ACPDAtmospheric Chemistry and Physics DiscussionsACPDAtmos. Chem. Phys. Discuss.1680-7375Copernicus GmbHGöttingen, Germany10.5194/acpd-15-2055-2015Understanding atmospheric peroxyformic acid chemistry: observation, modeling and implicationLiangH.ChenZ. M.zmchen@pku.edu.cnHuangD.WuQ. Q.HuangL. B.State Key Laboratory of Environmental Simulation and Pollution
Control, College of Environmental Sciences and Engineering, Peking
University, Beijing 100871, ChinaZ. M. Chen (zmchen@pku.edu.cn)22January2015152205520841December201424December2014This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://acp.copernicus.org/articles/.htmlThe full text article is available as a PDF file from https://acp.copernicus.org/articles/.pdf
The existence and importance of peroxyformic acid (PFA) in the
atmosphere has been under controversy. We present here, for the
first time, the observation data for PFA from four field
measurements carried out in China. These data provided powerful
evidence that PFA can stay in the atmosphere, typically in dozens of
pptv level. The relationship between PFA and other detected
peroxides was examined. The results showed that PFA had a strong
positive correlation with its homolog, peroxyacetic acid, due to
their similar sources and sinks. Through an evaluation of PFA
production and removal rates, we proposed that the reactions between
peroxyformyl radical (HC(O)O2) and formaldehyde or the
hydroperoxyl radical (HO2) were likely to be the major
source and degradation into formic acid (FA) was likely to be the
major sink for PFA. Based on a box model evaluation, we proposed
that the HC(O)O2 and PFA chemistry was a major source for FA
under low NOx conditions. Furthermore, it is found that the
impact of the HC(O)O2 and PFA chemistry on radical cycling
was dependent on the yield of HC(O)O2 radical from
HC(O) + O2 reaction. When this yield exceeded
50 %, the HC(O)O2 and PFA chemistry should not be
neglected for calculating the radical budget. To make clear the
exact importance of HC(O)O2 and PFA chemistry in the
atmosphere, further kinetic, field and modeling studies are
required.
Introduction
Organic peroxyacids (OPAs, RC(O)OOH) are known to play vital roles in
both gas and condensed phases in the atmosphere. First, they enhance
the acidity of the atmospheric aqueous phase, both in their own right
and by oxidizing dissolved tetravalent sulfur (S(IV)) into sulfuric
acid or sulfate (Stein and Saylor, 2012). Second, they serve as the
terminal products of the peroxy radicals and thereby ending the
radical cycling (e.g., Niki et al., 1985). Third, the formation of
OPAs are competing with the formation of peroxyacyl nitrates (PANs,
RC(O)OONO2), which are the critical pollutants and indicators
of the photochemical smog (Phillips et al., 2013). In addition, some
high molecular weight OPAs partition into the particle phase due to
their low volatility (Kroll and Seinfeld, 2008). These OPAs can
subsequently react with carbonyls in the particle phase to form
carboxylic acids or acyl peroxyhemiacetals, which have even lower
volatility than OPAs (Ziemann and Atkinson, 2012). These
low-volatility compounds can contribute to the components of secondary
organic aerosols (SOAs).
Using different measurement techniques, a number of OPAs, such as
peroxyacetic acid (PAA), peroxypropionic acid (PPA) and methylacrylic
peroxyacid (MPA), were detected in the atmosphere in previous studies
(He et al., 2010; Zhang et al., 2012; Phillips et al., 2013). However,
the peroxyformic acid (PFA, HC(O)OOH), the OPA with the simplest
molecular structure, has never been directly detected in the
atmosphere to our knowledge. Unfortunately, the difficult generation
and easy decomposition of PFA makes it impossible to determine the
existence of PFA theoretically. On one hand, PFA has a reactive
precursor. The generation of OPAs is known to predominantly result
from the reactions between peroxyacyl radicals (RC(O)O2) and
hydroperoxyl radicals (HO2). The corresponding precursor of
PFA in this reaction, i.e., the peroxyformyl radical
(HC(O)O2), was thought either not generated or rapidly
decomposed to yield CO and HO2 (Langford and Moore, 1984), as
shown in Reaction (R1). On the other hand, PFA itself can rapidly
decompose, in both gas and aqueous phases. Gaseous PFA can
spontaneously decompose into formic acid (FA, HC(O)OH) through
Reaction (R2), even at low temperature (Maker et al., 1977). When PFA
enters the aqueous phase, it can readily isomerize to carbonic acid
and decompose into CO2 and H2O through Reaction (R3).
HC(O)O2→CO+HO22HC(O)OOH→2HC(O)OH+O2HC(O)OOH(aq)→H2CO3(aq)→CO2+H2O
As a consequence of the non-detection, difficult generation and rapid
decomposition of PFA in the atmosphere, PFA-related chemistry has been
ignored by most of the current chemical mechanisms. However, the
conclusion that PFA does not exist in the atmosphere is in fact
imprudent. Actually, the ignorance of PFA-related chemistry may bring
potentially important impact on the predictions of radicals and other
trace gases. Therefore, the purpose of the current paper is to
demonstrate the existence of atmospheric PFA on the basis of a number
of specific observations, to report the concentrations and variations
of atmospheric PFA, to give a preliminary description to its
atmospheric budget and behaviors, and to discuss the possible
implications of PFA and HC(O)O2 related chemistry on
atmospheric chemistry.
MethodologySite description
Measurements for atmospheric peroxides were conducted separately at
two locations, namely, Peking University (PKU) Site and Gucheng (GC)
Site.
The PKU site (39.99∘ N, 116.30∘ E) is a typical
urban observation site located in the campus of Peking University,
northwest downtown of the Beijing city. A series of peroxide
measurements have been performed in this site since 2006 (He et al.,
2010; Zhang et al., 2010, 2012; Liang et al., 2013). The campus of PKU
is surrounded by two traffic thoroughfares where vehicles keep flowing
continuously. The instrumentations were mounted on the top of
a six-story building, with the sampling inlet ∼26m
above the ground. Three peroxide measurements, namely BJ-2012S (11
August–4 September 2012), BJ-2012F (18–28 September) and BJ-2013
(8–27 June 2013), were carried out in this site.
