140693. [DOI] [PubMed] [Google Scholar]
  41. Zhang Y.; Schauer J. J.; Zhang Y.; Zeng L.; Wei Y.; Liu Y.; Shao M. Characteristics of Particulate Carbon Emissions from Real-World Chinese Coal Combustion. Environ. Sci. Technol. 2008, 42, 5068–5073. 10.1021/es7022576. [DOI] [PubMed] [Google Scholar]
  42. Zhang Y.Study on Speciation of Particulate Organic Matter from Combustion Sources. Ph.D. Thesis, Peking University, Beijing, China, 2006. [Google Scholar]
  43. He L.-Y.; Hu M.; Zhang Y.-H.; Huang X.-F.; Yao T.-T. Fine Particle Emissions from On-Road Vehicles in the Zhujiang Tunnel, China. Environ. Sci. Technol. 2008, 42, 4461–4466. 10.1021/es7022658. [DOI] [PubMed] [Google Scholar]
  44. Coulter C. T.EPA-CMB8.2 Users Manual, 2004. http://www.epa.gov/scram001/models/receptor/EPA-CMB82Manual.pdf.
  45. Miller-Schulze J. P.; Shafer M.; Schauer J. J.; Heo J.; Solomon P. A.; Lantz J.; Artamonova M.; Chen B.; Imashev S.; Sverdlik L.; Carmichael G.; DeMinter J. Seasonal Contribution of Mineral Dust and Other Major Components to Particulate Matter at Two Remote Sites in Central Asia. Atmos. Environ. 2015, 119, 11–20. 10.1016/j.atmosenv.2015.07.011. [DOI] [Google Scholar]
  46. von Schneidemesser E.; Stone E. A.; Quraishi T. A.; Shafer M. M.; Schauer J. J. Toxic Metals in the Atmosphere in Lahore, Pakistan. Sci. Total Environ. 2010, 408, 1640–1648. 10.1016/j.scitotenv.2009.12.022. [DOI] [PubMed] [Google Scholar]
  47. Seinfeld J. H.; Pandis S. N.. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; Wiley: New York, NY, 1998. [Google Scholar]
  48. Ministry of Environmental Protection (MEP) . GB3095-2012: Ambient Air Quality Standards: Beijing, China, 2012.
  49. World Health Organization . WHO Air Quality Guidelines for Particulate Matter, Ozone, Nitrogen Dioxide and Sulfur Dioxide: Global Update 2005, 2006. [PubMed]
  50. Aunan K.; Hansen M. H.; Liu Z.; Wang S. The Hidden Hazard of Household Air Pollution in Rural China. Environ. Sci. Pol. 2019, 93, 27–33. 10.1016/j.envsci.2018.12.004. [DOI] [Google Scholar]
  51. Li W.; Wang C.; Wang H.; Chen J.; Yuan C.; Li T.; Wang W.; Shen H.; Huang Y.; Wang R.; Wang B.; Zhang Y.; Chen H.; Chen Y.; Tang J.; Wang X.; Liu J.; Coveney R. M.; Tao S. Distribution of Atmospheric Particulate Matter (PM) in Rural Field, Rural Village and Urban Areas of Northern China. Environ. Pollut. 2014, 185, 134–140. 10.1016/j.envpol.2013.10.042. [DOI] [PubMed] [Google Scholar]
  52. Lin C.; Lau A.; Li Y.; Fung J.; Li C.; Lu X.; Li Z. Difference in PM2.5 Variations between Urban and Rural Areas over Eastern China from 2001 to 2015. Atmosphere 2018, 9, 312. 10.3390/atmos9080312. [DOI] [Google Scholar]
  53. Vicente E. D.; Alves C. A. An Overview of Particulate Emissions from Residential Biomass Combustion. Atmos. Res. 2018, 199, 159–185. 10.1016/j.atmosres.2017.08.027. [DOI] [PubMed] [Google Scholar]
  54. Oros D. R.; Simoneit B. R. T. Identification and Emission Rates of Molecular Tracers in Coal Smoke Particulate Matter. Fuel 2000, 79, 515–536. 10.1016/s0016-2361(99)00153-2. [DOI] [Google Scholar]
  55. Chen X.-C.; Chuang H.-C.; Ward T. J.; Tian L.; Cao J.-J.; Ho S. S.-H.; Lau N.-C.; Hsiao T.-C.; Yim S. H.; Ho K.-F. Indoor, Outdoor, and Personal Exposure to PM2.5 and Their Bioreactivity among Healthy Residents of Hong Kong. Environ. Res. 2020, 188, 109780. 10.1016/j.envres.2020.109780. [DOI] [PubMed] [Google Scholar]
