Subscriber access provided by UNIV PRINCE EDWARD ISLAND
Article
Ethylene Glycol Emissions from On-road Vehicles Ezra C. Wood, W. Berk Knighton, Edward Charles Fortner, Scott C Herndon, Timothy B. Onasch, Jonathan P Franklin, Douglas R. Worsnop, Timothy R Dallmann, Drew Gentner, Allen H. Goldstein, and Robert A. Harley Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 20 Feb 2015 Downloaded from http://pubs.acs.org on February 21, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology 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 31
Environmental Science & Technology
1
Ethylene Glycol Emissions from On-road Vehicles
2 3 4
Ezra C. Wood1*, W. Berk Knighton2, Ed C. Fortner3, Scott C. Herndon3, Timothy B. Onasch3,
5
Jonathan P. Franklin3###, Douglas R. Worsnop3, Timothy R. Dallmann4#, Drew R. Gentner4##,
6
Allen H. Goldstein4,5, Robert A. Harley4
7 8
1
Department of Chemistry, University of Massachusetts, Amherst MA
9
2
Department of Chemistry and Biochemistry, Montana State University, Bozeman MT
10
3
Aerodyne Research, Inc., Billerica MA
11
4
Department of Civil and Environmental Engineering, University of California, Berkeley CA
12
5
Department of Environmental Science, Policy and Management, University of California,
13
Berkeley CA
14 15
*Corresponding Author.
16
[email protected]
17
413-545-4003 (phone)
18 19
# now at Center for Atmospheric Particle Studies, Carnegie Mellon University,
20
Pittsburgh PA
21
## now at Department of Chemical and Environmental Engineering, School of Forestry and
22
Environmental Studies, Yale University, New Haven CT
1
ACS Paragon Plus Environment
Environmental Science & Technology
23
### now at Department of Civil and Environmental Engineering, Massachusetts Institute of
24
Technology, Cambridge MA
2
ACS Paragon Plus Environment
Page 2 of 31
Page 3 of 31
Environmental Science & Technology
25
Abstract.
26
Ethylene glycol (HOCH2CH2OH), used as engine coolant for most on-road vehicles, is an
27
intermediate volatility organic compound (IVOC) with a high Henry’s Law Coefficient. We
28
present measurements of ethylene glycol (EG) vapor in the Caldecott Tunnel near San Francisco
29
using a proton transfer reaction mass spectrometer (PTR-MS). Ethylene glycol was detected at
30
mass to charge ratio 45 - usually interpreted as solely coming from acetaldehyde. EG
31
concentrations in bore 1 of the Caldecott Tunnel, which has a 4% uphill grade, were
32
characterized by infrequent (~once per day) events with concentrations exceeding ten times the
33
average concentration, likely from vehicles with malfunctioning engine coolant systems. Limited
34
measurements in tunnels near Houston and Boston are not conclusive regarding the presence of
35
EG in sampled air. Previous PTR-MS measurements in urban areas may have overestimated
36
acetaldehyde concentrations at times due to this interference by ethylene glycol. Estimates of EG
37
emission rates using the Caldecott Tunnel data are unrealistically high, suggesting that the
38
Caldecott data are not representative of emissions on a national or global scale. EG
39
emissions are potentially important because they can lead to the formation of secondary organic
40
aerosol following oxidation in the atmospheric aqueous phase.
41
3
ACS Paragon Plus Environment
Environmental Science & Technology
(TOC art)
42 43 44
4
ACS Paragon Plus Environment
Page 4 of 31
Page 5 of 31
Environmental Science & Technology
45
Introduction.
46
Vehicular emissions of organic compounds contribute greatly to ambient concentrations of
47
primary pollutants (e.g., volatile organic compounds, particulate matter) and to the formation of
48
secondary pollutants such as ozone and secondary organic aerosol (SOA).
49
secondary pollutants affect climate and have adverse effects on public health.