The GC Site (39.16∘ N, 115.74∘ E) is a rural site
located in the central meteorological bureau (CMB) farm in the North
China Plain, ∼100km southwest of the Beijing city. The
instrumentations were mounted in a container, which was surrounded by
vast wheat fields. The sampling inlet was ∼5m above the
ground. One peroxide measurement, i.e. GC-2013 (10–27 June 2013),
was carried out in this site.
Sampling and analyzing method
The sampler used to collect the peroxide components in the ambient air
is a scrubbing coil at ∼4∘C, which is a glass made
coil with an effective length of ∼100cm and an inside
diameter of 2 mm. The ambient air is pumped into this sampler
with a flow rate of 2.7 slm (standard liter per minute), to ensure the
air resides no more than 2 s in the sampler. In the sampler,
the air is mixed with the eluent,
5×10-3MH3PO4 in water, which is delivered
into the coil by a peristaltic pump at 0.2 mLmin-1. Then
the eluent, mixed with the sampled peroxide component, is delivered
into the high performance liquid chromatography (HPLC) system for detection. The residence time in the coil for
the eluent is estimated to be ∼1min.
The analyzing method for peroxides is the HPLC post column
derivatization method, which means that peroxides are derived into
fluorescent matters for detection after separated in the HPLC
column. Once collected, samples are injected into the HPLC system,
carried by the mobile phase, separated in the column and derived for
fluorescence detection. The mobile phase for the HPLC system is
5×10-3MH3PO4 with a flow rate of
0.5 mLmin-1. The derivatization reagent is a mixture of
Hemin and PHPAA (para-hydroxyphenylacetic acid). PHPAA reacts with
peroxides to form the fluorescent matter and Hemin serves as the
catalyst for the reaction. The derivatization reaction takes place in
the ∼3m Teflon tube at a temperature of ∼40∘C. More details about the sampling and analyzing method
for peroxides can be found in Hua et al. (2008).
The peroxide components are separated by an HPLC column, in which the
retention time for PFA chromatographic peak is ∼8.6min
(Huang et al., 2013), between the peaks of hydroxymethyl hydroperoxide
(HMHP, ∼7.1min) and Bis-hydroxymethyl peroxide (BHMP,
∼8.8min). The relative position of PFA peak coincides
with that reported in Kok et al. (1995).
PFA is unlikely to be significantly produced in the analytical
system. On one hand, the C1 precursor for PFA, HCHO (formaldehyde),
was not added in the eluent. On the other hand, H2O2+HCHO reaction is not able to significantly produce PFA in
solution, which is demonstrated by our condition experiments.
Measurement method for other trace gases
Besides peroxides, other trace gases including SO2, CO, NO,
NO2 and O3, were monitored simultaneously during each
measurement, using a set of corresponding online analyzers (Thermo
42i, 43i, 48i, and 49i analyzers for NO-NO2-NOx,
SO2, CO and O3, respectively). In addition, mass
concentrations of PM2.5 were determined by a TEOM 1400a
analyzer. The time resolution for the trace gases and PM2.5
were 1 min. At PKU site, meteorological parameters (i.e.,
ambient temperature, relative humidity, wind direction and wind speed)
were recorded continuously by a weather station (Met One Instruments
Inc., USA). At GC site, meteorological parameters were not recorded.
Instead, we use the 1 h average data provided by local weather
station.
Modeling method
A modeling approach was performed with a box model based on Version 6
of the Carbon Bond (CB6) Mechanism (Yarwood et al., 2010). The
original CB6 Mechanism contains 77 gas phase species and 218 gas phase
reactions (including photolysis reactions). In this study, the
mechanism is extended by PFA and HC(O)O2 radical related
chemistry (2 new species and 21 reactions). The reaction equations are
given in detail in Table S1 in the Supplement. The rates of these
reactions were assumed to be the same with the corresponding reactions
of PAA and peroxyacetyl radical. The initial concentrations of long
lived species were set as follows: 1 ppmv CO,
2 ppmvCH4, 30 ppbvO3 and 5 ppbv HCHO. Then
the model received emissions of NO and lumped volatile organic
compounds (VOCs) and these emissions were maintained throughout the
whole model runs. The emission rates of VOCs were calculated based on
the VOC emission ratio vs. CO in Beijing reported by Yuan
et al. (2012) and a CO emission rate of
2×1013moleculescm-2s-1. The calculated VOC
emission rates are listed in Table S2 in the Supplement. NO emission
rate was fixed constant throughout a single run and varied over the
range from 6.0×109moleculescm-2s-1 to
1.2×1012moleculescm-2s-1 in different
model cases. The photolysis frequencies of trace gases were calculated
with a Tropospheric Ultraviolet and Visible (TUV, version 5.0;
Madronich, 2002) radiation model. In this study, the box model is
applied to examine the influence of PFA and HC(O)O2 radical
chemistry on the production of formic acid and the budget of
ROx radicals. The physical losses of trace gases (i.e.,
deposition and phase transfer) and the changes of the planet boundary
layer height were not included in the model runs, thereby enabling the
results directly reflecting the upper limit of the potential impact of
the mechanistic changes.
