Water-soluble organic carbon (WSOC) is a ubiquitous and significant fraction
of fine particulate matter. Despite advances in aerosol thermodynamic
equilibrium models, there is limited understanding on the comprehensive
impacts of WSOC on aerosol acidity (pH). We address this limitation by
studying submicron aerosols that represent the two extremes in acidity levels
found in the atmosphere: strongly acidic aerosol from Baltimore, MD, and
weakly acidic conditions characteristic of Beijing, China. These cases are
then used to construct mixed inorganic–organic single-phase aqueous
particles and thermodynamically analyzed by the Extended Aerosol Inorganics Model (E-AIM) and ISORROPIA models
in combination with activity coefficient model AIOMFAC (Aerosol Inorganic–Organic
Mixtures Functional groups Activity Coefficient) to evaluate the
effects of WSOC on the H+ ion activity coefficients (γH+)
and activity (pH). We find that addition of organic acids and nonacid
organic species concurrently increases γH+ and aerosol liquid
water. Under the highly acidic conditions typical of the eastern US
(inorganic-only pH ∼1), these effects mostly offset each
other, giving pH changes of < 0.5 pH units even at organic aerosol
dry mass fractions in excess of 60 %. Under conditions with weaker acidity
typical of Beijing (inorganic-only pH ∼4.5), the nonacidic
WSOC compounds had similarly minor effects on aerosol pH, but organic acids
imparted the largest changes in pH compared to the inorganic-only
simulations. Organic acids affect pH in the order of their pKa values
(oxalic acid > malonic acid > glutaric acid). Although
the inorganic-only pH was above the pKa value of all three organic
acids investigated, pH changes in excess of 1 pH unit were only observed at
unrealistic organic acid levels (aerosol organic acid concentrations
> 35 µg m-3) in Beijing. The model simulations were
run at 70 %, 80 %, and 90 % relative humidity (RH) levels and the
effect of WSOC was inversely related to RH. At 90 % RH, WSOC altered
aerosol pH by up to ∼0.2 pH units, though the effect was up
to ∼0.6 pH units at 70 % RH. The somewhat offsetting nature
of these effects suggests that aerosol pH is sufficiently constrained by the
inorganic constituents alone under conditions where liquid–liquid phase
separation is not anticipated to occur.
Introduction
The acidity of atmospheric particles plays a critical role in many
physicochemical processes. Some of these processes include sulfur oxidation
and halogen chemistry, with important implications for the formation of
sulfates (Chameides, 1984); the oxidation of volatile organic
compounds (VOCs) and ozone formation in marine environments (Keene et
al., 1998); the gas–particle partitioning of many semivolatile species
(Ahrens et al., 2012; Keene et al., 2004); and enhancements to secondary
organic aerosol (SOA) formation
(Hallquist et al., 2009). The
inorganic salt constituents in atmospheric particles, such as ammonium
sulfate ((NH4)2SO4) and ammonium bisulfate
(NH4HSO4), contribute to particle acidity and water content, with
effects on aerosol radiative forcing (Seinfeld and Pandis, 2016). In
addition to the physicochemical effects within particles, their bulk acidity
can affect the health of both environmental ecosystems and the human
populations therein either directly (Gwynn et al., 2000; Peters et al.,
1996; Schindler, 1988; Spengler et al., 1996; Fang et al., 2017; Johnson et
al., 2008) or by their effects on nutrient deposition (Myriokefalitakis
et al., 2016, 2018; Kanakidou et al., 2016; Nenes
et al., 2011).
pH, the parameter serving to define and describe the acidity of aqueous
solutions, often has no direct correlation with proxy measurement methods
such as aerosol strong acidity (H+ contributed by strong acids that
dissociated completely at any pH level) or aerosol total acidity
(dissociated H+ and undissociated H+ bound to weak acids)
(Hennigan et al., 2015; Song et al., 2018b). The most accurate
predictions of aerosol pH come from aerosol thermodynamic equilibrium models
constrained by both aerosol- and gas-phase measurements (i.e., forward
mode calculations) or from the measured gas–particle partitioning of
semivolatile species, including ammonia, nitric acid, or oxalic acid, which
provide direct insight to pH (Hennigan et al., 2015).
Both approaches utilize aerosol- and gas-phase composition measurements,
along with the temperature and relative humidity (RH), to obtain aerosol pH
values. Consistent pH values are obtained when the assumptions about aerosol
mixing and equilibrium are met (Guo et al., 2018a, b).
Different aerosol thermodynamic equilibrium models have been developed
through the years, each with a unique set of assumptions, simplifications,
and approach to obtain the composition at thermodynamic equilibrium. The
Extended Aerosol Inorganics Model (E-AIM, http://www.aim.env.uea.ac.uk/aim/aim.php, last access: 28 September 2019) (Wexler and Clegg, 2002;
Friese and Ebel, 2010) and the ISORROPIA-II model (Greek for “equilibrium”,
https://www.epfl.ch/labs/lapi/isorropia/, last access: 28 September 2019) (Fountoukis and Nenes,
2007) are widely used to calculate aerosol pH for atmospheric and
experimental particle distributions (Guo et al., 2015, 2016, 2017;
Wang et al., 2016). The Aerosol Inorganic–Organic
Mixtures Functional groups Activity Coefficient (AIOMFAC) model (http://www.aiomfac.caltech.edu/model.html, last access: 28 September 2019) offers the most extensive
treatment of organic–inorganic interactions (Zuend et al., 2008, 2011) of models to date but is primarily an activity coefficient model
that does not solve full thermodynamic equilibrium calculations or phase
partitioning as E-AIM and ISORROPIA do. At present, E-AIM, ISORROPIA, and
AIOMFAC are widely used for atmospheric applications due to their
demonstrated predictive capabilities and their accessibility: they are
freely available online and include resources and user guides to facilitate
their application and use.
One key difference among the models is their treatment of organics.
ISORROPIA does not include organic species. E-AIM functions similarly to
ISORROPIA when considering inorganic species but in addition offers a
limited library of organic acids (included by UNIFAC methods or fitted
activity equations). AIOMFAC offers wide support for organic components but
is an activity coefficient model that does not solve the equilibrium
partitioning calculations for which the other models were designed. While
AIOMFAC has been used in combination with thermodynamic equilibrium models
such as ISORROPIA-II (Pye et al., 2018), these
are custom modifications to the models and not reflected in the online
versions used in this study. These models, and most others, do not treat
organics in a way that is comprehensive (that is, simultaneous consideration
of activity coefficient calculations and thermodynamic equilibrium
calculations). However, these simplified thermodynamic models do seem to
capture the partitioning of inorganic species well, even when organic
components are present in large quantities, which indicates that pH should
be captured well (Guo et al., 2018a). An additional
consideration between the models is their treatment of organic acids. E-AIM
offers support for limited (n=8) organic acid species and treats the
dissociation equilibrium of organic acids. In contrast, AIOMFAC treats
organic acids as nondissociating, a model difference that is discussed in
detail below. Note that the ion dissociation equilibria of inorganic species
(such as HSO4-/SO42-) are explicitly considered in the
equilibrium calculations of all three models employed in this study.