  56. Heo J.; Dulger M.; Olson M. R.; McGinnis J. E.; Shelton B. R.; Matsunaga A.; Sioutas C.; Schauer J. J. Source Apportionments of PM2.5 Organic Carbon Using Molecular Marker Positive Matrix Factorization and Comparison of Results from Different Receptor Models. Atmos. Environ. 2013, 73, 51–61. 10.1016/j.atmosenv.2013.03.004. [DOI] [Google Scholar]
  57. Bae M.-S.; Skiles M. J.; Lai A. M.; Olson M. R.; de Foy B.; Schauer J. J. Assessment of Forest Fire Impacts on Carbonaceous Aerosols Using Complementary Molecular Marker Receptor Models at Two Urban Locations in California’s San Joaquin Valley. Environ. Pollut. 2019, 246, 274–283. 10.1016/j.envpol.2018.12.013. [DOI] [PubMed] [Google Scholar]
  58. Liao J.; Zimmermann Jin A.; Chafe Z. A.; Pillarisetti A.; Yu T.; Shan M.; Yang X.; Li H.; Liu G.; Smith K. R. The Impact of Household Cooking and Heating with Solid Fuels on Ambient PM2.5 in Peri-Urban Beijing. Atmos. Environ. 2017, 165, 62–72. 10.1016/j.atmosenv.2017.05.053. [DOI] [Google Scholar]
  59. Liu J.; Mauzerall D. L.; Chen Q.; Zhang Q.; Song Y.; Peng W.; Klimont Z.; Qiu X.; Zhang S.; Hu M.; Lin W.; Smith K. R.; Zhu T. Air Pollutant Emissions from Chinese Households: A Major and Underappreciated Ambient Pollution Source. Proc. Natl. Acad. Sci. 2016, 113, 7756–7761. 10.1073/pnas.1604537113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wu F.; Yu Y.; Sun J.; Zhang J.; Wang J.; Tang G.; Wang Y. Characteristics, Source Apportionment and Reactivity of Ambient Volatile Organic Compounds at Dinghu Mountain in Guangdong Province, China. Sci. Total Environ. 2016, 548–549, 347–359. 10.1016/j.scitotenv.2015.11.069. [DOI] [PubMed] [Google Scholar]
  61. Wang M.; Shao M.; Lu S.-H.; Yang Y.-D.; Chen W.-T. Evidence of Coal Combustion Contribution to Ambient VOCs during Winter in Beijing. Chin. Chem. Lett. 2013, 24, 829–832. 10.1016/j.cclet.2013.05.029. [DOI] [Google Scholar]
  62. Schroder J. C.; Campuzano-Jost P.; Day D. A.; Shah V.; Larson K.; Sommers J. M.; Sullivan A. P.; Campos T.; Reeves J. M.; Hills A.; Hornbrook R. S.; Blake N. J.; Scheuer E.; Guo H.; Fibiger D. L.; McDuffie E. E.; Hayes P. L.; Weber R. J.; Dibb J. E.; Apel E. C.; Jaeglé L.; Brown S. S.; Thornton J. A.; Jimenez J. L. Sources and Secondary Production of Organic Aerosols in the Northeastern United States during WINTER. J. Geophys. Res. Atmos. 2018, 123, 7771–7796. 10.1029/2018jd028475. [DOI] [Google Scholar]
  63. Zhang F.; Shang X.; Chen H.; Xie G.; Fu Y.; Wu D.; Sun W.; Liu P.; Zhang C.; Mu Y.; Zeng L.; Wan M.; Wang Y.; Xiao H.; Wang G.; Chen J. Significant Impact of Coal Combustion on VOCs Emissions in Winter in a North China Rural Site. Sci. Total Environ. 2020, 720, 137617. 10.1016/j.scitotenv.2020.137617. [DOI] [PubMed] [Google Scholar]
  64. Hu W. W.; Hu M.; Yuan B.; Jimenez J. L.; Tang Q.; Peng J. F.; Hu W.; Shao M.; Wang M.; Zeng L. M.; Wu Y. S.; Gong Z. H.; Huang X. F.; He L. Y. Insights on Organic Aerosol Aging and the Influence of Coal Combustion at a Regional Receptor Site of Central Eastern China. Atmos. Chem. Phys. 2013, 13, 10095–10112. 10.5194/acp-13-10095-2013. [DOI] [Google Scholar]