ObservationsPFA levels
Gas phase PFA concentration data were collected from four field
measurements, i.e., BJ-2012S, BJ-2012F, BJ-2013 and GC-2013, as is
shown in Fig. 1. PAA concentration data were also presented in Fig. 1
as reference. The detection limit for PFA and PAA was estimated to be
10 pptv (parts per trillion by volume). It should be noted
that data under detection limit (D.L.) were not replaced in Fig. 1 in
order to better reveal the variation trend, whereas these data were
replaced with D.L./2 (i.e., 7.1 pptv) for statistical
calculation. The mean (±SD) concentrations of PFA were 21±19pptv for BJ-2012S, 10±7pptv for BJ-2012F,
31±37pptv for BJ-2013 and 12±13pptv for
GC-2013. It can be seen that PFA concentration showed an evident
seasonal and spatial variation. On one hand, PFA showed much higher
level in summer (BJ-2012S) than in fall (BJ-2012F) in the year 2012,
indicating that PFA is a photochemical product subject to the solar
radiation intensity. On the other hand, PFA observed at PKU site was
much higher than that observed at GC site at the same time, suggesting
that PFA was more readily formed and stayed in polluted area. The typical
concentration of PFA in the atmosphere, as is seen, was within the
range from several pptv to dozens of pptv, which contributed only
a minor part to the total detected peroxides. However, its total
production can be comparable with other peroxides if the rapid
degradation of PFA was taken into consideration. It was also found
that PFA and PAA were the most often detected peroxides over PKU and
GC sites. They were detected even at the time when H2O2 and
methyl hydroperoxide (MHP) were not detected. Like other peroxides, PFA had an obvious daily
cycle, with a rapid growth commencing from ∼10:00 LT and a peak concentration in the
late afternoon. High concentrations sometimes occurred at night,
implying gas phase photochemistry was not the exclusive source of PFA.
The Henry's law constant of PFA has never been determined in previous
studies. Due to the instability of PFA, its Henry's law constant was
not able to be determined using the previous estimation methods, e.g.,
HWINb (Meylan and Howard, 2000), SPARC (Hilal et al., 2008) and GROMHE
(Raventos-Duran et al., 2010). Here, we provide an estimation of the
Henry's law constant of PFA on the basis of the PFA observation data
in both gas phase and rainwater. PFA was determined in the rainwater
on 22 June 2013 during GC-2013 measurement. On that day, the rain kept
falling from the morning to ∼15:00 LT in the afternoon. Air and
rain samples were collected and determined alternately. Figure 2 shows
the temporal profile of PFA and PAA in the gas phase and rainwater. It
can be seen that PFA and PAA had a similar time variation in both gas
and aqueous phase, indicating their sources and sinks were
similar. OPAs in the rainwater generally come from two sources: mass
transfer from the gas phase and aqueous phase production. In our
previous study (Liang et al., 2013), we have concluded that the
partitioning of PAA deviated apparently from the Henry's law
equilibrium due to the aqueous phase reaction. Given the similar
temporal profile with PAA in both phases, it can be inferred that the
gas-aqueous phase partitioning of PFA also deviates from the Henry's
law equilibrium. Considering the similar behaviors of PFA and PAA,
here we use an indirect method to estimate the Henry's law constant of
PFA, that is, to make a comparison with PAA in both gas and aqueous
phase. As is seen in Fig. 2, the average ratio of PAA to PFA
concentration in the gas phase is ∼5 while the value turns into
∼20 in rainwater. Therefore, the Henry's law constant of PFA is
estimated to be a quarter of the value of PAA, i.e. ∼210Matm-1 at 298 K. Such low Henry's law
constant implies that PFA will majorly partition into the gas phase if
it is generated in the aqueous phase. Therefore, aqueous production
may serve as a potential source for gas phase PFA. The correlations
between PFA and PAA in the gas phase will be further discussed in
Sect. 3.2.1.
Correlations with other peroxidesPFA vs. PAA
Phillips et al. (2013) has reported a strong positive correlation
between PAA and three higher molecule weight OPAs. Likewise, we
present here a strong positive correlation between PFA and PAA, as
illustrated in Fig. 3. The strong correlations between different OPAs
indicate that they behave similarly in the atmosphere: they are
produced through the oxidation of hydrocarbons and removed through
OH-initiated reactions, photolysis and dry deposition. PFA and its
precursor, the HC(O)O2 radical, are instable in the gas
phase, as mentioned in Sect. 1. Therefore, PFA concentration is
typically lower than PAA. In fact, the ratio of PFA to PAA
(PFA/PAA) in the gas phase keeps changing, within
a range from 0.05 to 1. Several factors can affect the
PFA/PAA ratio. The most important factor seems to
be the solar radiation. As is shown in Fig. 1, observed
PFA/PAA in June (BJ-2012S) was about 0.10 on
average whereas the ratio in August and September (BJ-2012F) was about
0.33 on average. This can be either owing to an easier photochemical
production of PAA or an easier photochemical removal of PFA. The same
conclusion can as well be drawn from the daily cycle of
PFA/PAA: peaking at ∼07:00 LT in the early
morning and going down continuously in the daytime. The peak value of
PFA/PAA at ∼07:00 LT, reaching ∼0.70,
is significantly higher than the value at other time. This is because
both PFA and PAA tend to be under detection limit after strong removal
overnight. As PFA and PAA share a same detection limit, their ratio
will be unity when they are both under detection limit. Another
possible factor affecting the PFA/PAA is the
NOx concentration. It is well known that when NOx is
high, the peroxyacetyl radical, CH3C(O)O2, will combine
with NO2 to yield peroxyacetyl nitrate (PAN) and thus suppress
the production of PAA. For PFA, the corresponding peroxyacyl nitrate,
i.e. peroxyformyl nitrate has been proved to be very instable and has
never been detected in the atmosphere. It is likely that the
suppression of NO2 on PFA does not exist. As a result,
a higher PFA/PAA tended to be obtained at high
NOx period.
PFA vs. MHP
In general, the concentration of MHP greatly exceeded that of PFA. The
daily cycles of PFA and MHP are usually similar as both of them are
photochemical products. The ratio of their concentrations, however,
fluctuates greatly from day to day. On some occasions when the solar
radiation intensity is relatively stable, PFA tends to be low on the
days when MHP is very high, and vice versa. For example, MHP and PFA
seemed to be negatively correlated from 30 August to 4 September 2012
during BJ-2012S, as is shown in Fig. 4. This can be owing to the fact
that the production of MHP and PFA is in competition with each
other. Both MHP and PFA are C1 peroxides and are generated from
CH3O2 reactions. The combination of CH3O2 with
HO2 favors producing MHP (R4) while the combination of
CH3O2 with NO will yield HCHO (R6 and R7) and the
subsequent reactions will result in the production of PFA. However, as
is known, reactions between CH3O2 and HO2 will also
produce certain amount of HCHO (R5). In addition, the photolysis and
OH-initiated decomposition of CH3OOH will yield HCHO and
hence PFA. Therefore, PFA and MHP have a same source and a competing
source simultaneously, leading to the fluctuating ratio of PFA to MHP.