50
considerable uncertainty regarding the sources and atmospheric transformations of organic
51
aerosol: models under predict SOA concentrations and oxygen-to-carbon (O/C) ratios and over
52
predict aerosol volatility. 1, 6
1-3
Both primary and 4-5
There is
53 54
Although most vehicular organic compound emissions are from evaporation and incomplete
55
combustion of hydrocarbon fuels and lubrication oil,
56
vehicular source of organic compound emissions. The most common type of engine coolant used
57
is a 50% by volume solution of ethylene glycol (1,2-ethanediol, HOCH2CH2OH) in water. In
58
light-duty vehicles, approximately 5 to 15 liters of engine coolant are pumped at high pressure (2
59
bar) and temperature (up to 130° C) past the engine cylinders and through the vehicle’s radiator.
60
Leaks of coolant can develop in numerous places within a vehicle, e.g. the water pump’s shaft
61
seals, the engine head gaskets, the radiator cap, the rubber transport hoses, and the radiator itself.
62
Liquid EG can leak from a vehicle onto the road and subsequently evaporate or can evaporate
63
directly from a malfunctioning cooling system (e.g., as part of the visible mist emitted by an
64
overheating vehicle).
2, 7
engine coolant is another potential
65 66
Besides its use as an engine coolant, EG is also used in numerous industrial purposes such as the
67
production of polyethylene terephthalate (PET) bottles (~45% of global use)8 and for aircraft 5
ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 31
68
deicing. The US EPA’s 2011 Toxic Release Inventory lists airborne EG emissions of 0.8 Gg/yr
69
from point sources in the U.S., 9 but this is a lower limit given that not all EG-emitting facilities
70
are required to report emissions. To our knowledge, vehicular emissions of EG have not been
71
quantified in any emission inventory.
72 73
In the atmosphere, ethylene glycol can be oxidized either in the gas phase or in the aqueous
74
phase following uptake to a cloud, fog, or wet aerosol particle. Atmospheric oxidation of gaseous
75
EG by OH is rapid (rate constant 1.5 × 10-11 cm3 molecule-1 s-1)10, corresponding to an
76
atmospheric lifetime due to reaction with 106 molecules cm-3 OH of 19 hours. Glycolaldehyde
77
(HOCH2CHO), glycolic acid (HOCH2COOH), and formaldehyde are presumably the main
78
reaction products by analogy with larger diols.
79
glycolaldehyde are formaldehyde and carbon monoxide, and the main products of the reaction of
80
glycolaldehyde with OH in the gas phase are glyoxal and formaldehyde. 12 Gas-phase chemistry
81
and gas-particle partitioning of EG and its oxidation products do not lead to SOA formation.13
82
The high Henry’s Law coefficients of EG, glycolaldehyde, glyoxal, and glycolic acid (ΚH = 103
83
to 5×105 M atm-1)14-15, however, lead to facile uptake into the atmospheric aqueous phase,
84
wherein oxidation leads to highly oxidized products such as glyoxylic acid, oxalic acid
85
oligomers. 17 Following liquid water evaporation, these highly oxidized products can form SOA
86
with high oxygen/carbon ratios. 18 Due to current estimates of industrial and vehicular emissions,
87
EG is not currently considered a significant SOA precursor.
11
The major products from photolysis of
16
and
88 89
In this paper we present observations of unexpectedly high concentrations of EG in a California
90
highway tunnel and discuss possible emission mechanisms and atmospheric implications. 6
ACS Paragon Plus Environment
Page 7 of 31
Environmental Science & Technology
91 92
Experimental Methods.
93
Tunnel Sampling.
94
Pollutant concentrations in bore 1 of the Caldecott Tunnel near Oakland, CA were measured
95
from 22 July 2010 to July 28 2010 by a combination of real-time (1-second time resolution) and
96
integrated measurement devices housed in the East Fan Room or the Aerodyne Mobile
97
Laboratory (AML).
98
the roof of the tunnel. Bore 1 of the Caldecott Tunnel (1 km long) has a 4% uphill grade and the
99
preceding 4 km of road (Route 24) have grades of 4 to 6.5%. Diesel-fueled vehicles accounted
100
for 1 to 4% of the total traffic with the rest being gasoline-powered. 20 Ambient temperatures in
101
nearby Oakland ranged from 12 to 21° C. Details on the sampling configuration and
102
measurements are described elsewhere.
