manuscript submitted to
JGR: Atmospheres
Journal of Geophysical Research: Atmospheres
1
Supporting Information for
2
H
2
O
2
and CH
3
OOH (MHP) in the remote atmosphere.
3
I: Global distribution and regional influences
4
Hannah M. Allen
1
, John D. Crounse
2
, Michelle J. Kim
2
, Alexander P. Teng
2
,
5
Eric A. Ray
3
,
4
, Kathryn McKain
3
,
4
, Colm Sweeney
4
, Paul O. Wennberg
2
,
5
6
1
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA
7
2
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, USA
8
3
Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado, Boulder,
9
CO, USA
10
4
Earth System Research Laboratory, National Oceanic and Atmospheric Administration, Boulder, CO,
11
USA
12
5
Division of Engineering and Applied Sciences, California Institute of Technology, Pasadena, CA, USA
13
Contents of this file
14
•
Instrument Schematic (Figure S1)
15
•
Instrument Calibration (Figure S2)
16
•
Synthesis of CF
3
OOCF
3
17
•
Potential Interference in the MHP Measurement (Figure S3)
18
•
Back Trajectory Analysis
19
•
Biomass Burning Correlations (Figures S6, S7, and S8)
20
Introduction
21
This supporting information provides further details on the analytical methods used
22
to derive data and to support conclusions from this study. The instrument schematic (Fig-
23
ure S1), instrument calibration discussion (Figure S2), and CF
3
OOCF
3
supplement the
24
CIT-CIMS description in the main paper by providing specific details on how the H
2
O
2
25
and MHP data were collected and how mixing ratios were generated from raw signal.
26
The discussion on the potential interference in the MHP measurement provides a bound
27
on the extent to which a second atmospheric chemical species may interfere with the MHP
28
measurement (Figure S3). The back trajectory analysis discussion provides further de-
29
tails on the methodology used to generate Figure 6 in the main text. The biomass burn-
30
ing correlations (Figures S6, S7, and S8) supplement Figure 5 and the discussion of the
31
influence of biomass burning on hydroperoxides by showing the relationship between H
2
O
2
32
or MHP and biomass burning tracers (HCN and CO) mentioned in the main text and
33
by describing the instrumentation used to measure CO.
34
Corresponding author: Hannah M. Allen,
hallen@caltech.edu
–1–
manuscript submitted to
JGR: Atmospheres
S1. Instrument Schematic
35
A simplified schematic of the key components of the CIT-CIMS aboard the NASA
36
DC-8 is given below. The CIT-CIMS directs air from outside the aircraft into the instru-
37
ment via a partially-stopped tapered aluminum inlet from which a fraction of the air is
38
directed perpendicularly at a high flow rate into the cabin using a rear-cut inlet port.
39
The ambient air arrives at the “Y-block” where a small flow is sub-sampled by the C-
40
ToF instrument (300-350 sccm) and the remainder is directed to the triple quadrupole
41
instrument (sub-samples at
∼
350 sccm) or an exhaust port by which the majority of the
42
air exits the aircraft. The total flow through the inlet is generally
>
1000 slm when the
43
aircraft is in flight. A small stream of air from the inlet is diverted to the ambient ze-
44
roing system that consists of a bicarbonate denuder and bicarbonate-coated nylon wool
45
filter to assess background signals at ambient water vapor concentrations. Before enter-
46
ing either the C-ToF or the triple quadrupole, the ambient air sample is diluted with N
2
47
(factor of
∼
5) in the flow tube and mixed with CF
3
O
–
in the ion-molecule mixing re-
48
gion forming product ions. The CF
3
O
–
forms from gaseous CF
3
OOCF
3
(1 ppm CF
3
OOCF
3
49
in N
2
) ionized by a cylindrical Po-210 radioactive source (NRD, Po-2021). The sample
50
then enters the mass spectrometer chamber.