CH3O2+HO2→CH3OOH+O2CH3O2+HO2→HCHO+H2O+O2CH3O2+NO→CH3O+NO2CH3O+O2→HCHO+HO2
PFA vs. <inline-formula><mml:math display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">H</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:msub><mml:mi mathvariant="normal">O</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
The possible conversion from H2O2 to PFA in the condensed
phase has been discussed in Sect. 3.1. There is barely any direct
correlation between PFA and H2O2 in the gas phase. Therefore,
in many situations, their temporal profiles differ
substantially. H2O2 is often under detection limit when its
heterogeneous removal is important whereas PFA seems to be free from
heterogeneous removal and can always be detected even on aerosol
polluted episodes.
PFA budgetPFA sources
Table 1 presents a summary of the known formation and removal pathways
for PFA, and provides the possible range and a best guess of the
formation and removal rates in summer urban Beijing.
The reaction between peroxy radicals and HO2 is the major
known source for various atmospheric peroxides. For PFA, its
corresponding precursor, the HC(O)O2 radical, generated from
the HC(O) +O2 Reaction (R8), was thought to be
instable and can rapidly decompose into CO and HO2 (Reaction R1). For
this reason, Reactions (R8) and (R1) have been simplified as
Reaction (R9) in most mechanisms in photochemical models. However, PFA
cannot be produced unless stabilized HC(O)O2 radicals are
formed. Therefore, the importance of this source depends on whether
and how much stabilized HC(O)O2 radicals can be produced from
the HC(O)+O2 reactions.
HC(O)+O2→HC(O)O2HC(O)+O2→CO+HO2
To resolve this issue, efforts were made by a number of previous
studies. Yang et al. (2006) proposed that the radical product from the
HC(O)+O2 reaction could have a stable HC(O)O2
structure based on the ab initio approach. Villano et al. (2010)
observed the isolated HC(O)O2 radical in a laboratory study
using negative ion photoelectron spectroscopy. The corresponding
HC(O)O2 radical in their study, however, was produced by
electron photodetachment from the peroxyformyl anion rather than
directly from the HC(O)+O2 reaction. Therefore,
there is no conclusive evidence whether stabilized HC(O)O2
radical can be produced through the HC(O)+O2
reaction. An only existing recommendation for the yield of the
stabilized HC(O)O2 radical from the HC(O)+O2
reaction can be found in Demerjian et al. (1974), in which a yield of
∼30 % was suggested to best fit the experimental data.
To calculate the rate of the HC(O)O2+HO2 reaction,
we assumed that the concentration of HC(O)O2 radical was
1/100 to 1/10 of the HO2 concentration. The average
concentration of HO2 radical in summer Beijing, was typically
(1–8) ×108moleculescm-3 in sunny days, as
reported by Liu et al. (2012). Thus we estimated the concentration of
HC(O)O2 radical as (1–80) ×106moleculescm-3. The rate constant of the
HC(O)O2+HO2 reaction was assumed to be the same
with the CH3C(O)O2+HO2 reaction. The estimated
rate range for the HC(O)O2+HO2 reaction in summer
Beijing was listed in Table 1. To verify the exact contribution of
this source, the exact concentration of the HC(O)O2 radical
and the relevant kinetic parameters are required.
This reaction was proposed in early studies of HCHO photoreactions
(e.g., Sodeau and Lee, 1981) and the oxidation of formyl radicals
(Osif and Heicklen, 1976). Like the HC(O)O2+HO2
reaction, the importance of HC(O)O2+HCHO is also
dependent on the yield of stabilized HC(O)O2 produced from the
HC(O)+O2 reaction.
HC(O)O2+HCHO→HC(O)OOH+HC(O)
The rate constant for this reaction has not been reported yet and was
assumed the same with the HO2+HCHO reaction.
Chloroethylenes ozonolysis
Another known source for PFA is the ozonolysis of chloroethylenes,
which was confirmed in laboratory studies (Niki et al., 1982). This
ozonolysis reaction was suggested to produce the H(Cl)COO radical as
a Criegee intermediate, which could form PFA via the subsequent
reactions. Due to the low concentrations of chloroethylenes in the
real atmosphere, the contribution of this reaction to PFA source is
limited.
Aqueous photolysis of methylglyoxal
The aqueous phase photolysis was speculated to be a potentially
important source for PFA by analogy with the yield of PAA from
biacetyl photolysis (Faust et al., 1997). Because methylglyoxal (MG)
has a high solubility (with a Henry's law constant of ∼3×104Matm-1, Zhou and Mopper, 1990) and is abundant in
the gas phase, the typical concentration of MG in the aqueous phase is
about 100–300 µM. The listed range of rate in Table 1 is
the result of a clear day case. In a foggy day, the production rate of
PFA in this pathway could be 2 or 3 orders of magnitude higher than
the clear day value, owing to the greatly enhanced liquid water
content compared to clear days. An uncertainty for this source is that
MG is readily to form hydrate, which may greatly suppress the
photolysis (Faust et al., 1997).
In addition to the above sources, Francisco and Eisfeld (2009)
proposed that the OH-initiated reaction of hydroxymethyl hydroperoxide
(HMHP) could lead to the production of PFA with a small yield. Because
HMHP was usually not detected or at a very low level in summer Beijing
under most conditions, we presumed this reaction was of little
importance compared to other sources.
According to the range of rates in Table 1, gas phase reactions, in
particular HC(O)O2+HO2 and
HC(O)O2+HCHO reactions are likely to be the major
sources for PFA in urban Beijing.
PFA sinks
As reported in the literatures (e.g., Lee et al., 2000; Reeves and
Penkett, 2003), photolysis, OH-reactions and depositions are three
major removal pathways for peroxides. Therefore, we first give an
estimation of the rates of these pathways.
Photolysis
The rate constants for the photolysis of PFA are estimated by analogy
with the corresponding reactions of PAA.
OH-reaction
The rate constants for the OH-reaction of PFA are estimated by analogy
with the corresponding reactions of PAA.
