Subscriber access provided by - Access paid by the | UCSB Libraries
Characterization of Natural and Affected Environments 5
Diurnal variability and emission pattern of decamethylcyclopentasiloxane (D) from the application of personal care products in two North American cities. Matthew M. Coggon, Brian McDonald, Alexander Vlasenko, Patrick Veres, Francois Bernard, Abigail R. Koss, Bin Yuan, Jessica B. Gilman, Jeff Peischl, Kenneth C. Aikin, Justin DuRant, Carsten Warneke, Shao-Meng Li, and Joost A. de Gouw Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b00506 • Publication Date (Web): 16 Apr 2018 Downloaded from http://pubs.acs.org on April 16, 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 32
Environmental Science & Technology
Diurnal variability and emission pattern of decamethylcyclopentasiloxane (D5) from the application of personal care products in two North American cities. Matthew M. Coggon,∗,†,‡ Brian McDonald,†,‡ Alexander Vlasenko,¶,§ Patrick Veres,†,‡ François Bernard,†,‡ Abigail Koss,†,‡ Bin Yuan,†,‡ Jessica Gilman,† Jeff Peischl,†,‡ Ken Aikin,†,‡ Justin DuRant,k Carsten Warneke,∗,†,‡ Shao-Meng Li,∗,¶ and Joost de Gouw∗,‡ †NOAA Earth Systems Research Laboratory, Boulder, CO, USA ‡Cooperative Institute for Research in Environmental Sciences, Boulder, CO, USA ¶Air Quality Processes Research Section, Environment and Climate Change Canada, Toronto, Ontario, CA §now at Airzoneone Ltd. Mississauga, ON, CA kDepartment of Biology, University of South Carolina, Columbia, SC, USA E-mail:
[email protected];
[email protected];
[email protected];
[email protected] 1
2
] Abstract
3
Decamethylcyclopentasiloxane (D5 ) is a cyclic volatile methyl siloxane (cVMS) that
4
is widely used in consumer products and commonly observed in urban air. This study
1
ACS Paragon Plus Environment
Environmental Science & Technology
5
quantifies the ambient mixing ratios of D5 from ground sites in two North American
6
cities (Boulder, CO, USA and Toronto, ON, CA). From these data, we estimate the
7
diurnal emission profile of D5 in Boulder, CO. Ambient mixing ratios were consistent
8
with those measured at other urban locations; however, the diurnal pattern exhibited
9
similarities with those of traffic-related compounds such as benzene. Mobile measure-
10
ments and vehicle experiments demonstrate that emissions of D5 from personal care
11
products are coincident in time and place with emissions of benzene from motor ve-
12
hicles. During peak commuter times, the D5 /benzene ratio (wt/wt) is in excess of
13
0.3, suggesting that the mass emission rate of D5 from personal care product usage
14
is comparable to that of benzene due to traffic. The diurnal emission pattern of D5
15
is estimated using the measured D5 /benzene ratio and inventory estimates of benzene
16
emission rates in Boulder. The hourly D5 emission rate is observed to peak between
17
6:00 – 7:00 AM and subsequently follow an exponential decay with a time constant of
18
9.2 hr. This profile could be used by models to constrain temporal emission patterns
19
of personal care products.
20
Introduction
21
Cyclic volatile methyl siloxanes (cVMSs) are compounds widely used in the manufactur-
22
ing of consumer products, including shampoo, lotions, deodorants, and antiperspirants. 1,2
23
Cyclic siloxanes are also produced as intermediates for the synthesis of polydimethylsiloxane
24
(PDMS), which is used as a sealant, elastomer, and anti-foaming agent in industries such
25
as petroleum processing. 3–7 The widespread use of these compounds has led to numerous
26
government studies and academic reviews aimed at assessing the environmental and health
27
impacts of cVMS compounds, particularly octamethylcyclotetrasiloxane (D4 ), decamethyl-
28
cyclopentasiloxane (D5 ), and dodecamethylcyclohexasiloxane (D6 ). 1,6–12 Most assessments
29
conclude that D4 presents environmental risks and steps have been taken to phase out the
30
substance in personal care products; in contrast, the conclusions for D5 have been mixed. In
2
ACS Paragon Plus Environment
Page 2 of 32
Page 3 of 32
Environmental Science & Technology
31
a review by the UK Environmental Agency, 11 D5 was not identified to pose serious human
32
health effects, though it qualified as a very persistent and very bioaccumulative substance.