103
through a 1-2 µm PTFE filter and 34 meters of 0.95 cm (3/8”) inner diameter perfluoroalkoxy
104
(PFA) tubing at a flow rate of 11 standard liters per minute (SLPM). The inlet was periodically
105
flooded (overblow) upstream of the filter with dry zero air (AirGas) to monitor instrument
106
baselines. The AML also periodically measured outdoor air near the laboratory itself by
107
disconnecting the long sampling tube. Acetaldehyde (CH3CHO) and more than 200 other
108
individual volatile organic compounds (VOCs) and intermediate volatility organic compounds
109
(IVOCs) were measured in the East Fan Room by gas chromatography with mass spectrometry
110
and flame ionization detection. 21 VOC and IVOC measurements are averaged to 60 minutes and
111
have 2σ uncertainties of 10%.
112
The AML also sampled air in the Washburn tunnel outside Houston, TX during two round-trip
113
transits on 22 May 2009 as part of the Study of Houston Atmospheric Radical Precursors
19
The sampling point for all measurements was through an access grate at
20-21
Briefly, the AML sampled gas-phase compounds
7
ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 31
114
(SHARP) campaign. The AML conducted similar mobile measurements in the Central Artery I-
115
93 Tunnel in Boston, MA on 16 January 2008. Total sampling times (with a 1 m inlet) were 4
116
minutes in Houston and 9 minutes in Boston. Ambient temperatures were 21°C in Houston and
117
-5° C in Boston.
118 119
PTR-MS measurements.
120
Fast 1-second measurements of VOCs were made with an unmodified proton transfer reaction
121
mass spectrometer (PTR-MS, Ionicon) stationed in the AML. Ethylene glycol with a molecular
122
weight of 62 g mol-1 was detected at mass to charge ratio m/z = 45 (Figure 1) following a
123
dehydration fragmentation reaction (reactions 1 and 2) as first pointed out by Wisthaler et al.: 22
124 125
H3O+ + OHCH2CH2OH → (HOCH2CH2OH)H+ + H2O
(1)
126
(HOCH2CH2OH)H+ → (C2H5O)+ + H2O
(2)
127 128
Figure 1.
129
PTR-MS difference mass spectrum of air sampled from the headspace of in-service engine
130
coolant. m/z 61 and m/z 75 are not thought to be related to ethylene glycol. 8
ACS Paragon Plus Environment
Page 9 of 31
Environmental Science & Technology
131 132
CH3CHO is also detected at m/z 45, thus the total m/z 45 signal observed reflects the sum of
133
both compounds. Since CH3CHO was simultaneously measured by in-situ GC-FID, its
134
contribution to m/z 45 can easily be removed to calculate the EG concentration. Except at night,
135
the GC CH3CHO concentrations rarely exceeded 25% of the m/z 45 values. The only other
136
known contributors to m/z 45 are ethylene oxide (H2COCH2), which is an isomer of
137
acetaldehyde, and a weak interference from protonated carbon dioxide (CO2H+). Ethylene oxide
138
is an intermediate in the conversion of ethylene to ethylene glycol, and while there may be
139
industrial sources, there is no evidence that vehicles emit ethylene oxide. The high m/z 45
140
concentrations could only have been caused by ethylene oxide if there were numerous vehicles
141
transporting and leaking ethylene oxide at high emission rates in the tunnel. Endothermic charge
142
transfer to carbon dioxide forming CO2H+ has been observed to occur within the PTR-MS.
143
The interference from CO2 is small and no increases in m/z 45 were observed during spikes of
144
[CO2] of several hundred ppm in which there was little increase in [CO] or other pollutants.
145
Furthermore, CO2 would not produce the slow time response observed during the high m/z 45
146
events. Previous PTR-MS measurements in urban areas may have overestimated CH3CHO
147
concentrations at times due to this interference by ethylene glycol.