Deposition
The deposition rate is the value measured for organic peroxides, which
is smaller than or equals to the measured value for H2O2.
Decomposition into formic acid
Besides the above removal pathways, the spontaneous decomposition into
formic acid is a significant loss for PFA, even at a low temperature
(Giguère and Olmos, 1952). The rate constant for this
decomposition, however, has never been investigated to our
knowledge. The best guess of the rate in Table 1 is based on the
assumption that PFA has a lifetime of 45 min, according to the
fact that PFA was not detected but FA was detected after
a 45 min reaction of HC(O) oxidation in a laboratory
study (Osif and Heicklen, 1976).
Heterogeneous reactions
The heterogeneous reaction is of potential importance to the budget of
PFA at regions with high aerosol loading, such as urban areas. The
kinetics and mechanisms for the heterogeneous production or removal of
PFA, however, have not been investigated. The observations have
provided some evidence for the importance of the heterogeneous
reactions. First, high concentrations of PFA occasionally occurred at
night, indicating there is some nighttime PFA source. The known gas
phase sources, however, are highly subject to the solar radiation and
are not able to contribute to a nighttime production. Therefore,
nighttime PFA is likely to originate from heterogeneous
reactions. Second, in our previous studies, we have proposed that
H2O2 could convert to PAA rapidly in the aqueous phase under
certain conditions (Liang et al., 2013). As PFA and PAA have a similar
variation trend in the rainwater, the conversion from H2O2 to
PFA may also exist. H2O2 is known to have abundant sources
in the condensed phase, such as the decomposition of organic
complexes, the hydrolysis of hydroxyl hydroperoxides and the
transition metal ion catalyzed HO2+O2-
reactions (Arellanes et al., 2006). Some of these sources are
independent of solar radiation, providing favorable conditions for the
conversion from H2O2 to PFA at both daytime and
nighttime. Considering its low solubility and high volatility, PFA
could be quickly released into the gas phase once produced in the
condensed phase.
Atmospheric implications of <inline-formula><mml:math display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">HC</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="normal">O</mml:mi><mml:mo>)</mml:mo><mml:msub><mml:mi mathvariant="normal">O</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> and PFA chemistry
In this section, we will employ the box model to examine the impact of
PFA and HC(O)O2 radical related chemistry on oxidant budget
and formic acid production. PFA and HC(O)O2 related reactions
listed in Table S1 are included in the box model.
As is discussed in Sect. 4.1, the HC(O)O2 radical reactions
are likely to be the major source for PFA. In addition to the PFA
producing reactions, the HC(O)O2 radical can proceed other
reaction pathways. Peroxy radicals can react with NO, NO2,
NO3, other organic peroxy radicals and O3, as is
reviewed by Orlando and Tyndall (2012) recently. Therefore, we assume
that HC(O)O2 radical could proceed the same reaction pathways
and that the rate constants are the same with the corresponding
reactions of CH3C(O)O2 radical. Aqueous phase or
heterogeneous reactions are not included in the current studies, due
to the lack of sufficient information. Because the HC(O)O2
radical production yield from HC(O)+O2 reaction is
not sure, we use a series of yield values from 10 to 80 % in
different model cases. The model results are shown in Figs. 5 and
6.
According to the current knowledge, FA is formed photochemically from
the oxidation of biogenic VOCs (majorly isoprene) and their products
(Paulot et al., 2011). In the present study, we take into account two
new production pathways for FA, i.e., the reaction of the
HC(O)O2 radical with HO2 or organic peroxy radicals
and the degradation of PFA. Figure 5 compares the FA production from
HC(O)O2 chemistry (hereafter HC(O)O2 pathway) and PFA
chemistry (hereafter PFA pathway) and the photo-oxidation of isoprene
and its products (hereafter ISO pathway). It is apparent that FA
produced from both the HC(O)O2 and PFA pathways increase
linearly with the HC(O)O2 yield increase, whereas FA produced
from the ISO pathway is nearly unaffected by the yield change. It can
be seen that the NOx emission dependences of different
pathways are different. Under the high NO emission rate condition
(≥6.0×1011moleculescm2s-1),
HC(O)O2 and PFA pathways generally make minor contribution to
the FA production, compared to the ISO pathways. Under this condition,
PFA pathway can be comparable with ISO pathway only when the
HC(O)O2 yield reaches 80 %. When the NO emission rate is
low (≤1.2×1011moleculescm2s-1),
however, the HC(O)O2 and PFA pathways will dominate over the
ISO pathways, even if the HC(O)O2 yield is only 10 %. This
result indicates that the HC(O)O2 and PFA pathways are the
major FA sources under low NOx conditions. In our modeling
cases, the maximum FA production rates from HC(O)O2 and PFA
pathways are calculated to be from 0.004 to 0.03 ppbvh-1
at a HC(O)O2 yield of from 10 to 80 %. Due to the
chemical inertness of FA, this production rate will result in the
accumulation of FA in the atmosphere. In addition to the gas phase
reactions, PFA can also react with aldehydes in the aqueous or aerosol
phase to form FA. This part of FA production was not included in the
present study, as mentioned above. Because methane and formaldehyde,
the precursors of HC(O)O2 and PFA, are existing ubiquitously
in the global atmosphere, the FA formation from HC(O)O2 and
PFA pathways is of importance in the understanding of missing
secondary sources for FA (Paulot et al., 2011).