33
In a 2011 review, Environment Canada 10 concluded that D5 is persistent in the environment,
34
but not bioaccumulative and not a health threat to Canadian citizens. More recently, D5
35
has been found to accumulate in the livers of Atlantic cod 13,14 and a review by Rücker and
36
Kümmerer 7 has highlighted the persistence of cVMSs in different environmental media. Re-
37
cent restriction proposals by the EU Council on the Registration, Evaluation, Authorization,
38
and Restriction of Chemicals (REACH) have been implemented to limit D4 and D5 content
39
in wash-off personal care products to 0.1% by 2020. 15
40
D5 is of particular interest because its presence in the atmosphere largely results from
41
the evaporative emissions of personal care products. 6,7,16–18 Mackay et al. 6 estimates that
42
over 90% of D5 emitted to the atmosphere results from personal care product application;
43
however, residual amounts lost through product disposal may be emitted by wastewater
44
treatment plants 19 and landfills. 19–21 Release of D5 from PDMS is considered negligible. 5,6
45
Once released to the atmosphere, D5 is oxidized by OH radicals through hydrogen abstrac-
46
tion, 22 forming silanols 23 and dimeric oxidation products 24 which may partition favorably
47
into condensed phases. The global half-life for D5 is estimated to be around a week; con-
48
sequently long-range transport of D5 to rural and remote regions, including the Arctic, 25 is
49
possible. Recent atmospheric modeling of cVMS chemistry has demonstrated that siloxane
50
oxidation products could be ubiquitous throughout the US. 17 These results are consistent
51
with recent field measurements that have demonstrated a frequent occurrence of silicon-
52
containing nanoparticles in rural and urban areas. 26 Such particles are too small to originate
53
from mineral dust and are possibly secondary aerosols formed by siloxane oxidation. 24,26
54
Due to its strong association with personal care product usage and relatively long atmo-
55
spheric lifetime, D5 may act as a useful tracer for monitoring personal care product emissions
56
in indoor and outdoor environments. For example, a recent study by Tang et al. 27 demon-
57
strated that D4 , D5 , and D6 evaporated from personal care products constituted the largest
3
ACS Paragon Plus Environment
Environmental Science & Technology
58
component of indoor volatile organic compounds (VOCs) measured in an engineering class-
59
room. Recent work has suggested that personal care product emissions could significantly
60
contribute to the VOC burden in urban areas; 18,28 thus, constraining D5 emission rates,
61
and its relationship to other urban VOCs, could help to improve VOC emission invento-
62
ries. Urban D5 emission rates have been previously estimated based on modeling of ambient
63
measurements (190 mg person−1 d−1 , Chicago, IL, USA; 310 mg person−1 d−1 , Zurich,
64
Switzerland) 29 and sales of personal care products (260 mg person−1 d−1 , Canada). 6,29
65
D5 has been measured in urban (e.g. Chicago, USA, Toronto, Canada, and Zurich,
66
Switzerland) 30–32 and rural/remote atmospheres (e.g. Sweden, Arctic) 25,33,34 using low-time
67
resolution instrumentation. To date, one study has reported average D5 flux measurements
68
by high time-resolution mass spectrometry. 28 Measurements of hourly concentrations are
69
needed to evaluate the variability of D5 concentrations, infer emission mechanisms, and
70
constrain emission profiles in regional atmospheric models. Here, we report fast, real-time
71
measurements of ambient D5 in two urban, North-American locations (Boulder, Colorado,
72
USA and Toronto, Ontario, CA). From these data, we investigate D5 emission mechanisms,
73
calculate diurnal D5 emission patterns, and estimate city-wide D5 emissions for Boulder, CO.
74
Methods
75
Instrumentation
76
Measurements of VOCs in Toronto were conducted using a high mass-resolution proton-
77
transfer-reaction time-of-flight mass spectrometer (PTR-ToF-MS, Ionicon Analytik). A gen-
78
eral description of the instrument is published elsewhere. 35,36 Data were acquired on a 1
79
min accumulation period using TofDaq software (Tofwerk AG) and post-processed using the
80
PTRMS TOF Viewer software (Ionicon Analytik). Twenty background measurements were
81
conducted over the course of the campaign by passing air through a charcoal cartridge and
82
platinum catalyst heated to 350◦ C to remove VOCs. All data presented here have been 4
ACS Paragon Plus Environment
Page 4 of 32
Page 5 of 32
83
Environmental Science & Technology
background-corrected assuming a linear interpolation between zeroed samples.
84
The instrument response to a D5 standard was measured by the addition of a certified
85
gas mixture (1 ppm, Air Liquide). The sensitivity of the PTR-ToF-MS to the D5 signal at
86
the parent ion (m/z 371.10 H+ C10 H30 O5 Si5 ), was found to be 10±0.3 ncps/ppb at a reagent
87
ion signal of 1×106 cps. The instrument detection limit was found to be 2 ppt for a 10 min
88
data accumulation period.
89
Measurements in Boulder were also conducted using a PTR-ToF-MS (Tofwerk AG, Aero-
90
dyne Research Inc. ). 37 For clarity, we refer to this instrument as the NOAA PTR-ToF-MS.