The effects of water-soluble organic carbon (WSOC) on aerosol pH come through two primary means: dilution
of the aqueous phase by aerosol liquid water associated with the organic
fraction (Wo) and changes to the hydrogen ion activity coefficient and
thus hydrogen ion activity in solution (γH+ and aH+,
respectively). The total contribution of organics to aerosol water can be as
much as 30 %–50 % of total fine-particle aerosol water in the polluted
Beijing winter haze events (Tan et al., 2018; Huang et al., 2014),
40 %–50 % in the southeast United States (Nah et al., 2018; Guo et al.,
2015), and the eastern Mediterranean (Bougiatioti et al., 2016). The
effects of organics (soluble and insoluble) on aerosol pH under conditions
of liquid–liquid phase separation (LLPS) are more complex (Dallemagne et al., 2016). Free H+ ion
is predicted to have increased association with SO42- to form
HSO4- when organic compounds are in the same phase as inorganic
ions, resulting in a 0.1 pH unit increase in aerosol pH
(Pye et al., 2018). The isolation of the
organic components in a separate phase (LLPS condition) also alters the
partitioning behavior of NH3, a critical component that contributes to
aerosol pH. The inclusion, or lack thereof, of organic compounds was
predicted to have a greater effect on NH3 partitioning behavior than
the inclusion, or lack thereof, of nonvolatile cations, nitrate, and
chloride (Guo et al., 2018a). In addition to these effects,
AIOMFAC predicts that any organic presence in the same phase with inorganic
constituents drives free H+ to increased association with sulfate to
form bisulfate, a compound predicted to be more miscible with organics than
H+ and small cations. AIOMFAC was used to show that the organic phase of liquid–liquid phase separated particles still contains a significant
amount of inorganics, affecting the partitioning medium by inclusion of the
inorganic ions and their associated water, lowering the mole fraction and
activity of organics, and shifting the gas–particle partitioning of organic
compounds with O:C > 0.6 (Pye et
al., 2018). In the case where multiple phases do exist, there is anticipated
to be a primarily organic (PO) and primarily inorganic (PI) phase, each of
which contains H+ in equilibrium with the other phase (Zuend and Seinfeld, 2012). In cases where
the vast majority of inorganics are partitioned to the PI phase, the pH is
not anticipated to change drastically, as H+ is also required to be in
equilibrium with the other phases, which may explain the results of
Pye et al. (2018). In the case of phase
separation where the PO phase contains considerable amounts of inorganic
species, there exists the possibility of a PI phase with
substantially altered H+ activity and, therefore, the potential for
substantially altered aerosol pH (e.g., Dallemagne et al., 2016). The
present study avoids such complexity and instead considers mixed
organic–inorganic particles present in a single aqueous phase.
Aerosol pH can also be directly affected by organic acids, whose
dissociation produces H+ ions in the particle aqueous phase. Carboxylic
acids represent a highly abundant moiety in atmospheric organic aerosol (Yatavelli et
al., 2015; Kawamura and Bikkina, 2016; Nah et al., 2018). Although these
atmospheric organic acids are typically weaker acids with higher pKa
values than common inorganic acids (H2SO4 and HNO3), they may
contribute to particle acidity in some environments
(Trebs et al., 2005). However, this effect is not
present in all environments and is constrained to situations where the pH is
in the range of the pKa of the acid in question (Nah et al., 2018;
Song et al., 2018b). As both organic acids and nonacid organic species are
expected to be present, there are competing effects within the particle:
dilution by the water fraction associated with organic constituents, direct
acidification by the dissolution of organic acids, and the change in γH+ by interactions with the additional species in solution.
Oxalic acid (measured as the oxalate ion) is often the most abundant
carboxylic acid in atmospheric aerosols (Bikkina et al., 2015). Several
studies utilized E-AIM Model IV to evaluate the effect of oxalic acid on
particle acidity (Vasilakos et al., 2018; Song et al., 2018b). With
inorganics similar in composition to that of Baltimore, conditions applied
in this study, an increase of 25 %–50 % of oxalic acid compared to the base case had an insignificant effect on aerosol pH when only one liquid phase
was present. Pye et al. (2018) utilized data from the Southern Oxidant and
Aerosol Study (SOAS) in ISORROPIA and AIOMFAC to investigate the
gas–particle partitioning of ammonia, water, and organic compounds, and how
liquid–liquid phase separation (LLPS) in particles can affect aerosol pH,
predicting a 0.7 pH unit increase when the organic fraction and its diluting
effect was considered. Generally, aerosol processes are not affected by a pH
change of this magnitude, except in the regions on the thermodynamic sigmoid
curves of semivolatile species where partitioning is shown to vary greatly
for small changes in pH (Nah et al., 2018; Guo et al., 2018b).
In this work, we explore the effects of WSOC on aerosol pH in a systematic
way by utilizing inorganic data to construct combinations of single, aqueous-phase particulate compositions and utilizing aerosol thermodynamic models
to investigate the effects of different WSOC species and concentrations on
γH+ and aH+.
MethodsData
Inorganic and meteorological data used for this study were reported in prior
work. Briefly, data from Baltimore, MD, were taken from
Battaglia et al. (2017) and include speciated inorganic
PM2.5 concentrations, meteorological data, and gas-phase NH3
measurements. The data used as thermodynamic model inputs are summertime
(July) averages based on 3 or 5 years of monitoring. All model inputs and
outputs are available at https://knb.ecoinformatics.org/ (last access: 15 October 2019)
(urn:uuid:ae58e5f1-fb13-4e9c-800b-eeb8bcb14d84).
Aerosol inorganic composition, gas-phase NH3 measurements, and
meteorological parameters were obtained during a study of winter haze
formation in Beijing, China, in 2015 (Wang et al., 2016). These data
represent a contrast with Baltimore due to different source contributions,
differences in NH3 concentrations, T, RH, and inorganic aerosol levels.
The inorganic PM2.5 concentrations, and averaged seasonal T and RH,
along with NH3 gas concentration values were obtained as model-ready inputs of the Beijing winter haze data from Guo et al. (2018b), based on supplemental information from Wang et al. (2016).
Summary of model runs performed for the study. Inorganic
composition is invariant and taken from the sources provided. Organic
components were added in factorial fashion.
LocationInorganic dataInorganic equilibrium modelOrganic componentsNo. of pointsBaltimoreBattaglia et al. (2017)E-AIM Model IVOrganic acids1331BaltimoreBattaglia et al. (2017)E-AIM Model IVNonacid organics1331BeijingGuo et al. (2018b)ISORROPIA 2.3Organic acids1331E-AIM Model IVBeijingGuo et al. (2018b)ISORROPIA 2.3Nonacid organics1331E-AIM Model IVGeneral approach
The general approach to this study was to utilize the inorganic PM and
NH3 data described above, in combination with various additional WSOC
constituents, as inputs to aerosol thermodynamic equilibrium models to
investigate the effects on model-predicted aerosol pH and γH+.
Inorganic data were modeled in either E-AIM IV or ISORROPIA-II to obtain
equilibrium concentrations of aerosol liquid water (ALW) along with all
inorganic aerosol ionic species. Organic constituents were then added to
this invariant inorganic matrix (assuming the added organic mass was at
equilibrium), at identical T and RH, and the resulting particle compositions
were modeled in AIOMFAC to obtain aerosol H+ ion activity (aH+ and
γH+) and thus aerosol pH. The average inorganic composition,
gas-phase NH3, and meteorological conditions were held approximately
constant for each location, while WSOC composition and concentrations were
systematically varied. A matrix was constructed to examine multiple
combinations of the selected organic component composition levels (factorial
design), and their effects were evaluated on the basis of the organic-to-inorganic
ratio (OIR) or organic mass fraction, both computed on a dry particle basis.
This full factorial design consists of three factors for each acid or
nonacid condition (the identity of each species), each with discrete
possible values (air concentrations in µg m-3), where the
experiment incorporates all possible combinations of these values across all
factors (Keppel, 1991). For each location, this resulted in a total of
7986 model simulations, with 1331 simulations run for both cases of
organic compounds selected, and at each of three distinct RH levels, as
described below. A summary of the models run for each location is shown in
Table 1. The RH in all simulations was fixed at either ∼70%, ∼80 %, or ∼90 %, with inorganic
system inputs calculated and invariant at each RH level based on the initial
input data from either Baltimore or Beijing to ensure deliquescence of
inorganic aerosol particles, to understand the sensitivity of the
model-predicted aerosol pH to changes in RH (ALW), and to avoid
liquid–liquid phase separation as a potential cause of organic-influenced
aerosol pH changes (Pye et al., 2018). For all
of the results presented in this analysis, aerosol pH was computed as the
negative base-10 logarithm of the hydrogen ion activity taken from the
E-AIM or AIOMFAC output (pH =-log10aH+) on a molality basis.