51
Finally, two types of calibration gases may be used prior to or during flights: the
52
“cold cal” and the “hot cal” systems. The cold cal system is used extensively by the triple
53
quadrupole and periodically by the ToF instrument and consisted of peroxyacetic acid
54
(PAA,CH
3
CO
3
H), urea-H
2
O
2
, and isotopically labeled MHP (CD
3
OOH) at 0
◦
C to main-
55
tain a constant temperature and to slow hydroperoxide decomposition. The hot cal sys-
56
tem is contained isotopically labeled acetic acid (AA,
13
CH
3
13
COOH) and ethylene gly-
57
col ((CD
2
OH)
2
) maintained at 50
◦
C. Both cal systems are controlled to a constant pres-
58
sure (2000 mbar) and constant flows (ranging from 5–55 sccm) are maintained through
59
the use of glass critical orifices. Additional details related to this instrumentation can
60
be found in Crounse et al. (2006) and St. Clair et al. (2010).
61
Figure S1.
A simplified schematic of the key components of the CIT-CIMS aboard the NASA
DC-8 aircraft. A more detailed version of a previous iteration of the ToF schematic can be found
in Crounse et al. (2006) and of the triple quadrupole in St. Clair et al. (2010).
–2–
manuscript submitted to
JGR: Atmospheres
S2. Instrument Calibration
62
Figure S2.
Relative sensitivity of ToF signals for H
2
O
2
(
m/z
119, solid) and MHP (
m/z
133,
dashed) as a function of the water mixing ratio in the instrument flow tube (left) or as a function
of instrument flow tube temperature (right) for the range of water and temperature encountered
during the ATom campaign.
CIT-CIMS signals are sensitive to variations in temperature and water vapor in the
63
instrument’s ion-molecule reaction region and therefore must be accounted for using tem-
64
perature and water-dependent calibrations. The ToF H
2
O
2
signal (
m/z
119) was cal-
65
ibrated in the laboratory for these dependencies by introducing a known quantity of H
2
O
2
66
into the flow tube and either (1) measuring the signal under dry conditions at 298 K to
67
assess the absolute calibration; (2) introducing water vapor using a variable flow from
68
a Teflon pillow bag with a known mixing ratio of water vapor in N
2
while maintaining
69
a constant flow tube pressure (35 mbar) and temperature, to determine the water sen-
70
sitivity; or (3) varying temperature by using LN2 and heat gun to cool and heat the flow
71
tube (in lab) under low water conditions or by in-flight calibrations across various tem-
72
peratures (in field) to determine the temperature dependency. Figure S2 shows the re-
73
sults of these calibrations indicating how the ToF H
2
O
2
signal changes with water and
74
temperature. The H
2
O
2
calibrations were conducted during the ATom deployments prior
75
to each flight by the addition of a small flow of N
2
through urea-H
2
O
2
/glass-wool mix-
76
ture contained in a U-tube and temperature-controlled to 0
◦
C. Variable amounts of wa-
77
ter vapor were added to the flow tube by passing humidified N
2
through a flow controller.
78
In addition, H
2
O
2
calibrations were conducted approximately once every three hours dur-
79
ing ATom flights.