23
148 149
The PTR-MS m/z 45 measurements presented here use the instrumental response of CH3CHO,
150
quantified by dilution of a multicomponent gas standard (Apel-Riemer). In order to estimate the
151
PTR-MS sensitivity response factor for ethylene glycol, we calculate the reaction rate constants
152
for the proton transfer reaction of H3O+ with acetaldehyde and ethylene glycol using classic ion-
153
molecule theory, similar to that presented by Zhao and Zhang. 24 These calculations suggest that 9
ACS Paragon Plus Environment
Environmental Science & Technology
154
the PTR-MS sensitivity response factor for ethylene glycol is ~10% lower than that of
155
acetaldehyde. Values for the dipole moment and polarizability for acetaldehyde and ethylene
156
glycol are from the CRC Handbook of Chemistry and Physics. 25
Page 10 of 31
157 158
Concentrations determined by the PTR-MS are the difference between measurements of ambient
159
air and measurements of VOC-free air. The instrumental background of the PTR-MS was
160
checked using the dry zero air overblows and also by periodic sampling (every 67 min) of VOC-
161
free air generated with a heated platinum catalyst connected to the PTR-MS by 1 m of 1/8” OD
162
PFA tubing. After the onset of high EG events the background signal at m/z 45 increased and
163
slowly returned to normal levels indicating adsorption and subsequent desorption of EG
164
onto/from the 34 m sampling tube and the internal surfaces of the PTR-MS. This behavior was
165
evident by a diminished time response of the m/z 45 signal during dry zero air overblows and
166
when the PTR-MS switched to sampling air from its catalyst. When m/z 45 mixing ratios were
167
high (over 20 ppb), the 1/e time response in the m/z 45 signal exceeded one minute when the
168
PTR-MS switched to sampling catalyst air (102 ± 5 sec on 27 July 2010), compared to ~2
169
seconds for other VOCs measured (see Figure SI-1 in the Supporting Information). Similarly,
170
m/z 45 levels rapidly decreased during zero air overblows but remained elevated (Figure SI-2 in
171
the Supporting Information). Reversible sorption leads to errors in short time scale
172
measurements (e.g., 1 Hz) but likely cancels out when longer averaging times are considered
173
(i.e., PTR-MS measurements are likely low when [EG] is increasing and high when [EG] is
174
decreasing). The time response of the sampling system and PTR-MS to CH3CHO was
175
comparable to that other VOCs, evident from standard additions of CH3CHO (Figure SI-3 in the
176
Supporting Information). 10
ACS Paragon Plus Environment
Page 11 of 31
Environmental Science & Technology
177 178
Several additional diagnostic tests demonstrate that the PTR-MS measurements of EG were
179
indeed reflective of tunnel air and not long-term contamination of the filter, sampling tube, or
180
internal surfaces of the PTR-MS. If the high m/z 45 measurements from the tunnel (described in
181
the Results section) were solely caused by long-term desorption of EG from the sampling tube
182
when sampling humid air, then sampling outdoor air would have yielded similarly high (20+
183
ppb) m/z 45 readings. At the conclusion of the study on 27 July 2010, the PTR-MS sampled
184
outdoor ambient air using the same long sampling tube with a used filter (Figure SI-3 in the
185
Supporting Information). In contrast to the 35 ppb measured in the tunnel only 1.5 hours earlier,
186
the PTR-MS measured an m/z 45 mixing ratio of only 4 ppb, consistent with normal outdoor
187
CH3CHO mixing ratios. The low m/z 45 mixing ratios observed by the PTR-MS at night, which
188
at times agreed with the CH3CHO measurements by the GC within the measurement
189
uncertainties (see Results section), further demonstrate that the PTR-MS and sampling inlet were
190
able to “recover” from the high concentrations of EG sampled during the day.
191 192
The PTR-MS was periodically offline for calibrations and other instrumental maintenance.