Organic constituents
Water-soluble organic compounds were selected by broadly classifying them as
organic acids or nonacid organics. Within each category, three individual
species were selected based on their detection in atmospheric particles and
their availability in the predefined list of AIOMFAC organic species
available on the AIOMFAC web interface, or the ability to reasonably
construct them using the functional groups approach of AIOMFAC. In addition,
nonacid organics were selected from three different primary moiety groups
from among the AIOMFAC standard species. Oxalic acid (C2H2O4
pKa1=1.23, pKa2=4.19) (Lide, 1994), glutaric acid
(C5H8O4, pKa1=4.31, pKa2=5.41) (Lide,
1994), and malonic acid (C3H3O4, pKa1=2.83,
pKa2=5.69) (Lide, 1994) were selected as the three dicarboxylic
acid species. Levoglucosan (C6H10O5), tetrahydrofuran
((CH2)4O), and 2-methyltetrol (1-methylbutane-1,2,3,4-tetrol,
C5H12O4), three organic species observed in ambient aerosols,
were selected as the nonacid WSOC species. Concentration levels were not
constrained by observations but were instead selected to achieve similar
organic-to-inorganic mass ratios for each of the two geographic regions
being considered. For Beijing, typical organic mass fractions can be on the
order of 50 %–70 % of total aerosol mass (Zhou et
al., 2018) and 20 %–60 % of total aerosol mass for continental midlatitude locations like Baltimore (Carlton et al., 2009). For each
geographic region, 11 different concentrations were chosen for each WSOC
compound (0–4 µg m-3 for Baltimore; 0–40 µg m-3 for Beijing) and combined in factorial fashion: each organic acid
concentration level combination of the three organic acids was examined in
combination with every other level of the remaining two and vice versa for
the nonacid organic species. Combinations of organic acids and nonacid
organic species were not explicitly considered here; only combinations of
organic acids with organic acids or combinations of nonacids with
nonacids were examined experimentally. All model inputs and outputs are
available at https://knb.ecoinformatics.org/ (last access: 15 October 2019).
Thermodynamic model input configuration and equilibrium model evaluations
E-AIM Model IV provides thermodynamic equilibrium modeling of the
H+-NH4+-Na+-SO42--NO3--Cl--H2O
system at temperatures from 263.15 to 330 K for subsaturated systems that
contain NH4+ and Cl-, or Na+ in combination with other
ions (Friese and Ebel, 2010). Data for Baltimore and Beijing were
formatted for E-AIM input in the following ways: average inorganic species
concentrations (µg m-3) were converted to moles per cubic meter; the
average daily temperature for the same period was used as the temperature
input; and the relative humidity of the system was fixed (at 70 %, 80 %, or
90 %) both to ensure the inorganic system was in a deliquesced state and
because of the RH restrictions (subsaturated solution requirements, RH > 0.6) on E-AIM Model IV inputs. In addition to fixing system RH
at 70 %, 80 %, or 90 %, the aerosol metastable (solid precipitate
formation disabled) mode was enforced on the model by disabling the
formation of all solids in the model input matrix, according to the analysis
and recommendation of Guo et al. (2018b, 2015). Crustal species (Ca2+,
Mg2+, K+) not supported by the model were not considered, and the
persistent cation deficiency was corrected by the addition of H+ to the
system to ensure electroneutrality. The amount of H+ added was a
not-insignificant amount, comprising approximately 65 % of the amount of
NH4+ included in the model for Baltimore, but makes sense given
the expected acidic nature of eastern US, sulfate-rich aerosols
(Weber et al., 2016). For Beijing, a persistent anion deficiency
was addressed by addition of OH- to the system to ensure
electroneutrality. The amount of OH- added to the system for the
Beijing case was 1 order of magnitude lower than the cation species but
on the same order of magnitude and 4–7 times lower than any other anion
except Cl-.
E-AIM offers support for certain organic acid species. For the Baltimore and
Beijing simulations, the organic acid species were added directly to the
E-AIM model inputs. In the case of organic acid model runs, factorial
combinations of the organic acid species at 0.0, 0.01, 0.02, 0.04, 0.08,
0.16, 0.32, 0.5, 1, 2, and 4 µg m-3 (Baltimore) and 0.0, 1, 2,
5, 10, 15, 20, 25, 30, 35, and 40 µg m-3 (Beijing) and were input into the model after concentrations were converted (mol m-3). Formation of organic solids was also
disabled as part of the metastable equilibrium condition. For the nonacid
organics, the addition of the selected species to the E-AIM equilibrium
calculation was not possible, and the model was run with the inorganic
constituents only. E-AIM provides output of the aqueous species mole
fractions and mole-fraction-based activity coefficients; this mole-fraction-based aerosol pH was converted to a molality-based aerosol pH
utilizing known thermodynamic relations (Robinson
and Stokes, 1965; Jia et al., 2018).
ISORROPIA-II model input configuration and equilibrium model evaluations
ISORROPIA-II provides thermodynamic equilibrium modeling for the
H+-NH4+-Na+-SO42--NO3--Cl--Ca2+-Mg2+-K+-H2O system across a wide range of temperature and RH values without limitation
based on the input composition (Fountoukis and Nenes, 2007). Data for
Beijing were already formatted for use in ISORROPIA-II as described above
(Guo et al., 2018b). The formation of solids in the
model was disabled (leading to potential supersaturated aerosols, metastable
mode operation), based on the justifications in previous studies (Guo et
al., 2015, 2018b) and to maintain consistency with the E-AIM
model conditions. An initial model run was performed to verify that
identical model outputs were obtained using the inputs of Guo et al. (2018b).
For the purposes of this investigation, the RH value was changed from the
Beijing average ambient value of 56 % to 70 %, 80 %, or 90 %,
consistent with the model input for the Baltimore data for the same reasons
discussed above. The Beijing average ambient temperature of 274.05 K was
used in the Beijing inorganic model calculations with the three RH values.
Flow diagram for adjusting total moles in the organic–inorganic
mixed system inputs for the AIOMFAC model. Total moles were adjusted
(representing increased Wo) by the first 11 and last 11 model points
until the AIOMFAC-output RH was within 5 % of the fixed model input value
(70 %, 80 %, or 90 %). These 22 points were then used to fit a
polynomial function to correct the total system moles for the remaining 1309 data points.
AIOMFAC model input configuration and equilibrium model evaluations
E-AIM was utilized to determine the equilibrium composition of the inorganic
aerosol, including the NH3 phase partitioning and the aerosol liquid water content. Outputs from E-AIM were then used as inputs into AIOMFAC to
characterize the organic effects on aerosol H+ activity and γH+ (Fig. 1). The E-AIM outputs (AIOMFAC inputs) were also checked
for consistency with ISORROPIA to ensure that the applied model assumptions
(H+ and OH- as balancing species to achieve electroneutrality)
provided reasonable results.
Whole species inorganic inputs used in AIOMFAC modeling at 70 %,
80 %, or 90 % RH. Inorganic equilibrium outputs were used to assign
anions to cations to form whole species, represented here. Entries marked
with a dash represent components not used for the given location as a result
of the anion or cation not being present in the composition data (species not
measured).
The particle-phase outputs from the E-AIM model runs were used as inputs to
AIOMFAC; however, this required significant adjustments to the format to fit
the AIOMFAC model. AIOMFAC requires inorganic species inputs to be entered
as ionic pairs (whole molecular species entered as a cation and anion pair) in
order to guarantee electroneutrality. Therefore, the ionic species outputs
of E-AIM were converted to molecular species inputs by assigning pairs and
then performing a stoichiometric balance until all ions were accounted for
(i.e., E-AIM H+ and SO42- being combined in stoichiometric fashion
as H2SO4 with corresponding reductions in the pool of E-AIM
H+ and SO42-). In the Baltimore case for the pure inorganic
input (all organic species modeled at 0.0 µg m-3 concentration),
E-AIM Model IV provided particle-phase output for the following ions:
H+, NH4+, Na+, HSO4-, SO42-,
NO3-, and OH-. In order to format these concentrations for
AIOMFAC-specific inputs (that is, to compute the necessary mole fraction
format of molecular species in the aerosol), the ions were assigned in the
following ways. First, all SO42- was associated with H+ for
the H2SO4 pair in AIOMFAC. All NO3- was associated with
Na+ for the NaNO3 pair. The remaining Na+ was associated with
SO42-, then NH4+ with HSO4-, and the remaining NH4+ with the remaining SO4-. This allocation process
proceeded similarly for the Beijing data. The selected species and order of
allocation of the ionic species appears to be dependent solely on the user
and a priori knowledge of which molecular species are likely to exist in the aerosol
particle as the dissociated ionic species. The selection of species is
unlikely to affect model outcomes, as this is simply a way to account for
the ionic species present in the AIOMFAC model inputs, which require matched
cation–anion pairs and are expected to be fully dissociated in the aqueous
phase during model evaluation. The end result is a mixture of inorganic
molecular species containing the full concentration values generated by
E-AIM assumed to be dissociated within the aerosol where each functional
group can contribute to species activity based on the AIOMFAC model
paradigm; assignment of molecular species pairings is performed only on the
basis of formatting specifically for the AIOMFAC model. The inorganic inputs
used in the AIOMFAC models for both Baltimore and Beijing simulations are
given in Table 2.