80
MHP observations from the ATom Mission were derived from the triple quadrupole
81
and the ToF instrument, which were calibrated using different methods. MHP mixing
82
ratios from the triple quadrupole, reported for the majority of the ATom deployments,
83
were calculated by comparing the ambient MHP signal (
m/z
133
→
m/z
85) to the sig-
84
nal from a standard addition of isotopically labeled MHP (CD
3
OOH,
m/z
136
→
m/z
85
85). The absolute calibration of the labeled MHP source was found in the laboratory us-
86
ing an FTIR cross section of 3.20
×
10
−
19
cm
2
molecule
−
1
at 2963.8 cm
−
1
(Niki et al.,
87
1983). During the deployments, the calibration source was kept in a diffusion vial main-
88
tained at a constant temperature (0
◦
C) and pressure (2 bar) and a constant output was
89
assumed. For ATom-4, changes in the synthetic approach to generate the reagent ion pre-
90
cursor, CF
3
OOCF
3
, combined with exclusion of new PFA tubing from the system, re-
91
sulted in instrumental
m/z
133 backgrounds that were low enough to enable the ToF
92
to report MHP. New PFA tubing has relatively high emissions of CF
3
C(O)OH, which
93
reacts with CF
3
O
–
to yield a product ion (CF
3
C(O)O – . HF,
m/z
133) and has the same
94
nominal mass as that of MHP; exclusion of new PFA tubing therefore reduces this in-
95
–3–
manuscript submitted to
JGR: Atmospheres
terference. The ToF was calibrated for MHP by comparing changes in the signal observed
96
at
m/z
136 from the isotopically-labeled MHP calibration source across changes in flow
97
tube water and temperature collected during ATom preflight and in-flight calibration pe-
98
riods. Figure S2 shows the results of these comparisons indicating how the ToF MHP
99
signal changes with water and temperature. Absolute MHP mixing ratios were found
100
by comparing the relative ToF MHP signals to calibrated Triple signals.
101
S3. Synthesis of CF
3
OOCF
3
102
Beginning with ATom-3, the CIT-CIMS switched to using CF
3
OOCF
3
reagent (pre-
103
cursor to CF
3
O
–
ion) obtained from a new source. Prior to this change, we gratefully
104
acquired CF
3
OOCF
3
from talented synthetic chemists at Clemson University (Darryl
105
DesMarteau and more recently Joseph Thrasher). This material was generated through
106
methods largely developed by DesMarteau (Des Marteau, 1970). While generally being
107
quite pure, the material contained trace levels of several sulfur species, including SO
2
108
and SF
6
, which proved difficult to remove and severely limited the CIMS ability to de-
109
tect ambient SO
2
. In an effort to overcome this limitation, we have developed a new syn-
110
thetic route to CF
3
OOCF
3
. This synthesis will be described in a separate manuscript.
111
Briefly, CF
3
OOOCF
3
is formed from the photolysis of trifluoroacetic anhydride in the
112
gas phase, collected and purified, and finally thermally converted to CF
3
OOCF
3
as has
113
been described before (Des Marteau, 1970; Hohorst et al., 1973). This results in mate-
114
rial containing more than 50 times less sulfur compounds (SO
2
and SF
6
) than the ma-
115
terial obtained from Clemson, and thus reduces the CIMS instrumental SO
2
background
116
signals from
∼
3 ambient ppbv to
<
50 pptv for dry conditions.
117
S4. Potential Interference in the MHP Measurement
118
Laboratory investigations indicate that methylene diol (mediol, HOCH
2
OH) is mea-
119
sured with high efficiency on the CIT-CIMS at the same mass as MHP. This compound
120
is detected as HOCH
2
OH
·
CF
3
O
–
observed at
m/z
133 on the ToF and at both triple
121
quadrupole masses of
m/z
133
→
m/z
85 and
m/z
133
→
m/z
133. Because MHP ap-
122
pears primarily at
m/z
133
→
m/z
85 in the triple quadrupole instrument, in theory mediol
123
can be distinguished from MHP. However, analytical challenges in obtaining a precise
124
known and stable quantity of mediol limit an accurate calibration of the instrument sen-
125
sitivity towards the diol at this time. In addition, the
m/z
133
→
m/z
133 mass has not
126
been measured over the range of temperatures and water vapor mixing ratios needed to
127
assess the variations in the relative ratio of the
m/z
85 to
m/z
133 fragments over the
128
temperature and humidity regimes sampled during ATom. Instead, in this section we use
129
estimated Henry’s Law coefficients and HCHO measurements to assess the magnitude
130
of mediol mixing ratios in the atmosphere and thus the potential for methylene diol to
131
impact the CIT-CIMS MHP ambient observations.