Figure 6 shows the percentage change of OH, HO2 and O3
concentrations as well as HO2 production rate after the
inclusion of HC(O)O2 and PFA chemistry in the box
model. Interestingly, we found that the HC(O)O2 and PFA
chemistry leads to a decrease of OH and HO2 under low NO
emission conditions (≤1.2×1011moleculescm2s-1) but an increase of OH and
HO2 under high NO emission conditions (≥6.0×1011moleculescm2s-1). Because the
HC(O)+O2 Reaction (R9) is one of the major primary
sources of HO2, the addition of the HC(O)O2 radical
production pathway (R8) in the model would suppress the HO2
production. However, the formation of the HC(O)O2 radical
will not lead to the loss of total ROx (OH, HO2 and
RO2) radicals, as HC(O)O2 radical can recycle back
to HO2 and subsequently OH via its reaction with NO. The loss
of ROx radical occurs until the HC(O)O2 radical
combines with HO2 to form PFA and PFA is removed through dry
deposition or degradation into FA. Under low NO emission conditions,
the HC(O)O2 radical favors combining with HO2 radical
to form PFA, thus resulting in the net loss of ROx
radicals. Under high NO emission conditions, however, the
HC(O)O2 radical favors reacting with NO to yield HO2,
thus leading to no net loss of ROx radicals but the weakened
suppression of NO on HOx. Overall, the impact of mechanistic
change on OH and HO2 is not significant when the
HC(O)O2 yield is less than 20 %. If the HC(O)O2
yield is 50 % or higher, however, the HC(O)O2 and PFA
chemistry has to be considered to perform a better simulation for both
the concentrations and budgets of OH and HO2. In addition, the
mechanistic change will not result in a great change of ozone
concentration even if the HC(O)O2 yield is up to 80 %.
Summary and conclusions
For the first time, the observation data of peroxyformic acid (PFA)
obtained from four field measurements, namely, BJ-2012S, BJ-2012F,
BJ-2013 and GC-2013, were presented in the current study. These
measurements have made clear that PFA can stay in the atmosphere in
a certain amount, typically dozens of pptv. Like other peroxides, PFA
was photochemically produced, reaching a peak concentration in the
late afternoon. Rainwater measurements demonstrated that PFA can also
stay in the atmospheric aqueous phase. The Henry's law constant for
PFA, however, is very low, reaching only a quarter of that for
PAA. This value suggested that PFA would rapidly release into the gas
phase once formed in the aqueous phase. On the basis of an analysis of
the correlations between PFA and other peroxides, we found that PFA
had a strong positive correlation with PAA, which may be due to their
similar sources and sinks. In contrast, PFA was sometimes inversely
correlated with MHP, due to their competitive production.
Based on the comparison of possible reaction rates of each source or
sink, we proposed that the peroxyformyl radical (HC(O)O2)
reactions are the major sources and the degradation into formic acid
is the major sink for PFA. Box model calculations based on Carbon
Bond 6 (CB6) mechanism were performed to evaluate the impact of
HC(O)O2 and PFA chemistry on FA production and radical
budget. The results showed that HC(O)O2 and PFA reactions are
significant sources for FA under low NO emission conditions, with
a maximum production rate of from 0.004 to 0.030 ppbvh-1,
depending on the yield of HC(O)O2 radical from
HC(O)+O2 reaction. The neglect of HC(O)O2 and
PFA chemistry will cause a model bias of less than 10 % for OH and
HO2 concentrations, provided the HC(O)O2 radical yield
from HC(O)+O2 reaction is below 20 %. However, if
this yield reaches 50 % or higher, HC(O)O2 and PFA
chemistry should be considered for radical prediction.
The Supplement related to this article is available online at doi:10.5194/acpd-15-2055-2015-supplement.
Acknowledgements
This work was financed by the National Natural Science Foundation of
China (grants 41275125, 21190053, 21477002). The authors acknowledge
the Gucheng observation team for supporting the field peroxide
measurement. The authors wish to thank Y. S. Wu and M. Hu (Peking
University) for supporting the data of the meteorological factors
and common gases used in this publication.
ReferencesArellanes, C., Paulson, S. E., Fine, P. M., and Sioutas, C.:
Exceeding of Henry's law by hydrogen peroxide associated with urban
aerosols, Environ. Sci. Technol., 40, 4859–4866,
doi:10.1021/es0513786, 2006. Demerjian, K. L., Kerr, J. A., and Calvert, J. G.:
Advances in Environmental Science and Technology, Wiley, New York,
189 pp., 1974.Faust, B. C., Powell, K., Rao, C. J., and Anastasio, C.:
Aqueous-phase photolysis of biacetyl (an α-dicarbonyl
compound): a sink for biacetyl, and a source of acetic acid,
peroxyacetic acid, hydrogen peroxide, and the highly oxidizing
acetylperoxyl radical in aqueous aerosols, fogs, and clouds,
Atmos. Environ., 31, 497–510,
doi:10.1016/s1352-2310(96)00171-9,
1997.Francisco, J. S. and Eisfeld, W.: Atmospheric oxidation mechanism of
hydroxymethyl hydroperoxide, J. Phys. Chem. A, 113, 7593–7600,
doi:10.1021/jp901735z, 2009.Giguère, P. A. and Olmos, A. W.: A spectroscopic study
of hydrogen bonding in performic and peracetic acids, Can. J. Chem.,