An additional key step in formatting the E-AIM output for input to AIOMFAC
is in the model treatment of the water associated with organic constituents,
Wo. E-AIM provided output of Wo as a part of the total aerosol
liquid water (ALW; Wi+Wo) for the organic acid simulations
but provided no estimate of Wo for the nonacid simulations. RH is not
an input to the AIOMFAC model runs. Rather, AIOMFAC requires the input of
all species (inorganic and organic) in mole fractions and assumes the
difference between the total inputs and unity is contributed by water, with
water activity equal to the ambient relative humidity.
Therefore, accounting for the water contributed by the organic species was
an additional step in formatting the E-AIM outputs for AIOMFAC input as
described below.
For two of the four cases (Baltimore and Beijing inorganics plus nonacid
organics), Wo was added to the system by the following process, a flow
diagram of which is shown in Fig. 1. For the first 11 points of the
factorial design (representing the addition of only the first organic
constituent at each concentration level) and the final 11 points of the
factorial design (representing the 11 highest organic addition points,
including the addition of all three organic species at their maximum
selected concentration), total system moles were varied manually by
increasing the inorganic model-predicted moles of aerosol water. AIOMFAC
inputs (as mole fractions) were calculated using this adjusted total mole
value. The 22 manually adjusted points were modeled in AIOMFAC. If the
option for liquid water is selected (as it was in all of our simulations),
AIOMFAC assumes that water makes up the difference between the mole or mass
fraction of all inputs summed together and unity. To achieve consistency
with the inorganic model results, the total moles of the system were
manually adjusted until the RH output generated by the AIOMFAC model was
within ∼5 % of the RH value fixed for the inorganic
systems. Once this close fit was achieved for the 22 selected points, they
were used to generate polynomial fits of the total moles added to the system
as Wo versus total organic mass (regardless of species). These
polynomial fits were then applied to all model points to adjust the total
system moles through the addition of liquid water associated with organic
mass, resulting in AIOMFAC-predicted RH values within 5 % of the E-AIM RH
values of 70 %, 80 %, or 90 %. This method of accounting for Wo
is a mathematical construct and does not reflect the use of a
species-dependent organic hygroscopicity parameter, which would have been
prohibitive to apply for each point across all cases and RH levels.
Additionally, following the introduction of the adjusted Wo to the
system, the gas phase was not allowed to re-equilibrate to the new water
content contributed by the organic species. This provides a conservative
(high) constraint on the effect of Wo (Guo et al., 2015).
Evaluation of LLPS and accommodations for LLPS scenarios
The O:C ratio is a key factor that determines whether LLPS occurs in
organic-containing particles (Song et al., 2018a; Freedman, 2017). We
followed the parameterization found experimentally by Bertram et al. (2011)
to evaluate the presence of LLPS in our simulations. This method uses the
overall mixture O:C ratio to determine the separation RH of the mixture. If
the modeled (in this case, specified/enforced) system RH is lower than the
parameterized separation RH (RHLLPS), LLPS is likely to occur. This was
performed for each of the nonacid mixtures for both Baltimore and Beijing
data to verify the claim that LLPS was not anticipated to occur. For cases
where the parameterized RHLLPS was higher than the predicted system RH,
LLPS was anticipated to occur, and the point was flagged and excluded from
further analysis. Out of 1331 simulations, Baltimore had 55 % (n=732),
70 % (n=932), and 75 % (n=998) simulations that met the non-LLPS
criteria at 70 %, 80 %, and 90 % RH respectively. Beijing had 85 %
(n=1131), 89 % (n=1185), and 93 % (n=1238) of simulations meet
the non-LLPS conditions at 70 %, 80 %, and 90 % RH respectively.
Experimental work by You et al. (2013) indicates that glutaric
acid, malonic acid, oxalic acid, or their mixtures do not undergo LLPS at
any of the RHs investigated.
Results and discussionEffects of WSOC on aerosol <inline-formula><mml:math id="M195" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">γ</mml:mi><mml:mrow class="chem"><mml:msup><mml:mi mathvariant="normal">H</mml:mi><mml:mo>+</mml:mo></mml:msup></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>
AIOMFAC-predicted aerosol pH and γH+ versus the organic dry
mass fraction (total mass of organics / sum of inorganics + organics, excluding ALW), along
with aerosol liquid water used in the model evaluations for the nonacid
species runs in Baltimore and Beijing at all RH levels, are shown in Figs. 2 and 3. For the case of nonacid WSOC compounds at 80 % RH (Figs. 2d and 3d), ALW increases from 4.7×10-9 to 9.7×10-9 L m-3 and from 9.6×10-8 to 1.8×10-7 L m-3 for Baltimore and
Beijing, respectively, as the organic mass fraction increases. Similar
trends follow for the 70 % and 90 % RH scenarios in both cities. This
behavior makes sense, because the inorganic species concentrations and RH
were fixed, so adding increasing levels of water-soluble organics increases
the ALW. Increasing the organic dry mass fraction increases the value of
γH+, from initial values of 0.10 and 0.16 (80 % RH) for
Baltimore and Beijing under inorganic-only conditions to 2.4 for Baltimore
(Fig. 2d) and 1.6 for Beijing (Fig. 3d). The higher absolute ALW levels in
the Beijing simulations are due to the significantly higher inorganic and
organic aerosol loadings.
AIOMFAC-modeled aerosol pH (molality basis) versus organic dry
mass fraction with the factorial addition of nonacid organic species and
AIOMFAC-modeled γH+ (molality basis) and aerosol liquid water
(ALW, polynomial fit to AIOMFAC output) versus organic dry mass fraction
(right panel) for Baltimore at (a, b) 70 % RH, (c, d) 80 % RH, and
(e, f) 90 % RH.
The results follow for the additional RH values studied. For the case of
nonacid organics at 70 % RH (Figs. 2 and 3), increasing the organic dry
mass fraction increases the value of γH+, from initial values
of 0.11 and 0.18 for Baltimore and Beijing under inorganic-only conditions
to 1.3 for Baltimore (at an organic dry mass fraction of 0.65) and 2.5 for
Beijing (organic dry mass fraction of 0.67). ALW follows a similar trend at
70 % RH as it does at 80 % RH, but with lower absolute levels. For
the case of nonacid organics at 90 % RH, increasing the organic dry mass
fraction increases γH+, from initial values of 0.12 and 0.20 for
Baltimore and Beijing under inorganic-only conditions to 1.2 for Baltimore
(at an organic dry mass fraction of 0.79) and 0.78 for Beijing (organic dry
mass fraction of 0.64). For these simulations, the ALW increases from 9.1×10-9 to 2.4×10-8 L m-3
and from 2.1×10-7 to 3.6×10-7 L m-3 for Baltimore and Beijing, respectively. In each case, the
data plotted in Figs. 2 and 3 are those that are determined not to have
LLPS according to the parameterization of Bertram et al. (2011).