132
Mediol exists in the atmosphere in both the gas and aqueous phase due to its for-
133
mation from the hydration of formaldehyde. See, for example, Franco et al. (2021). This
134
conversion proceeds via:
135
HCHO
(g)
←−→
HCHO
(aq)
(1)
HCHO
(aq)
+ H
2
O
←−→
CH
2
(OH)
2(aq)
(2)
CH
2
(OH)
2(aq)
←−→
CH
2
(OH)
2(g)
(3)
The equilibration time for dissolved HCHO and methanediol is very fast (k = 2.04
×
10
5
e
−
2936
/T
136
s
−
1
for the forward reaction of Eq. 2) and in solution approximately 99% of formaled-
137
hyde has been hydrated to the diol (K =
e
3769
/T
−
5
.
494
) (Winkelman et al., 2002). How-
138
ever, the Henry’s Law coefficient for mediol has not been previously reported and there-
139
fore the extent to which Eq. 3 occurs remains uncertain. The closest analogue with a
140
measured Henry’s Law coefficient is hydroxymethyl hydroperoxide (HMHP, HOCH
2
OOH,
141
–4–
manuscript submitted to
JGR: Atmospheres
1.7
×
10
6
M atm
−
1
at 298 K), which has a reported temperature-dependent coefficient
142
of H = e
−
18
.
79+9870
/T
(Burkholder et al., 2015). Other analogues are ethylene glycol (HOC
2
H
4
OH,
143
6.6
×
10
5
M atm
−
1
at 298 K) and several higher carbon diols that have Henry’s Law co-
144
efficients in the range of 10
5
to 10
6
M atm
−
1
at 298 K.
145
Figure S3.
Estimated contribution of mediol to the
m/z
133 signal on the CIT-CIMS, rel-
ative to the MHP contribution, over the range of ambient water encountered during the ATom
campaign. The solid black line indicates the average and error bars indicate one standard devi-
ation, while the shaded region shows the range when the estimated Henry’s Law coefficient is a
factor of 10 bigger or smaller.
The extent of the potential mediol interference in the
m/z
133 signal during the
146
ATom campaign is estimated using measured HCHO mixing ratios and the reported temperature-
147
dependent Henry’s Law coefficient for HMHP (H = e
−
18
.
79+9870
/T
, Burkholder et al. (2015)),
148
assumed to be similar to that of mediol. Gas-phase HCHO and gas-phase mediol are as-
149
sumed to be in equilibrium with each other as connected by the condensed-phase con-
150
version process. HCHO was measured onboard the NASA DC-8 during the ATom cam-
151
paign using the in situ airborne formaldehyde (ISAF) instrument that employs laser in-
152
duced fluorescence (LIF) to measure atmospheric HCHO with high sensitivity. The es-
153
timated mixing ratios of gas-phase mediol in the regions of the atmosphere sampled dur-
154
ing ATom range from below the CIMS detection limit (
<
1 pptv) to a maximum of 12
155
pptv. We note, however, this estimate contains considerable uncertainty and the Henry’s
156
Law constant for mediol is not known (factor of 10 uncertainty is assumed here). As in-
157
dicated in Eq. 2, mediol is most prevalent in regions with high water vapor, which cor-
158
responds to regions with very low MHP sensitivity for the CIT-CIMS. Analytical chal-
159
lenges preclude obtaining a water-dependent calibration of mediol with the CIT-CIMS;
160
instead, we assume a scenario in which mediol interactions with the reagent ion behave
161
similarly to that of HMHP and instrument sensitivity increases with increasing water
162
vapor. Using the water-dependent calibration of HMHP as an analogue for the diol, the
163
CIT-CIMS may be up to
∼
150 times more sensitive to mediol than to MHP at the high-
164
est water vapor mixing ratios. The relative contribution of mediol to the
m/z
133 sig-
165
nal during ATom thus is estimated to vary from an average of
<
1% in the free tropo-
166
sphere to 10% or higher (max of 35%) in the marine boundary layer (Figure S3). A sen-
167
sitivity test reducing the mediol Henry’s Law coefficient by a factor of 10 indicates that
168
the contribution could be an average of 45% (max of 85%) in the marine boundary layer.