  65. Guo H.; Wang T.; Simpson I. J.; Blake D. R.; Yu X. M.; Kwok Y. H.; Li Y. S. Source Contributions to Ambient VOCs and CO at a Rural Site in Eastern China. Atmos. Environ. 2004, 38, 4551–4560. 10.1016/j.atmosenv.2004.05.004. [DOI] [Google Scholar]
  66. Dai Q.; Bi X.; Song W.; Li T.; Liu B.; Ding J.; Xu J.; Song C.; Yang N.; Schulze B. C.; Zhang Y.; Feng Y.; Hopke P. K. Residential Coal Combustion as a Source of Primary Sulfate in Xi’an, China. Atmos. Environ. 2019, 196, 66–76. 10.1016/j.atmosenv.2018.10.002. [DOI] [Google Scholar]
  67. Clements N.; Eav J.; Xie M.; Hannigan M. P.; Miller S. L.; Navidi W.; Peel J. L.; Schauer J. J.; Shafer M. M.; Milford J. B. Concentrations and Source Insights for Trace Elements in Fine and Coarse Particulate Matter. Atmos. Environ. 2014, 89, 373–381. 10.1016/j.atmosenv.2014.01.011. [DOI] [Google Scholar]
  68. Duan J.; Tan J. Atmospheric Heavy Metals and Arsenic in China: Situation, Sources and Control Policies. Atmos. Environ. 2013, 74, 93–101. 10.1016/j.atmosenv.2013.03.031. [DOI] [Google Scholar]
  69. Li R.; Li J.; Cui L.; Wu Y.; Fu H.; Chen J.; Chen M. Atmospheric Emissions of Cu and Zn from Coal Combustion in China: Spatio-Temporal Distribution, Human Health Effects, and Short-Term Prediction. Environ. Pollut. 2017, 229, 724–734. 10.1016/j.envpol.2017.05.068. [DOI] [PubMed] [Google Scholar]
  70. Chow J. C.; Watson J. G. Review of PM2.5 and PM10 Apportionment for Fossil Fuel Combustion and Other Sources by the Chemical Mass Balance Receptor Model. Energy Fuels 2002, 16, 222–260. 10.1021/ef0101715. [DOI] [Google Scholar]
  71. Sheesley R. J.; Schauer J. J.; Bean E.; Kenski D. Trends in Secondary Organic Aerosol at a Remote Site in Michigan’s Upper Peninsula. Environ. Sci. Technol. 2004, 38, 6491–6500. 10.1021/es049104q. [DOI] [PubMed] [Google Scholar]
  72. He X.; Huang X. H. H.; Chow K. S.; Wang Q.; Zhang T.; Wu D.; Yu J. Z. Abundance and Sources of Phthalic Acids, Benzene-Tricarboxylic Acids, and Phenolic Acids in PM2.5 at Urban and Suburban Sites in Southern China. ACS Earth Space Chem. 2018, 2, 147–158. 10.1021/acsearthspacechem.7b00131. [DOI] [Google Scholar]
  73. Moreno T.; Querol X.; Alastuey A.; Gibbons W. Identification of Chemical Tracers in the Characterisation and Source Apportionment of Inhalable Inorganic Airborne Particles: An Overview. Biomarkers 2009, 14, 17–22. 10.1080/13547500902965435. [DOI] [PubMed] [Google Scholar]
  74. Harrison R. M.; Jones A. M.; Gietl J.; Yin J.; Green D. C. Estimation of the Contributions of Brake Dust, Tire Wear, and Resuspension to Nonexhaust Traffic Particles Derived from Atmospheric Measurements. Environ. Sci. Technol. 2012, 46, 6523–6529. 10.1021/es300894r. [DOI] [PubMed] [Google Scholar]
  75. Pant P.; Harrison R. M. Estimation of the Contribution of Road Traffic Emissions to Particulate Matter Concentrations from Field Measurements: A Review. Atmos. Environ. 2013, 77, 78–97. 10.1016/j.atmosenv.2013.04.028. [DOI] [Google Scholar]
  76. Rauert C.; Rødland E. S.; Okoffo E. D.; Reid M. J.; Meland S.; Thomas K. V. Challenges with Quantifying Tire Road Wear Particles: Recognizing the Need for Further Refinement of the ISO Technical Specification. Environ. Sci. Technol. Lett. 2021, 8, 231. 10.1021/acs.estlett.0c00949. [DOI] [Google Scholar]