AIOMFAC-modeled aerosol pH (molality basis) versus organic dry
mass fraction with the factorial addition of nonacid organic species and
AIOMFAC-modeled γH+ (molality basis) and aerosol liquid water
(ALW, polynomial fit to AIOMFAC output) versus organic dry mass fraction
(right panel) for Beijing at (a, b) 70 % RH, (c, d) 80 % RH, and (e, f) 90 % RH.
The plots of ALW display distinct behaviors attributable to the way in which
the water content was derived for the model systems. For the organic acid
simulations, the ALW was taken directly from the E-AIM Model IV output of
aqueous-phase water (mol m-3) run with inorganic and organic acid
inputs. For all nonacid organic cases, total ALW (Wi+Wo) was
determined according to the manual AIOMFAC output fitting/polynomial fit
correlation described in the methods section (Fig. 1). This results in
system water behavior described by polynomial fits of additional water
versus organic dry mass fraction.
Effects of WSOC on aerosol pH
The model-predicted effects of WSOC on aerosol pH are shown in Figs. 2–5. As the dry organic mass fraction increases, ALW increases as well, since
the RH and inorganics are held constant. This suggests a diluting effect,
which would increase pH, in agreement with Guo et al. (2015). On the other
hand, γH+ also increases with increasing dry organic mass
fraction, indicating that the addition of WSOC compounds increases the
acidity (decreases pH).
E-AIM-modeled (a) aerosol pH (molality basis) and (b) ALW at
70 % (blue), 80 % (red), and 90 % RH (green) levels for the Baltimore
simulations with organic acids.
For the case of nonacid WSOC additions (Figs. 2 and 3), increasing the
organic mass fraction decreases the predicted aerosol pH from the initial
inorganic-only values from 1.64 to a max of 1.94 (Baltimore) and from 4.29
to a max of 4.38 (Beijing) at 80 % RH. For the 70 % RH simulations, the model predicted pH changes from 1.49 to a max of 1.88 for Baltimore and
from 4.10 to a max of 4.33 for Beijing. For the 90 % RH case, the model
predicted pH changes from 1.85 to a max of 2.09 for Baltimore and from 4.52
to a max of 4.56 for Beijing. The transition in the pH plots are smooth,
where the contour lines reflect individual levels of the factorial design
and highlight the overall trend: as nonacidic WSOC is added,
AIOMFAC-predicted aerosol pH increases for both the Baltimore and Beijing
conditions. Since the WSOC leads to ALW uptake (diluting acidity), the
increase in pH comes about due to the increase ALW having a stronger effect
than the increase in γH+.
E-AIM-modeled (a) aerosol pH (molality basis) and (b) ALW at
70 % (blue), 80 % (red), and 90 % RH (green) levels for the Beijing
simulations with organic acids.
For the case of organic acids, increasing the organic mass fraction results
in only slight changes in the predicted aerosol pH for Baltimore (Fig. 4)
but more pronounced changes for Beijing (Fig. 5). At 80 % RH, the
predicted pH ranges from an initial (inorganic-only) value of 1.49
(Baltimore) and 4.2 (Beijing) to 1.34 and 2.6, respectively (total range =0.4 and 1.68 pH units respectively). Similarly, there is a change from 1.33
to 1.22 (range =0.34 pH units) for Baltimore and a change from 4.06 to
2.52 (range =1.58 pH units) for the Beijing simulations at 70 % RH.
Finally, there is a change from 1.75 to 1.54 (range =0.47) for Baltimore
and a change from 4.44 to 2.68 (range =1.78) for Beijing at 90 % RH.
The ranges represent the total range spread from the highest to lowest
model-predicted pH. For Baltimore, organic acids are predicted to have only
a slight effect on aerosol pH. Under the highly acidic conditions typical of
the eastern US (inorganic-only pH ∼1), pH changes are
always < 0.5 pH units, even when the dry organic aerosol mass
fraction exceeds 60 % (corresponding to total aerosol organic acid
concentrations up to 12 µg m-3). This is likely due to the pH
being sufficiently acidic that the organic acid dissociation is largely
inhibited. The undissociated organic acids still contribute ALW and affect
γH+, but the combined effects produce very minor modifications
to pH. For Beijing, aerosol pH changes are predicted to be me more
substantial with the addition of organic acids, due to the initially higher
aerosol pH. As organic acids are added, they can dissociate and contribute
free H+. However, pH changes in excess of 1 pH unit only occur at dry
organic mass fractions > 0.5. Given the high inorganic aerosol
concentrations in Beijing, such pH changes in excess of 1 pH unit correspond
to unrealistically high aerosol organic acid mass concentrations
(> 35 µg m-3). The relatively minor effect of organic
acids on aerosol pH in Beijing is partly due to the high concentrations of
ammonia (the sum of NH3 and NH4+ in Beijing =32.8µg m-3), which also
contribute to the much higher inorganics-only pH compared to the eastern
US conditions.
ΔpH vs. discrete organic acid levels for the organic acid
pH calculations carried out in E-AIM for Baltimore and Beijing. ΔpH
represents the difference between the initial inorganic-only pH prediction
and all other model evaluations where organic acids were present.
The effect of organic acids on pH is closely tied to acid strength (i.e.,
pKa value). Figure 6 shows that organic acids affect pH in the order of
their pKa values, with oxalic acid (pKa1=1.23) > malonic acid (pKa1=2.83) > glutaric acid (pKa1=4.31) (Lide, 1994). The simulations with a single organic acid demonstrate
this effect most clearly: addition of 40 µg m-3 oxalic, malonic,
and glutaric acid produce pH changes of -1.3, -0.5, and -0.2 pH units,
respectively. The pH changes are all negative, indicating that the organic
acids have increased particle acidity (H+). Note that although the
molar amounts added are not equivalent, the observed pH changes represent
log-scale changes to the H+ activity and the effect does proceed in the
order of acid strength.
The magnitudes of these observed pH changes, with the exception of the
Beijing organic acids case at high organic mass fractions (> 35 µg m-3 acids concentration), are not expected to significantly
alter particle conditions or lead to substantial changes in particle
chemistry. For example, ∼0.5 pH unit changes should not
significantly alter IEPOX uptake (Xu et al., 2015) or metal
dissolution (Fang et al., 2017), two processes affected by particle acidity.
An exception would be conditions where the pH is close to the point where a
given species is almost equally partitioned between the gas and particle
phases (i.e., on the center/vertical portion of the titration-style sigmoid
curves). This effect is demonstrated in the work of Guo et al. (2018b) and
Vasilakos et al. (2018): when the pH lays on or near the inflection point of
the sigmoid curve, a change of 0.5 pH units can have a significant effect on
species partitioning; however, when the pH is in the flatter regions of the
curve above or below the rapid transitional region, a change of 0.5 pH units
will have a negligible effect on partitioning and thus particle chemistry.
Taken together, these results indicate that, despite organic mass fractions
greater than 60 % (dry particle mass basis), the combined effects of WSOC
species on model-predicted aerosol pH is only about 0.5 pH units, maximum,
with most pH changes < 0.2 pH units. This result is observed for
nonacidic WSOC species and realistic concentrations of organic acids, as well as for simulations with a single organic compound added or for mixtures (Table 2). This suggests that the overall effect of WSOC on aerosol pH is quite
minimal in conditions where LLPS does not occur. This finding holds only for
systems in which there is no LLPS and the solvent is H2O. For systems
in which LLPS does occur, a condition expected in systems with O:C ratio of
the organic material ≤0.5, or RH < 60 % with organic–sulfate
mass ratio < 1 (Bertram et al., 2011; You et al., 2013), the
situation becomes more complicated. As LLPS scenarios still require
equilibrium between both predominantly aqueous and predominately organic
phases, there are both water and inorganic ions (including H+) in the
organic phase and organics in the inorganic-rich aqueous phase (Zuend
and Seinfeld, 2012; Pye et al., 2018). Thus the IUPAC definition of pH could
be applied to either phase so long as H+ activity could be defined,
necessitating an understanding of if and when LLPS occurs, as well as the phase for
which pH is being reported.
This work stands apart from but connects to related works.