193
During the seven days of sampling, the PTR-MS was online for 39 out of 84 total hours of
194
daytime measurements (07:00 – 19:00) and 70 out of 84 hours of nighttime measurements (19:00
195
– 07:00). Daytime and nighttime data were averaged to 1 sec and 14 sec, respectively. Because
196
of the uncertainty in the internal PTR-MS response to EG and the ad/desorption of EG to both
197
the long sampling tube and the internal surfaces of the PTR-MS, we estimate the 2σ uncertainty
198
of the time-averaged EG measurements as (+ 100% / - 50%). This large uncertainty does not
199
affect the conclusions of this paper. At the Caldecott tunnel, the contribution of CH3CHO (as 11
ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 31
200
measured by the GC) to the total m/z 45 concentration (usually
Article
Ethylene Glycol Emissions from On-road Vehicles Ezra C. Wood, W. Berk Knighton, Edward Charles Fortner, Scott C Herndon, Timothy B. Onasch, Jonathan P Franklin, Douglas R. Worsnop, Timothy R Dallmann, Drew Gentner, Allen H. Goldstein, and Robert A. Harley Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 20 Feb 2015 Downloaded from http://pubs.acs.org on February 21, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology 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 31
Environmental Science & Technology
1
Ethylene Glycol Emissions from On-road Vehicles
2 3 4
Ezra C. Wood1*, W. Berk Knighton2, Ed C. Fortner3, Scott C. Herndon3, Timothy B. Onasch3,
5
Jonathan P. Franklin3###, Douglas R. Worsnop3, Timothy R. Dallmann4#, Drew R. Gentner4##,
6
Allen H. Goldstein4,5, Robert A. Harley4
7 8
1
Department of Chemistry, University of Massachusetts, Amherst MA
9
2
Department of Chemistry and Biochemistry, Montana State University, Bozeman MT
10
3
Aerodyne Research, Inc., Billerica MA
11
4
Department of Civil and Environmental Engineering, University of California, Berkeley CA
12
5
Department of Environmental Science, Policy and Management, University of California,
13
Berkeley CA
14 15
*Corresponding Author.
16
[email protected]
17
413-545-4003 (phone)
18 19
# now at Center for Atmospheric Particle Studies, Carnegie Mellon University,
20
Pittsburgh PA
21
## now at Department of Chemical and Environmental Engineering, School of Forestry and
22
Environmental Studies, Yale University, New Haven CT
1
ACS Paragon Plus Environment
Environmental Science & Technology
23
### now at Department of Civil and Environmental Engineering, Massachusetts Institute of
24
Technology, Cambridge MA
2
ACS Paragon Plus Environment
Page 2 of 31
Page 3 of 31
Environmental Science & Technology
25
Abstract.
26
Ethylene glycol (HOCH2CH2OH), used as engine coolant for most on-road vehicles, is an
27
intermediate volatility organic compound (IVOC) with a high Henry’s Law Coefficient. We
28
present measurements of ethylene glycol (EG) vapor in the Caldecott Tunnel near San Francisco
29
using a proton transfer reaction mass spectrometer (PTR-MS). Ethylene glycol was detected at
30
mass to charge ratio 45 - usually interpreted as solely coming from acetaldehyde. EG
31
concentrations in bore 1 of the Caldecott Tunnel, which has a 4% uphill grade, were
32
characterized by infrequent (~once per day) events with concentrations exceeding ten times the
33
average concentration, likely from vehicles with malfunctioning engine coolant systems. Limited
34
measurements in tunnels near Houston and Boston are not conclusive regarding the presence of
35
EG in sampled air. Previous PTR-MS measurements in urban areas may have overestimated
36
acetaldehyde concentrations at times due to this interference by ethylene glycol. Estimates of EG
37
emission rates using the Caldecott Tunnel data are unrealistically high, suggesting that the
38
Caldecott data are not representative of emissions on a national or global scale. EG
39
emissions are potentially important because they can lead to the formation of secondary organic
40
aerosol following oxidation in the atmospheric aqueous phase.
41
3
ACS Paragon Plus Environment
Environmental Science & Technology
(TOC art)
42 43 44
4
ACS Paragon Plus Environment
Page 4 of 31
Page 5 of 31
Environmental Science & Technology
45
Introduction.
46
Vehicular emissions of organic compounds contribute greatly to ambient concentrations of
47
primary pollutants (e.g., volatile organic compounds, particulate matter) and to the formation of
48
secondary pollutants such as ozone and secondary organic aerosol (SOA).
49
secondary pollutants affect climate and have adverse effects on public health.