169
This interference may help explain anomalously low convective transmission of MHP ob-
170
served previously (see, for example, Barth et al. (2016)).
171
–5–
manuscript submitted to
JGR: Atmospheres
S5. Altitude Averages
172
Figure S4.
H
2
O
2
mixing ratios binned for different altitudes separated into Pacific (PO) and
Atlantic (AO) Ocean basins. The altitudes are
<
1.5 km (roughly the marine boundary layer),
between 1.5 and 8 km (roughly free troposphere), above 8 km (roughly upper troposphere, ex-
cluding stratosphere), and stratosphere (strat, found by filtering for points above 7 km in which
O
3
reached above 100 ppbv). Each box has lines at the lower quartile, median, and upper quar-
tile values with whiskers that extend to 1.5 times the interquartile range.
Figure S5.
MHP mixing ratios binned for different altitudes separated into Pacific (PO) and
Atlantic (AO) Ocean basins. See Fig. S4 for details.
–6–
manuscript submitted to
JGR: Atmospheres
S5. Back Trajectory Analysis
173
Back trajectories were calculated using the Traj3D model (Bowman, 1993; Bow-
174
man & Carrie, 2002) run with the National Centers for Environmental Predictions (NCEP)
175
Global Forecast System (GFS) 0.5
◦
by 0.5
◦
resolution meteorology. A cluster of 245 tra-
176
jectories was initialized in a cube with dimensions of 0.3
◦
longitude by 0.3
◦
latitude by
177
20 hPa pressure centered on one minute intervals on the aircraft position along the flight
178
track and run backwards for 10 days. The latitude, longitude, and pressure altitude for
179
each of the 245 trajectories were then averaged to a single latitude, longitude, and pres-
180
sure for each one minute point along the flight track.
181
S6. Biomass Burning Correlations
182
Atmospheric CO, CH
4
, and CO
2
were measured using the NOAA Picarro (G2401m,
183
Picarro, Santa Clara, CA), a commercial instrument that uses wavelength-scanned cav-
184
ity ring down spectroscopy (WS-CRDS) as a detection method. CRDS is a time-based
185
measurement employing a near-infrared laser to measure spectral properties of compounds
186
in an optical measurement cavity with an effective path length of up to 20 km. The NOAA
187
Picarro instrument on ATom was modified to have a lower cell pressure set point (80 torr
188
instead of 140 torr) as well as to have a shorter measurement interval (
∼
1.2 seconds for
189
ATom-1 and -2,
∼
2.0 seconds for ATom-3 and -4, compared with
∼
2.4 seconds origi-
190
nally) by reducing the number of CO spectroscopic peak scans. As a result, the CO mea-
191
surement is slightly less precise than in the original configuration (1
σ
of the raw 1-2 sec-
192
ond measurements was
∼
9 ppb for ATom-1 and -2 and
∼
4.5 for ATom-3 and -4). CO
193
measurements are reported with 1 Hz frequency for ATom-1 and -2 and with 0.5 Hz fre-
194
quency for ATom-3 and -4. See Crosson (2008) and Chen et al. (2013) for more details.
195
Figure S6.
Correlation between CO and H
2
O
2
for the ATom deployments, colored by lati-
tude. The strong correlation between H
2
O
2
and CO, a major biomass burning tracer, indicates
the production of H
2
O
2
in regions influenced by biomass burning emissions, primarily in the
equatorial region (latitudes of
−
20
◦
to 20
◦
) throughout all times of the year sampled.
–7–
manuscript submitted to
JGR: Atmospheres
Figure S7.
Correlation between HCN and MHP for the ATom deployments, colored by lat-
itude. MHP has a moderate to very low correlation with the biomass burning tracer HCN in
regions where the correlation of this tracer with H
2
O
2
is very high (latitudes of
−
20
◦
to 20
◦
).
However, MHP does show some enhancement with high HCN at polar latitudes (
>
60
◦
) in Au-
gust.
Figure S8.