Pye et al. (2018) specifically examined the
effects of LLPS, but the present study examines a different particle regime
altogether (single aqueous phase with water, inorganics, and organics);
instances where LLPSs were predicted to occur were excluded from the analysis
for this reason.
Our findings are supported by the work of Song et
al. (2018b), who utilized E-AIM Model IV and ISORROPIA to model the same
Beijing winter haze conditions and found that addition of oxalic acid (set
at 20 % of the sulfate concentration) to their model in E-AIM produced
reductions in pH of only 0.07 pH units. Our results are also consistent with
those of Vasilakos et al. (2018), who observed a similarly
minor effect of oxalate addition on aerosol pH in the eastern US,
and (Nah et al., 2018), where oxalic acid/oxalate
gas–particle partitioning predicted without considering organic species in
the thermodynamic analysis was in reasonable agreement with measurements.
Our results indicate that additions of weaker organic acids, even at higher
concentrations, would have even less of an effect on pH.
A limitation of this study is that the model simulations were only run at
three RH levels (70 %, 80 %, and 90 %), with metastable conditions
enforced at all times. However, aerosol particles progress through a wider
RH range in the atmosphere, with concomitant effects on aerosol liquid water
and phase transitions. Future work would need to expand on the RH range in
order to elucidate the behavior as the system transitions from the LLPS
condition to the fully mixed aqueous condition, as well as the contribution of
changing ALW. Additionally, the use of E-AIM Model IV imposes composition
limitations on the inputs (i.e., no support for Ca2+, Mg2+, or
K+; limited support for Na+ in the presence of NH4+ and
Cl-), necessitating the use of equivalent cations to maintain
electroneutrality in the model inputs. Combined with the use of metastable
calculations, there exists a potential source of error in the solution
activity if these species are considered and allowed to precipitate out in
the thermodynamic model calculations (e.g., CaSO4). As AIOMFAC relies
on specific, uniquely defined functional group interactions in the
composition of activity coefficients, the exchange of an unsupported cation
in E-AIM for a charge-equivalent cation may have effects on the output
unknown to us carried through to the calculation of the species activity
coefficients in AIOMFAC; this is a limitation of thermodynamic models that
has been previously discussed (Jacobson, 1999; Kim and Seinfeld, 1995).
Another limitation of this study is consideration of only six WSOC species,
despite hundreds or thousands being present in atmospheric particles. This
is a limitation we acknowledge but is based on the significant number of
model runs given the factorial design paradigm, as well as the decision to utilize
only compounds predefined in the thermodynamic models (particularly the
AIOMFAC model, which allows users to create organic molecules by combining
subgroups). Because the compounds selected here have relatively low
molecular weight (MW), it is possible that higher MW compounds, such as
humic-like substances (HULIS), may impart a different effect. However, given
the consistent results found here for both Baltimore and Beijing conditions
across the 70 %–90 % RH range, and at organic dry mass fractions that range
from 0 % to 60 % utilizing WSOC containing four moieties, we feel our results
do represent conditions in atmospheric particles. Future studies would be
necessary to expand the selection of WSOC compounds and thus broaden the
results reported here. Because we have forced the metastable mode on our use
of the models, the system mixing state becomes another potential source of error. Here we have considered only internally mixed
aerosol particles without LLPS, a case that may not exist given the
concentration of organic species utilized in the model study; formation of
solid precipitates may occur, which has the potential to drastically alter
the aqueous phase activity values. The most significant restriction of this
study is the lack of observational data for comparison. Direct measurements
of particle pH have so far been restricted to simple laboratory particles of
specific supermicron sizes and compositions (Rindelaub et al.,
2016).
Conclusions and implications
In this work, the effects of WSOC on model-predicted aerosol pH were
evaluated. Different inorganic datasets from Baltimore and Beijing winter
haze conditions representing distinct inorganic composition regimes were
first modeled in aerosol thermodynamic equilibrium models (E-AIM or
ISORROPIA) and then combined with six different organic species in AIOMFAC to
determine the effects on aqueous-phase γH+ and aH+. We
find that the effects of nonacid WSOC species to each of the regions have
only a modest effect on aerosol pH (< 0.5 pH units, with most
< 0.2 pH units). These small effects on pH were predicted even up to
organic dry mass in excess of 60 %. Organic acids are predicted to have a
similarly small effect on pH in the eastern US. In Beijing, organic acids
can have larger effects on pH (in excess of 2 pH units) but require
stronger organic acids (pKa values lower than the inorganic-only pH)
present at high concentrations, on the same order of molar concentration as
the dissociating inorganic species. The magnitude of these changes to
aerosol pH is consistent with the results predicted by previous studies
that considered only inorganic aerosol components combined with a priori knowledge
of organic mass, organic water contribution, and organic species
hygroscopicity (Guo et al., 2015; Bougiatioti et al., 2016).
The results of this study have important implications for the aerosol
modeling community as well as for experimental studies that utilize phase
partitioning data to constrain aerosol pH. Previous studies have postulated
on the effect of organic species while ignoring their inclusion, or included
them in order to elucidate the effects of LLPS, but this study demonstrates
that, in the case of single-phase systems, including these species may only
contribute unnecessary complexity to the model runs. As their effects are
predominantly < 0.5 pH units, it is not expected that the inclusion
of organics will cause the pH of the system to reach any significant
transitions unless the organic components have already driven the system to
a sensitive portion of the species partitioning sigmoid curves, or aerosols
in which there is significant phase separation. This work demonstrates that
inclusion of large quantities of organic components does not appear to have
a significant effect on model-predicted aerosol pH, consistent with the
findings of Guo et al. (2018a) and Vasilakos et al. (2018). Based on the species and concentrations of the organics studied
here, future aerosol modeling studies carried out under conditions where
LLPS is not occurring may be justified in the use of inorganic-only aerosol
thermodynamic equilibrium models to predict aerosol pH without the direct
inclusion of organic species.
Data availability
All model inputs and outputs are available at
https://knb.ecoinformatics.org/
(urn:uuid:ae58e5f1-fb13-4e9c-800b-eeb8bcb14d84) (Hennigan and Battaglia, 2019).
Author contributions
CH and MB devised the study. MB performed all modeling analyses. CH, MB, RW, and AN collaborated on interpreting the results. MB prepared the manuscript, with significant contributions from CH, RW, and AN.
Competing interests
The authors declare that they have no conflict of interest.
Acknowledgements
Michael A. Battaglia Jr. and Christopher J. Hennigan were supported by NSF grant no. CHE-1454763. Athanasios Nenes
acknowledges support from the project PyroTRACH (ERC-2016-COG) funded from
H2020-EU.1.1. – Excellent Science – European Research Council (ERC), project
ID 726165. Rodney J. Weber recognizes support from the US EPA under STAR grant 835882. It is noted that the results of this research have not been formally reviewed by the EPA and the views expressed in this document are solely those of the authors and do not necessarily reflect those of the EPA. The EPA does not endorse any products or commercial services mentioned in this publication.
Financial support
This research has been supported by the National Science Foundation, Directorate for Mathematical and Physical Sciences (grant no. CHE-1454763); the European Research Council (grant no. PyroTRACH (726165)); and the US EPA (grant no. 835882).
Review statement
This paper was edited by Alexander Laskin and reviewed by two anonymous referees.