50
considerable uncertainty regarding the sources and atmospheric transformations of organic
51
aerosol: models under predict SOA concentrations and oxygen-to-carbon (O/C) ratios and over
52
predict aerosol volatility. 1, 6
1-3
Both primary and 4-5
There is
53 54
Although most vehicular organic compound emissions are from evaporation and incomplete
55
combustion of hydrocarbon fuels and lubrication oil,
56
vehicular source of organic compound emissions. The most common type of engine coolant used
57
is a 50% by volume solution of ethylene glycol (1,2-ethanediol, HOCH2CH2OH) in water. In
58
light-duty vehicles, approximately 5 to 15 liters of engine coolant are pumped at high pressure (2
59
bar) and temperature (up to 130° C) past the engine cylinders and through the vehicle’s radiator.
60
Leaks of coolant can develop in numerous places within a vehicle, e.g. the water pump’s shaft
61
seals, the engine head gaskets, the radiator cap, the rubber transport hoses, and the radiator itself.
62
Liquid EG can leak from a vehicle onto the road and subsequently evaporate or can evaporate
63
directly from a malfunctioning cooling system (e.g., as part of the visible mist emitted by an
64
overheating vehicle).
2, 7
engine coolant is another potential
65 66
Besides its use as an engine coolant, EG is also used in numerous industrial purposes such as the
67
production of polyethylene terephthalate (PET) bottles (~45% of global use)8 and for aircraft 5
ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 31
68
deicing. The US EPA’s 2011 Toxic Release Inventory lists airborne EG emissions of 0.8 Gg/yr
69
from point sources in the U.S., 9 but this is a lower limit given that not all EG-emitting facilities
70
are required to report emissions. To our knowledge, vehicular emissions of EG have not been
71
quantified in any emission inventory.
72 73
In the atmosphere, ethylene glycol can be oxidized either in the gas phase or in the aqueous
74
phase following uptake to a cloud, fog, or wet aerosol particle. Atmospheric oxidation of gaseous
75
EG by OH is rapid (rate constant 1.5 × 10-11 cm3 molecule-1 s-1)10, corresponding to an
76
atmospheric lifetime due to reaction with 106 molecules cm-3 OH of 19 hours. Glycolaldehyde
77
(HOCH2CHO), glycolic acid (HOCH2COOH), and formaldehyde are presumably the main
78
reaction products by analogy with larger diols.
79
glycolaldehyde are formaldehyde and carbon monoxide, and the main products of the reaction of
80
glycolaldehyde with OH in the gas phase are glyoxal and formaldehyde. 12 Gas-phase chemistry
81
and gas-particle partitioning of EG and its oxidation products do not lead to SOA formation.13
82
The high Henry’s Law coefficients of EG, glycolaldehyde, glyoxal, and glycolic acid (ΚH = 103
83
to 5×105 M atm-1)14-15, however, lead to facile uptake into the atmospheric aqueous phase,
84
wherein oxidation leads to highly oxidized products such as glyoxylic acid, oxalic acid
85
oligomers. 17 Following liquid water evaporation, these highly oxidized products can form SOA
86
with high oxygen/carbon ratios. 18 Due to current estimates of industrial and vehicular emissions,
87
EG is not currently considered a significant SOA precursor.
11
The major products from photolysis of
16
and
88 89
In this paper we present observations of unexpectedly high concentrations of EG in a California
90
highway tunnel and discuss possible emission mechanisms and atmospheric implications. 6
ACS Paragon Plus Environment
Page 7 of 31
Environmental Science & Technology
91 92
Experimental Methods.
93
Tunnel Sampling.
94
Pollutant concentrations in bore 1 of the Caldecott Tunnel near Oakland, CA were measured
95
from 22 July 2010 to July 28 2010 by a combination of real-time (1-second time resolution) and
96
integrated measurement devices housed in the East Fan Room or the Aerodyne Mobile
97
Laboratory (AML).
98
the roof of the tunnel. Bore 1 of the Caldecott Tunnel (1 km long) has a 4% uphill grade and the
99
preceding 4 km of road (Route 24) have grades of 4 to 6.5%. Diesel-fueled vehicles accounted
100
for 1 to 4% of the total traffic with the rest being gasoline-powered. 20 Ambient temperatures in
101
nearby Oakland ranged from 12 to 21° C. Details on the sampling configuration and
102
measurements are described elsewhere.