  77. Simoneit B. R. T. Biomass Burning—a Review of Organic Tracers for Smoke from Incomplete Combustion. Appl. Geochem. 2002, 17, 129–162. 10.1016/s0883-2927(01)00061-0. [DOI] [Google Scholar]
  78. Chen X.-C.; Jahn H. J.; Engling G.; Ward T. J.; Kraemer A.; Ho K.-F.; Yim S. H. L.; Chan C.-Y. Chemical Characterization and Sources of Personal Exposure to Fine Particulate Matter (PM2.5) in the Megacity of Guangzhou, China. Environ. Pollut. 2017, 231, 871–881. 10.1016/j.envpol.2017.08.062. [DOI] [PubMed] [Google Scholar]
  79. Duan X.; Jiang Y.; Wang B.; Zhao X.; Shen G.; Cao S.; Huang N.; Qian Y.; Chen Y.; Wang L. Household Fuel Use for Cooking and Heating in China: Results from the First Chinese Environmental Exposure-Related Human Activity Patterns Survey (CEERHAPS). Appl. Energy 2014, 136, 692–703. 10.1016/j.apenergy.2014.09.066. [DOI] [Google Scholar]
  80. Rodes C. E.; Kamens R. M.; Wiener R. W. The Significance and Characteristics of the Personal Activity Cloud on Exposure Assessment Measurements for Indoor Contaminants. Indoor Air 1991, 1, 123–145. 10.1111/j.1600-0668.1991.03-12.x. [DOI] [Google Scholar]
  81. Wallace L. Correlations of Personal Exposure to Particles with Outdoor Air Measurements: A Review of Recent Studies. Aerosol Sci. Technol. 2000, 32, 15–25. 10.1080/027868200303894. [DOI] [Google Scholar]
  82. Simoneit B. R. T.; Schauer J. J.; Nolte C. G.; Oros D. R.; Elias V. O.; Fraser M. P.; Rogge W. F.; Cass G. R. Levoglucosan, a Tracer for Cellulose in Biomass Burning and Atmospheric Particles. Atmos. Environ. 1999, 33, 173–182. 10.1016/s1352-2310(98)00145-9. [DOI] [Google Scholar]
  83. Fabbri D.; Torri C.; Simoneit B. R. T.; Marynowski L.; Rushdi A. I.; Fabiańska M. J. Levoglucosan and Other Cellulose and Lignin Markers in Emissions from Burning of Miocene Lignites. Atmos. Environ. 2009, 43, 2286–2295. 10.1016/j.atmosenv.2009.01.030. [DOI] [Google Scholar]
  84. Yan C.; Zheng M.; Sullivan A. P.; Shen G.; Chen Y.; Wang S.; Zhao B.; Cai S.; Desyaterik Y.; Li X.; Zhou T.; Gustafsson Ö.; Collett J. L. Residential Coal Combustion as a Source of Levoglucosan in China. Environ. Sci. Technol. 2018, 52, 1665–1674. 10.1021/acs.est.7b05858. [DOI] [PubMed] [Google Scholar]
  85. Rosenthal J.; Balakrishnan K.; Bruce N.; Chambers D.; Graham J.; Jack D.; Kline L.; Masera O.; Mehta S.; Mercado I. R.; Neta G.; Pattanayak S.; Puzzolo E.; Petach H.; Punturieri A.; Rubinstein A.; Sage M.; Sturke R.; Shankar A.; Sherr K.; Smith K.; Yadama G. Implementation Science to Accelerate Clean Cooking for Public Health. Environ. Health Perspect. 2017, 125, A3–A7. 10.1289/EHP1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Li X.; Clark S.; Floess E.; Baumgartner J.; Bond T.; Carter E. Personal Exposure to PM2.5 of Indoor and Outdoor Origin in Two Neighboring Chinese Communities with Contrasting Household Fuel Use Patterns. Sci. Total Environ. 2021, 800, 149421. 10.1016/j.scitotenv.2021.149421. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

es1c01368_si_001.pdf (1MB, pdf)

Articles from Environmental Science & Technology are provided here courtesy of American Chemical Society

RESOURCES