ReferencesAhrens, L., Harner, T., Shoeib, M., Lane, D. A., and Murphy, J. G.: Improved characterization of gas–particle partitioning
for per- and polyfluoroalkyl substances in the atmosphere using annular
diffusion denuder samplers, Environ. Sci. Technol., 46,
7199–7206, 10.1021/es300898s, 2012.Battaglia, M. A., Douglas, S., and Hennigan, C. J.: Effect of the Urban Heat
Island on Aerosol pH, Environ. Sci. Technol., 51,
13095–13103, 10.1021/acs.est.7b02786, 2017.Bertram, A. K., Martin, S. T., Hanna, S. J., Smith, M. L., Bodsworth, A., Chen, Q., Kuwata, M., Liu, A., You, Y., and Zorn, S. R.: Predicting the relative humidities of liquid-liquid phase separation, efflorescence, and deliquescence of mixed particles of ammonium sulfate, organic material, and water using the organic-to-sulfate mass ratio of the particle and the oxygen-to-carbon elemental ratio of the organic component, Atmos. Chem. Phys., 11, 10995–11006, 10.5194/acp-11-10995-2011, 2011.Bikkina, S., Kawamura, K., and Miyazaki, Y.: Latitudinal distributions of
atmospheric dicarboxylic acids, oxocarboxylic acids, and -dicarbonyls over
the western North Pacific: Sources and formation pathways,
J. Geophys. Res.-Atmos., 120, 5010–5035, 10.1002/2014jd022235,
2015.Bougiatioti, A., Nikolaou, P., Stavroulas, I., Kouvarakis, G., Weber, R., Nenes, A., Kanakidou, M., and Mihalopoulos, N.: Particle water and pH in the eastern Mediterranean: source variability and implications for nutrient availability, Atmos. Chem. Phys., 16, 4579–4591, 10.5194/acp-16-4579-2016, 2016.Carlton, A. G., Wiedinmyer, C., and Kroll, J. H.: A review of Secondary Organic Aerosol (SOA) formation from isoprene, Atmos. Chem. Phys., 9, 4987–5005, 10.5194/acp-9-4987-2009, 2009.Chameides, W. L.: The Photochemistry of a Remote Marine Stratiform Cloud,
J. Geophys. Res.-Atmos., 89, 4739–4755,
10.1029/JD089iD03p04739, 1984.Dallemagne, M. A., Huang, X. Y., and Eddingsaas, N. C.: Variation in pH of
Model Secondary Organic Aerosol during Liquid-Liquid Phase Separation,
J. Phys. Chem. A, 120, 2868–2876, 10.1021/acs.jpca.6b00275,
2016.Fang, T., Guo, H. Y., Zeng, L. H., Verma, V., Nenes, A., and Weber, R. J.: Highly Acidic Ambient Particles, Soluble Metals, and
Oxidative Potential: A Link between Sulfate and Aerosol Toxicity, Environ.
Sci. Technol., 51, 2611–2620, 10.1021/acs.est.6b06151, 2017.Fountoukis, C. and Nenes, A.: ISORROPIA II: a computationally efficient thermodynamic equilibrium model for K+–Ca2+–Mg2+–NH4+–Na+–SO42−–NO3−–Cl−–H2O aerosols, Atmos. Chem. Phys., 7, 4639–4659, 10.5194/acp-7-4639-2007, 2007.Freedman, M. A.: Phase separation in organic aerosol, Chem. Soc. Rev., 46,
7694–7705, 10.1039/c6cs00783j, 2017.Friese, E. and Ebel, A.: Temperature Dependent Thermodynamic Model of the
System H+-NH4+-Na+-SO42-NO3--Cl--H2O, J. Phys. Chem. A, 114,
11595–11631, 10.1021/jp101041j, 2010.Guo, H., Xu, L., Bougiatioti, A., Cerully, K. M., Capps, S. L., Hite Jr., J. R., Carlton, A. G., Lee, S.-H., Bergin, M. H., Ng, N. L., Nenes, A., and Weber, R. J.: Fine-particle water and pH in the southeastern United States, Atmos. Chem. Phys., 15, 5211–5228, 10.5194/acp-15-5211-2015, 2015.Guo, H., Sullivan, A. P., Campuzano-Jost, P., Schroder, J. C., Lopez-Hilfiker, F. D., Dibb, J. E., Jimenez, J. L., Thornton, J. A., Brown, S. S., Nenes, A., and Weber, R. J.: Fine particle pH and the partitioning of nitric acid during
winter in the northeastern United States,
J. Geophys. Res.-Atmos., 121, 10355–10376, 10.1002/2016JD025311, 2016.Guo, H., Liu, J., Froyd, K. D., Roberts, J. M., Veres, P. R., Hayes, P. L., Jimenez, J. L., Nenes, A., and Weber, R. J.: Fine particle pH and gas–particle phase partitioning of inorganic species in Pasadena, California, during the 2010 CalNex campaign, Atmos. Chem. Phys., 17, 5703–5719, 10.5194/acp-17-5703-2017, 2017.Guo, H., Nenes, A., and Weber, R. J.: The underappreciated role of nonvolatile cations in aerosol ammonium-sulfate molar ratios, Atmos. Chem. Phys., 18, 17307–17323, 10.5194/acp-18-17307-2018, 2018a.Guo, H., Otjes, R., Schlag, P., Kiendler-Scharr, A., Nenes, A., and Weber, R. J.: Effectiveness of ammonia reduction on control of fine particle nitrate, Atmos. Chem. Phys., 18, 12241–12256, 10.5194/acp-18-12241-2018, 2018b.Gwynn, R. C., Burnett, R. T., and Thurston, G. D.: A time-series analysis of
acidic particulate matter and daily mortality and morbidity in the Buffalo,
New York, region, Environ. Health Persp., 108, 125–133, 10.2307/3454510,
2000.Hallquist, M., Wenger, J. C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M., Dommen, J., Donahue, N. M., George, C., Goldstein, A. H., Hamilton, J. F., Herrmann, H., Hoffmann, T., Iinuma, Y., Jang, M., Jenkin, M. E., Jimenez, J. L., Kiendler-Scharr, A., Maenhaut, W., McFiggans, G., Mentel, Th. F., Monod, A., Prévôt, A. S. H., Seinfeld, J. H., Surratt, J. D., Szmigielski, R., and Wildt, J.: The formation, properties and impact of secondary organic aerosol: current and emerging issues, Atmos. Chem. Phys., 9, 5155–5236, 10.5194/acp-9-5155-2009, 2009.
Hennigan, C. J. and Battaglia Jr., M. A.: Thermodynamic modeling of WSOC effects on aerosol pH, Knowledge Network for Biocomplexity, urn:uuid:ae58e5f1-fb13-4e9c-800b-eeb8bcb14d84, 2019.Hennigan, C. J., Izumi, J., Sullivan, A. P., Weber, R. J., and Nenes, A.: A critical evaluation of proxy methods used to estimate the acidity of atmospheric particles, Atmos. Chem. Phys., 15, 2775–2790, 10.5194/acp-15-2775-2015, 2015.Huang, R.-J., Zhang, Y., Bozzetti, C., Ho, K.-F., Cao, J.-J., Han, Y., Daellenbach, K. R., Slowik, J. G., Platt, S. M., Canonaco, F., Zotter, P., Wolf, R., Pieber, S. M., Bruns, E. A., Crippa, M., Ciarelli, G., Piazzalunga, A., Schwikowski, M., Abbaszade, G., Schnelle-Kreis, J., Zimmermann, R., An, Z., Szidat, S., Baltensperger, U., El Haddad, I., and Prevot, A. S. H.: High secondary aerosol contribution to particulate
pollution during haze events in China, Nature, 514, 218–222,
10.1038/nature13774, 2014.Jacobson, M. Z.: Studying the effects of calcium and magnesium on
size-distributed nitrate and ammonium with EQUISOLV II, Atmos. Environ., 33,
3635–3649, 10.1016/S1352-2310(99)00105-3, 1999.Jia, S., Wang, X., Zhang, Q., Sarkar, S., Wu, L., Huang, M., Zhang, J., and Yang, L.: Technical note: Comparison and interconversion of pH based on different standard states for aerosol acidity characterization, Atmos. Chem. Phys., 18, 11125–11133, 10.5194/acp-18-11125-2018, 2018.Johnson, A. H., Moyer, A., Bedison, J. E., Richter, S. L., and Willig, S. A.: Seven Decades of Calcium Depletion in Organic
Horizons of Adirondack Forest Soils, Soil Sci. Soc. Am.