103
through a 1-2 µm PTFE filter and 34 meters of 0.95 cm (3/8”) inner diameter perfluoroalkoxy
104
(PFA) tubing at a flow rate of 11 standard liters per minute (SLPM). The inlet was periodically
105
flooded (overblow) upstream of the filter with dry zero air (AirGas) to monitor instrument
106
baselines. The AML also periodically measured outdoor air near the laboratory itself by
107
disconnecting the long sampling tube. Acetaldehyde (CH3CHO) and more than 200 other
108
individual volatile organic compounds (VOCs) and intermediate volatility organic compounds
109
(IVOCs) were measured in the East Fan Room by gas chromatography with mass spectrometry
110
and flame ionization detection. 21 VOC and IVOC measurements are averaged to 60 minutes and
111
have 2σ uncertainties of 10%.
112
The AML also sampled air in the Washburn tunnel outside Houston, TX during two round-trip
113
transits on 22 May 2009 as part of the Study of Houston Atmospheric Radical Precursors
19
The sampling point for all measurements was through an access grate at
20-21
Briefly, the AML sampled gas-phase compounds
7
ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 31
114
(SHARP) campaign. The AML conducted similar mobile measurements in the Central Artery I-
115
93 Tunnel in Boston, MA on 16 January 2008. Total sampling times (with a 1 m inlet) were 4
116
minutes in Houston and 9 minutes in Boston. Ambient temperatures were 21°C in Houston and
117
-5° C in Boston.
118 119
PTR-MS measurements.
120
Fast 1-second measurements of VOCs were made with an unmodified proton transfer reaction
121
mass spectrometer (PTR-MS, Ionicon) stationed in the AML. Ethylene glycol with a molecular
122
weight of 62 g mol-1 was detected at mass to charge ratio m/z = 45 (Figure 1) following a
123
dehydration fragmentation reaction (reactions 1 and 2) as first pointed out by Wisthaler et al.: 22
124 125
H3O+ + OHCH2CH2OH → (HOCH2CH2OH)H+ + H2O
(1)
126
(HOCH2CH2OH)H+ → (C2H5O)+ + H2O
(2)
127 128
Figure 1.
129
PTR-MS difference mass spectrum of air sampled from the headspace of in-service engine
130
coolant. m/z 61 and m/z 75 are not thought to be related to ethylene glycol. 8
ACS Paragon Plus Environment
Page 9 of 31
Environmental Science & Technology
131 132
CH3CHO is also detected at m/z 45, thus the total m/z 45 signal observed reflects the sum of
133
both compounds. Since CH3CHO was simultaneously measured by in-situ GC-FID, its
134
contribution to m/z 45 can easily be removed to calculate the EG concentration. Except at night,
135
the GC CH3CHO concentrations rarely exceeded 25% of the m/z 45 values. The only other
136
known contributors to m/z 45 are ethylene oxide (H2COCH2), which is an isomer of
137
acetaldehyde, and a weak interference from protonated carbon dioxide (CO2H+). Ethylene oxide
138
is an intermediate in the conversion of ethylene to ethylene glycol, and while there may be
139
industrial sources, there is no evidence that vehicles emit ethylene oxide. The high m/z 45
140
concentrations could only have been caused by ethylene oxide if there were numerous vehicles
141
transporting and leaking ethylene oxide at high emission rates in the tunnel. Endothermic charge
142
transfer to carbon dioxide forming CO2H+ has been observed to occur within the PTR-MS.
143
The interference from CO2 is small and no increases in m/z 45 were observed during spikes of
144
[CO2] of several hundred ppm in which there was little increase in [CO] or other pollutants.
145
Furthermore, CO2 would not produce the slow time response observed during the high m/z 45
146
events. Previous PTR-MS measurements in urban areas may have overestimated CH3CHO
147
concentrations at times due to this interference by ethylene glycol.