Correlation between CO and MHP for the ATom deployments, colored by lati-
tude. MHP has moderate to low correlation with the biomass burning tracer CO in the regions
where the correlation of this tracer with H
2
O
2
is very high (latitudes of
−
20
◦
to 20
◦
).
References
196
Barth, M. C., Bela, M. M., Fried, A., Wennberg, P. O., Crounse, J. D., St. Clair,
197
J. M., . . . Cantrell, C. A.
(2016).
Convective transport and scavenging of
198
peroxides by thunderstorms observed over the central U.S. during DC3.
J.
199
Geophys. Res. Atmos.
,
121
, 4272-4295.
200
Bowman, K. P. (1993). Large-scale isentropic mixing properties of the Antarctic po-
201
lar vortex from analyzed winds.
J. Geophys. Res. Atmos.
,
98
, 23013-23027.
202
Bowman, K. P., & Carrie, G. D. (2002). The mean-meridional transport circulation
203
of the troposphere in an idealized GCM.
J. Atmos. Sci.
,
59
, 1502-1514.
204
Burkholder, J. B., Sander, S. P., Abbatt, J., Barker, J. R., Huie, R. E., Kolb, C. E.,
205
. . . Wine, P. H.
(2015).
Chemical kinetics and photochemical data for use
206
in atmospheric studies, evaluation no. 18.
Jet Propulsion Laboratory
,
JPL
207
Pupblication 15-10
.
208
Chen, H., Karion, A., Rella, C. W., Winderlich, J., Gerbig, C., Filges, A., . . . Tans,
209
P. P. (2013). Accurate measurements of carbon monoxide in humid air using
210
the cavity ring-down spectroscopy (CRDS) technique.
Atmos. Meas. Tech.
,
6
,
211
1031-1040.
212
–8–
manuscript submitted to
JGR: Atmospheres
Crosson, E. R. (2008). A cavity ring-down analyzer for measuring atmospheric levels
213
of methane, carbon dioxide, and water vapor.
Applied Phys. B
,
92
, 403-408.
214
Crounse, J. D., McKinney, K. A., Kwan, A. J., & Wennberg, P. O. (2006). Measure-
215
ment of gas-phase hydroperoxides by chemical ionization mass spectrometry.
216
Anal. Chem.
,
78
, 6726-6732.
217
Des Marteau, D. D. (1970). Reaction of bis(fluoroxy)difluoromethane with cesium
218
trifluoromethoxide.
Inorg. Chem.
,
9
, 2179-2181.
219
Franco, B., Blumenstock, T., Cho, C., Clarisse, L., Clerbaux, C., Coheur, P.-F., . . .
220
Taraborrelli, D.
(2021).
Ubiquitous atmospheric production of organic acids
221
mediated by cloud droplets.
Nature
,
593
, 233-237.
222
Hohorst, F. A., D., D. D., Anderson, L. R., Gould, D. E., & Fox, W. B. (1973). Re-
223
actions of bis(trifluoromethyl) trixoide.
J. Am. Chem. Soc.
,
95
, 3866-3869.
224
Niki, H., Maker, P. D., Savage, C. M., & Breitenback, L. P.
(1983).
A Fourier
225
transform infrared study of the kinetics and mechanism for the reaction HO +
226
CH
3
OOH.
J. Phys. Chem.
,
87
, 2190-2193.
227
St. Clair, J. M., McCabe, D. C., Crounse, J. D., Steiner, U., & Wennberg, P. O.
228
(2010). Chemical ionization tandem mass spectrometer for the in situ measure-
229
ment of methyl hydrogen peroxide.
Rev. Sci. Instru.
,
81
, 094102.
230
Winkelman, J. G. M., Voorwinde, O. K., Ottens, M., Beenackers, A. A. C. M., &
231
Janssen, L. P. B. M. (2002). Kinetics and chemical equilibrium of the hydra-
232
tion of formaldehyde.
Chem. Eng. Sci.
,
57
, 4067-4076.
233
–9–