30, 821–830,
doi:10.1139/v52-099, 1952.Hall, B. D. and Claiborn, C. S.: Measurements of the dry
deposition of peroxides to a Canadian boreal
forest, J. Geophys. Res., 102, 29343–29353,
doi:10.1029/97JD01113, 1997.He, S. Z., Chen, Z. M., Zhang, X., Zhao, Y., Huang, D. M.,
Zhao, J. N., Zhu, T., Hu, M., and Zeng, L. M.: Measurement of
atmospheric hydrogen peroxide and organic peroxides in Beijing
before and during the 2008 Olympic Games: chemical and physical
factors influencing their concentrations, J. Geophys. Res., 115,
D17307,
doi:10.1029/2009jd013544, 2010. Hilal, S. H., Ayyampalayam, S. N., and Carreira, L. A.:
Air-liquid partition coefficient for a diverse set of organic
compounds: Henry's law constant in water and hexadecane,
Environ. Sci. Technol., 42, 9231–9236, 2008.Hua, W., Chen, Z. M., Jie, C. Y., Kondo, Y.,
Hofzumahaus, A., Takegawa, N., Chang, C. C., Lu, K. D.,
Miyazaki, Y., Kita, K., Wang, H. L., Zhang, Y. H., and Hu, M.:
Atmospheric hydrogen peroxide and organic hydroperoxides during
PRIDE-PRD'06, China: their concentration, formation mechanism and
contribution to secondary aerosols, Atmos. Chem. Phys., 8,
6755–6773,
doi:10.5194/acp-8-6755-2008,
2008.Huang, D., Chen, Z. M., Zhao, Y., and Liang, H.: Newly
observed peroxides and the water effect on the formation and removal
of hydroxyalkyl hydroperoxides in the ozonolysis of isoprene,
Atmos. Chem. Phys., 13, 5671–5683,
doi:10.5194/acp-13-5671-2013,
2013.Kok, G. L., McLaren, S. E., and Stafflbach, T. A.: HPLC
determination of atmospheric organic
hydroperoxides, J. Atmos. Ocean. Tech., 12, 282–289,
doi:10.1175/1520-0426(1995)012<0282:HDOAOH>2.0.CO;2,
1995.Kroll, J. H. and Seinfeld, J. H.: Chemistry of secondary
organic aerosol: formation and evolution of low-volatility organics
in the atmosphere, Atmos. Environ., 42, 3593–3624,
doi:10.1016/j.atmosenv.2008.01.003, 2008.Langford, A. O. and Moore, C. B.: Reaction and relaxation
of vibrationally excited formyl radicals, J. Chem. Phys., 80,
4204–4210,
doi:10.1063/1.447251, 1984.Lee, M., Heikes, B. G., and O'Sullivan, D. W.: Hydrogen
peroxide and organic hydroperoxide in the troposphere: a review,
Atmos. Environ., 34, 3475–3494,
doi:10.1016/s1352-2310(99)00432-x, 2000.Liang, H., Chen, Z. M., Huang, D., Zhao, Y., and
Li, Z. Y.: Impacts of aerosols on the chemistry of atmospheric trace
gases: a case study of peroxides and HO2 radicals,
Atmos. Chem. Phys., 13, 11259–11276,
doi:10.5194/acp-13-11259-2013, 2013.Liu, Z., Wang, Y., Gu, D., Zhao, C., Huey, L. G.,
Stickel, R., Liao, J., Shao, M., Zhu, T., Zeng, L., Amoroso, A.,
Costabile, F., Chang, C.-C., and Liu, S.-C.: Summertime
photochemistry during CAREBeijing-2007: ROx budgets and O3
formation, Atmos. Chem. Phys., 12, 7737–7752,
doi:10.5194/acp-12-7737-2012, 2012.Madronich, S.: The Tropospheric Visible Ultra-violet
(TUV) model web page, available at:
http://www.acd.ucar.edu/TUV/ (last access: 21 October 2013),
2002. Madronich, S. and Calvert, J. G.: Permutation reactions
of organic peroxy-radicals in the troposphere, J. Geophys. Res., 95,
5697–5715, 1990.Maker, P. D., Niki, H., Savage, C. M., and
Breitenbach, L. P.: Fourier transform infrared spectrometric
determination of gaseous performic acid, Anal. Chem., 49,
1346–1347,
doi:10.1021/ac50017a016, 1977. Meylan, W. M. and Howard, P. H.: Src's epi suite, v3.20,
Syracuse Research Corporation, Syracuse, NY, 2000.Niki, H., Maker, P. D., Savage, C. M.,
Breitenbach, L. P., Martinez, R. I., and Herron, J. T.: A Fourier
transform infrared study of the gas-phase reactions of ozone with
chloroethylenes. Detection of peroxyformic acid, J. Phys. Chem.-US, 86,
1858–1861,
doi:10.1021/j100207a024, 1982.Niki, H., Maker, P. D., Savage, C. M., and
Breitenbach, L. P.: FTIR study of the kinetics and mechanism for
Cl-atom-initiated reactions of acetaldehyde, J. Phys. Chem.-US, 89,
588–591,
doi:10.1021/j100250a008, 1985.Orlando, J. J. and Tyndall, G. S.: Laboratory studies of
organic peroxy radical chemistry: an overview with emphasis on
recent issues of atmospheric significance, Chem. Soc. Rev., 41,
6294–6317,
doi:10.1039/C2cs35166h, 2012.Osif, T. L. and Heicklen, J.: Oxidation of formyl
radicals, J. Phys. Chem.-US, 80, 1526–1531,
doi:10.1021/j100555a002,
1976.Paulot, F., Wunch, D., Crounse, J. D., Toon, G. C.,
Millet, D. B., DeCarlo, P. F., Vigouroux, C., Deutscher, N. M.,
González Abad, G., Notholt, J., Warneke, T., Hannigan, J. W.,
Warneke, C., de Gouw, J. A., Dunlea, E. J., De Mazière, M.,
Griffith, D. W. T., Bernath, P., Jimenez, J. L., and
Wennberg, P. O.: Importance of secondary sources in the atmospheric
budgets of formic and acetic acids, Atmos. Chem. Phys., 11,
1989–2013,
doi:10.5194/acp-11-1989-2011,
2011.Phillips, G. J., Pouvesle, N., Thieser, J., Schuster, G.,
Axinte, R., Fischer, H., Williams, J., Lelieveld, J., and
Crowley, J. N.: Peroxyacetyl nitrate (PAN) and peroxyacetic acid
(PAA) measurements by iodide chemical ionisation mass spectrometry:
first analysis of results in the boreal forest and implications for
the measurement of PAN fluxes, Atmos. Chem. Phys., 13, 1129–1139,
doi:10.5194/acp-13-1129-2013, 2013.Raventos-Duran, T., Camredon, M., Valorso, R.,
Mouchel-Vallon, C., and Aumont, B.: Structure-activity relationships