23
148 149
The PTR-MS m/z 45 measurements presented here use the instrumental response of CH3CHO,
150
quantified by dilution of a multicomponent gas standard (Apel-Riemer). In order to estimate the
151
PTR-MS sensitivity response factor for ethylene glycol, we calculate the reaction rate constants
152
for the proton transfer reaction of H3O+ with acetaldehyde and ethylene glycol using classic ion-
153
molecule theory, similar to that presented by Zhao and Zhang. 24 These calculations suggest that 9
ACS Paragon Plus Environment
Environmental Science & Technology
154
the PTR-MS sensitivity response factor for ethylene glycol is ~10% lower than that of
155
acetaldehyde. Values for the dipole moment and polarizability for acetaldehyde and ethylene
156
glycol are from the CRC Handbook of Chemistry and Physics. 25
Page 10 of 31
157 158
Concentrations determined by the PTR-MS are the difference between measurements of ambient
159
air and measurements of VOC-free air. The instrumental background of the PTR-MS was
160
checked using the dry zero air overblows and also by periodic sampling (every 67 min) of VOC-
161
free air generated with a heated platinum catalyst connected to the PTR-MS by 1 m of 1/8” OD
162
PFA tubing. After the onset of high EG events the background signal at m/z 45 increased and
163
slowly returned to normal levels indicating adsorption and subsequent desorption of EG
164
onto/from the 34 m sampling tube and the internal surfaces of the PTR-MS. This behavior was
165
evident by a diminished time response of the m/z 45 signal during dry zero air overblows and
166
when the PTR-MS switched to sampling air from its catalyst. When m/z 45 mixing ratios were
167
high (over 20 ppb), the 1/e time response in the m/z 45 signal exceeded one minute when the
168
PTR-MS switched to sampling catalyst air (102 ± 5 sec on 27 July 2010), compared to ~2
169
seconds for other VOCs measured (see Figure SI-1 in the Supporting Information). Similarly,
170
m/z 45 levels rapidly decreased during zero air overblows but remained elevated (Figure SI-2 in
171
the Supporting Information). Reversible sorption leads to errors in short time scale
172
measurements (e.g., 1 Hz) but likely cancels out when longer averaging times are considered
173
(i.e., PTR-MS measurements are likely low when [EG] is increasing and high when [EG] is
174
decreasing). The time response of the sampling system and PTR-MS to CH3CHO was
175
comparable to that other VOCs, evident from standard additions of CH3CHO (Figure SI-3 in the
176
Supporting Information). 10
ACS Paragon Plus Environment
Page 11 of 31
Environmental Science & Technology
177 178
Several additional diagnostic tests demonstrate that the PTR-MS measurements of EG were
179
indeed reflective of tunnel air and not long-term contamination of the filter, sampling tube, or
180
internal surfaces of the PTR-MS. If the high m/z 45 measurements from the tunnel (described in
181
the Results section) were solely caused by long-term desorption of EG from the sampling tube
182
when sampling humid air, then sampling outdoor air would have yielded similarly high (20+
183
ppb) m/z 45 readings. At the conclusion of the study on 27 July 2010, the PTR-MS sampled
184
outdoor ambient air using the same long sampling tube with a used filter (Figure SI-3 in the
185
Supporting Information). In contrast to the 35 ppb measured in the tunnel only 1.5 hours earlier,
186
the PTR-MS measured an m/z 45 mixing ratio of only 4 ppb, consistent with normal outdoor
187
CH3CHO mixing ratios. The low m/z 45 mixing ratios observed by the PTR-MS at night, which
188
at times agreed with the CH3CHO measurements by the GC within the measurement
189
uncertainties (see Results section), further demonstrate that the PTR-MS and sampling inlet were
190
able to “recover” from the high concentrations of EG sampled during the day.
191 192
The PTR-MS was periodically offline for calibrations and other instrumental maintenance.
193
During the seven days of sampling, the PTR-MS was online for 39 out of 84 total hours of
194
daytime measurements (07:00 – 19:00) and 70 out of 84 hours of nighttime measurements (19:00
195
– 07:00). Daytime and nighttime data were averaged to 1 sec and 14 sec, respectively. Because
196
of the uncertainty in the internal PTR-MS response to EG and the ad/desorption of EG to both
197
the long sampling tube and the internal surfaces of the PTR-MS, we estimate the 2σ uncertainty
198
of the time-averaged EG measurements as (+ 100% / - 50%). This large uncertainty does not
199
affect the conclusions of this paper. At the Caldecott tunnel, the contribution of CH3CHO (as 11
ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 31
200
measured by the GC) to the total m/z 45 concentration (usually