91
A complete description of the instrument is provided by Yuan et al. 37 The acquired data
92
were processed following the recommendations of Stark et al. 38 using the CIMS Tofware
93
package (v. 2.5.7) in Igor Pro (WaveMetrics, Lake Oswego, OR). In this study, the NOAA
94
PTR-ToF-MS was deployed twice to measure ambient D5 concentrations (see Measurement
95
Description). In 2015, the instrument was operated using the configuration described by
96
Yuan et al.; 37 in 2017, the instrument was operated in the absence of the small segmented
97
quadrupole (SSQ). We refer to this instrumental configuration as the "modified" NOAA
98
PTR-ToF-MS. The SSQ was removed to reduce the VOC humidity dependence associated
99
with the previous instrument design. 37 All data presented here have been corrected for hu-
100
midity using the water cluster ratio (m/z 37/m/z 19), as prescribed by Yuan et al. 37 A full
101
description of this correction is provided in the Supporting Information.
102
In both Boulder deployments, data were acquired on a 30 s accumulation period. Every
103
two hours, an instrument zero was conducted by passing ambient air through a platinum
104
catalyst heated to 350◦ C. These data have been background-corrected assuming a linear
105
interpolation between zeroed samples. Every six hours, the instrument was calibrated with
106
a 10 component, gravimetrically prepared gas standard containing VOCs with known sensi-
107
tivities. The instrument response to this calibration standard was stable to within 15% over
108
the course of each deployment. These fluctuations in instrument response are influenced
109
by uncertainties in the calibration standard (20%) and normal variations in the instrument
5
ACS Paragon Plus Environment
Environmental Science & Technology
110
transmission curve. 39
111
In 2015, the NOAA PTR-ToF-MS response to D5 at the parent ion (m/z 371.10) was
112
calibrated by permeation cell followed by catalytic conversion to CO2 . 40 This method is a
113
standard technique for calibrating the response of proton-transfer-reaction mass spectrome-
114
ters to many VOCs; 37,40 however, employing this method with silicon-containing compounds
115
presents uncertainties due to the potential for catalytic poisoning by SiO2 formed during
116
the combustion process. Here, the catalyst was operated at a temperature of 350◦ C and
117
a flow rate of 14 sccm. Prior to instrument calibration, the CO2 concentration resulting
118
from D5 combustion was stable at 0.94±0.04ppm for over 6 hr. Because this method has
119
the potential to negatively bias the conversion of D5 to CO2 , the resulting calibration factor
120
(7.1±0.5 ncps/ppb at a reagent ion signal of 1.3×106 cps, Fig. S1C) should be treated as
121
an upper-limit. The limit of detection (3σ), determined based on the signal noise during
122
background measurements, was 5 ppt for a 30 s data accumulation period.
123
In 2017, the modified NOAA PTR-ToF-MS was calibrated using a home-built liquid
124
calibration unit (LCU). This method was employed to avoid the uncertainties associated
125
with catalytic conversion of D5 to CO2 . Briefly, D5 and benzene (added as a reference
126
standard) were mixed with n-hexane (used as a solvent) and injected into a steady stream
127
of N2 (∼ 0.5 L min−1 ) using a high precision syringe pump (Harvard Apparatus). The
128
temperature of the septum and injection port was maintained at 50 – 60◦ C to ensure complete
129
compound volatilization. Before entering the NOAA PTR-ToF-MS, the VOC mixture was
130
diluted by an additional stream of N2 . Calibration experiments were conducted for a variety
131
of benzene and D5 concentrations (0.01–0.1 wt%) and N2 dilution flows (1.5-3 L min−1 ),
132
resulting in a range of D5 concentrations (50 ppt – 20 ppb). This calibration method was
133
tested for VOCs routinely monitored by the NOAA PTR-ToF-MS (e.g. benzene, toluene,
134
and acetone) and was found to reproduce the sensitivity to within 5%. Benzene, toluene, and
135
acetone are typically calibrated by gravimetrically prepared gas standards with uncertainty
136
of 20%; consequently, we assume 20% uncertainty for the D5 calibration. The sensitivity of
6
ACS Paragon Plus Environment
Page 6 of 32
Page 7 of 32
Environmental Science & Technology
137
the PTR-ToF-MS to D5 was measured to be 27 ncps/ppb at a reagent ion signal of 9×106 cps
138
(Fig. S1C). The limit of detection of the modified NOAA PTR-ToF-MS was found to be 2 ppt
139
for a 30 s data accumulation period. We note that the difference in sensitivity and detection
140
limits between the 2015 and 2017 measurements results from differences in instrumental
141
setups. Despite these differences, the D5 /benzene ratio, which we use to constrain the D5
142
emission pattern, agree to within 30% between both sampling periods (see Results Section
143
and Fig. S8).
144
We also report PTR-ToF-MS measurements of benzene (m/z 79.05), which is largely
145
associated with vehicle exhaust when measured in urban air. For both instruments, the re-
146
sponse to benzene was calibrated by the addition of certified or gravimetrically prepared gas
147
standards. We note that the calibration ranges described above do not extend to the lowest
148
ambient D5 and benzene concentrations measured (