to estimate the effective Henry's law constants of organics of
atmospheric interest, Atmos. Chem. Phys., 10, 7643–7654,
doi:10.5194/acp-10-7643-2010, 2010.Reeves, C. E. and Penkett, S. A.: Measurements of
peroxides and what they tell us, Chem. Rev., 103, 5199–5218,
doi:10.1021/cr0205053, 2003.Sander, S. P., Abbatt, J., Barker, J. R.,
Burkholder, J. B., Friedl, R. R., Golden, D. M., Huie, R. E.,
Kolb, C. E., Kurylo, M. J., Moortgat, G. K., Orkin, V. L., and
Wine, P. H.: Chemical Kinetics and Photochemical Data for Use in
Atmospheric Studies, Evaluation No. 17, JPL Publication 10-6, Jet
Propulsion Laboratory, Pasadena, CA, USA, available at:
http://jpldataeval.jpl.nasa.gov (last access:
June 2012), 2011.Sodeau, J. S. and Lee, E. K.: Photoreactions of
formaldehyde, Rev. Chem. Intermed., 4, 259–296,
doi:10.1007/BF03052417, 1981.Stein, A. F. and Saylor, R. D.: Sensitivities of sulfate
aerosol formation and oxidation pathways on the chemical mechanism
employed in simulations, Atmos. Chem. Phys., 12, 8567–8574,
doi:10.5194/acp-12-8567-2012, 2012.Villano, S. M., Eyet, N., Wren, S. W., Ellison, G. B.,
Bierbaum, V. M., and Lineberger, W. C.: Photoelectron spectroscopy
and thermochemistry of the peroxyformate anion, J. Phys. Chem. A,
114, 191–200,
doi:10.1021/jp907569w, 2010.Yang, H. C., Chen, H. L., and Ho, J. J.: Ab initio study of
intramolecular hydrogen transfer in formylperoxy
radical, J. Mol. Struct., 774, 35–41,
doi:10.1016/j.theochem.2006.07.003, 2006.Yarwood, G., Jung, J., Whitten, G. Z., Heo, G.,
Mellberg, J., and Estes, M.: Updates to the Carbon Bond mechanism
for version 6 (CB6), in: 2010 CMAS Conference, Chapel Hill, NC,
11–13 October,
avalaible at:
http://www.cmascenter.org/conference/2010/abstracts/emery_updates_carbon_2010.pdf
(last access: 22 October 2013), 2010.Yuan, B., Shao, M., de Gouw, J., Parrish, D. D., Lu, S.,
Wang, M., Zeng, L., Zhang, Q., Song, Y., Zhang, J., and Hu, M.:
Volatile organic compounds (VOCs) in urban air: how chemistry
affects the interpretation of positive matrix factorization (PMF)
analysis, J. Geophys. Res., 117, D24302,
doi:10.1029/2012jd018236,
2012.Zhang, J., Hatakeyama, S., and Akimoto, H.: Rate
constants of the reaction of ozone with trans-1,2-dichloroethene and
vinyl-chloride in air, Int. J. Chem. Kinet., 15, 655–668,
doi:10.1002/kin.550150707, 1983.Zhang, X., Chen, Z. M., He, S. Z., Hua, W., Zhao, Y., and
Li, J. L.: Peroxyacetic acid in urban and rural atmosphere:
concentration, feedback on PAN-NOx cycle and implication
on radical chemistry, Atmos. Chem. Phys., 10, 737–748,
doi:10.5194/acp-10-737-2010, 2010.Zhang, X., He, S. Z., Chen, Z. M., Zhao, Y., and Hua, W.:
Methyl hydroperoxide (CH3OOH) in urban, suburban and rural
atmosphere: ambient concentration, budget, and contribution to the
atmospheric oxidizing capacity, Atmos. Chem. Phys., 12, 8951–8962,
doi:10.5194/acp-12-8951-2012, 2012. Zhou, X. and Mopper, K.: Apparent partition coefficients
of 15 carbonyl compounds between air and seawater and between air
and freshwater; implications for air–sea exchange,
Environ. Sci. Technol., 24, 1864–1869, 1990.Ziemann, P. J. and Atkinson, R.: Kinetics, products, and
mechanisms of secondary organic aerosol formation, Chem. Soc. Rev.,
41, 6582–6605,
doi:10.1039/C2CS35122F, 2012.
An estimation of the rates of atmospheric PFA
sources and sinks in Beijing.
Sourceor sinkNo.ReactionsReferenceRange of rates (ppbvh-1)Best guess of rates (ppbvh-1)sourceP1HC(O)OO+HO2→0.4HC(O)OOHSander et al. (2011)8.2×10-5–5.2×10-16.6×10-3sourceP2HC(O)OO+HCHO→HC(O)OOHOsif and Heicklen (1976)1.8×10-5–1.41.8×10-2sourceP3chloroethylenes ozonolysisZhang et al. (1983)1.8×10-5–3.5×10-32.2×10-4sourceP4aqueous methyl glyoxal photolysisFaust et al. (1997)4.9×10-7–2.4×10-47.6×10-6sourceP5HOCH2OOH+OH→HC(O)OOHFrancisco and Eisfeld (2009)––sinkL1HC(O)OOH+hν→productsSander et al. (2011)3.6× 10-6- 3.6× 10-37.2× 10-5sinkL2HC(O)OOH+OH→productsNCAR MM (Madronich and Calvert, 1990)3.6×10-4–3.2×10-22.9×10-3sinkL3dry depositionHall and Claiborn (1997)9.0×10-5–2.9×10-27.2×10-4sinkL42HC(O)OOH→2HC(O)OH+O2––2.7×10-2unknownU1heterogeneous reactions–––
P represents the production pathways; L represents
the loss pathways; U represents that the pathway is
uncertain to be a source or sink.
Temporal profiles of PFA and PAA during four field
measurements.
PFA and PAA in the rainwater (a) and gas phase
(b) on 22 June 2013, GC-2013.
Relationship between PFA and PAA and the hourly average ratio
of PFA to PAA for all data during BJ-2012S.
Daily average MHP and PFA from 30 August to 4 September 2012
during BJ-2012S.
The production rates of formic acid (FA) from different
pathways at different NO emission rate. Black line: FA production
rates from HC(O)O2 reaction; red line: FA production rates
from PFA decomposition; blue line: FA production rates from other
pathways; green line: the sum of the above three pathways. Y: the
yield of HC(O)O2 from HC(O)+O2 reactions.
The percentage change of OH, HO2 concentrations and
HO2 production rate due to the mechanistic change at
different NO emission rates and different yield of HC(O)O2
from HC(O)+O2 reactions. Y: the yield of
HC(O)O2 from HC(O)+O2 reactions.