J., 72, 1824–1830, 10.2136/sssaj2006.0407, 2008.Kanakidou, M., Myriokefalitakis, S., Daskalakis, N., Fanourgakis, G., Nenes, A., Baker, A. R., Tsigaridis, K., and Mihalopoulos, N.: Past, Present, and Future Atmospheric Nitrogen
Deposition, J. Atmos. Sci., 73, 2039–2047,
10.1175/jas-d-15-0278.1, 2016.Kawamura, K. and Bikkina, S.: A review of dicarboxylic acids and related
compounds in atmospheric aerosols: Molecular distributions, sources and
transformation, Atmos. Res., 170, 140–160, 10.1016/j.atmosres.2015.11.018,
2016.Keene, W. C., Sander, R., Pszenny, A. A. P., Vogt, R., Crutzen, P. J., and Galloway, J. N.: Aerosol pH in the marine boundary layer: A review and
model evaluation, J. Aerosol Sci., 29, 339–356, 10.1016/S0021-8502(97)10011-8, 1998.Keene, W. C., Pszenny, A. A. P., Maben, J. R., Stevenson, E., and Wall, A.: Closure evaluation of size-resolved aerosol pH in the
New England coastal atmosphere during summer, J. Geophys. Res.-Atmos., 109, D23307, 10.1029/2004jd004801, 2004.
Keppel, G.: Design and Analysis: A Researcher's Handbook, 3rd ed.,
Prentice-Hall, Inc., Englewood Cliffs, NJ, US, 1991.Kim, Y. P. and Seinfeld, J. H.: Atmospheric Gas-Aerosol Equilibrium 3.
Thermodynamics of Crustal Elements Ca2+, K+, and Mg2+, Aerosol Sci.
Tech., 22, 93–110, 10.1080/02786829408959730, 1995.
Lide, D. R. (Ed.):
CRC Handbook of Chemistry and Physics, 75th Edition
CRC Press, Inc.
Ann Arbor, MI, 1994.Myriokefalitakis, S., Nenes, A., Baker, A. R., Mihalopoulos, N., and Kanakidou, M.: Bioavailable atmospheric phosphorous supply to the global ocean: a 3-D global modeling study, Biogeosciences, 13, 6519–6543, 10.5194/bg-13-6519-2016, 2016.Myriokefalitakis, S., Ito, A., Kanakidou, M., Nenes, A., Krol, M. C., Mahowald, N. M., Scanza, R. A., Hamilton, D. S., Johnson, M. S., Meskhidze, N., Kok, J. F., Guieu, C., Baker, A. R., Jickells, T. D., Sarin, M. M., Bikkina, S., Shelley, R., Bowie, A., Perron, M. M. G., and Duce, R. A.: Reviews and syntheses: the GESAMP atmospheric iron deposition model intercomparison study, Biogeosciences, 15, 6659–6684, 10.5194/bg-15-6659-2018, 2018.Nah, T., Guo, H., Sullivan, A. P., Chen, Y., Tanner, D. J., Nenes, A., Russell, A., Ng, N. L., Huey, L. G., and Weber, R. J.: Characterization of aerosol composition, aerosol acidity, and organic acid partitioning at an agriculturally intensive rural southeastern US site, Atmos. Chem. Phys., 18, 11471–11491, 10.5194/acp-18-11471-2018, 2018.Nenes, A., Krom, M. D., Mihalopoulos, N., Van Cappellen, P., Shi, Z., Bougiatioti, A., Zarmpas, P., and Herut, B.: Atmospheric acidification of mineral aerosols: a source of bioavailable phosphorus for the oceans, Atmos. Chem. Phys., 11, 6265–6272, 10.5194/acp-11-6265-2011, 2011.
Peters, A., Goldstein, I. F., Beyer, U., Franke, K., Heinrich, J., Dockery, D. W., Spengler, J. D., and Wichmann, H. E.: Acute health effects of exposure to high levels of air
pollution in eastern Europe, Am. J. Epidemiol., 144, 570–581, 1996.Pye, H. O. T., Zuend, A., Fry, J. L., Isaacman-VanWertz, G., Capps, S. L., Appel, K. W., Foroutan, H., Xu, L., Ng, N. L., and Goldstein, A. H.: Coupling of organic and inorganic aerosol systems and the effect on gas–particle partitioning in the southeastern US, Atmos. Chem. Phys., 18, 357–370, 10.5194/acp-18-357-2018, 2018.Rindelaub, J. D., Craig, R. L., Nandy, L., Bondy, A. L., Dutcher, C. S., Shepson, P. B., and Ault, A. P.: Direct Measurement of pH in Individual Particles
via Raman Microspectroscopy and Variation in Acidity with Relative Humidity,
J. Phys. Chem. A, 120, 911–917, 10.1021/acs.jpca.5b12699,
2016.
Robinson, R. A. and Stokes, R. H.: Electrolyte solutions, 2nd revised ed.,
Butterworths, London, 1965.Schindler, D. W.: Effects of acid rain on freshwater ecosystems, Science,
239, 149–157, 10.1126/science.239.4836.149, 1988.
Seinfeld, J. H. and Pandis, S. N.: Atmospheric chemistry and physics: from
air pollution to climate change, 3 ed., John Wiley & Sons, Hoboken, 2016.Song, M., Ham, S., Andrews, R. J., You, Y., and Bertram, A. K.: Liquid–liquid phase separation in organic particles containing one and two organic species: importance of the average O : C, Atmos. Chem. Phys., 18, 12075–12084, 10.5194/acp-18-12075-2018, 2018a.Song, S., Gao, M., Xu, W., Shao, J., Shi, G., Wang, S., Wang, Y., Sun, Y., and McElroy, M. B.: Fine-particle pH for Beijing winter haze as inferred from different thermodynamic equilibrium models, Atmos. Chem. Phys., 18, 7423–7438, 10.5194/acp-18-7423-2018, 2018b.Spengler, J. D., Koutrakis, P., Dockery, D. W., Raizenne, M., and Speizer, F. E.: Health effects of acid aerosols on North American
children: Air pollution exposures, Environ. Health Persp., 104, 492–499, 10.2307/3432989, 1996.Tan, T. Y., Hu, M., Li, M. R., Guo, Q. F., Wu, Y. S., Fang, X., Gu, F. T., Wang, Y., and Wu, Z. J.: New insight into PM2.5 pollution patterns in Beijing
based on one-year measurement of chemical compositions, Sci. Total
Environ., 621, 734–743, 10.1016/j.scitotenv.2017.11.208, 2018.Trebs, I., Metzger, S., Meixner, F. X., Helas, G. N., Hoffer, A., Rudich, Y., Falkovich, A. H., Moura, M. A. L., da Silva, R. S., Artaxo, P., Slanina, J., and Andreae, M. O.: The NH4+-NO3--Cl--SO42--H2O aerosol system and its gas
phase precursors at a pasture site in the Amazon Basin: how relevant are
mineral cations and soluble organic acids?, J. Geophys. Res.-Atmos., 110, D07303, 10.1029/2004JD005478, 2005.Vasilakos, P., Russell, A., Weber, R., and Nenes, A.: Understanding nitrate formation in a world with less sulfate, Atmos. Chem. Phys., 18, 12765–12775, 10.5194/acp-18-12765-2018, 2018.Wang, G., Zhang, R., Gomez, M. E., Yang, L., Levy Zamora, M., Hu, M., Lin, Y., Peng, J., Guo, S., Meng, J., Li, J., Cheng, C., Hu, T., Ren, Y., Wang, Y., Gao, J., Cao, J., An, Z., Zhou, W., Li, G., Wang, J., Tian, P., Marrero-Ortiz, W., Secrest, J., Du, Z., Zheng, J., Shang, D., Zeng, L., Shao, M., Wang, W., Huang, Y., Wang, Y., Zhu, Y., Li, Y., Hu, J., Pan, B., Cai, L., Cheng, Y., Ji, Y., Zhang, F., Rosenfeld, D., Liss, P. S., Duce, R. A., Kolb, C. E., and Molina, M. J.: Persistent sulfate formation from London Fog to Chinese
haze, P. Natl. Acad. Sci. USA,
113, 13630–13635, 10.1073/pnas.1616540113, 2016.Weber, R. J., Guo, H., Russell, A. G., and Nenes, A.: High aerosol acidity
despite declining atmospheric sulfate concentrations over the past 15 years,
Nat. Geosci., 9, 282–285, 10.1038/ngeo2665, 2016.Wexler, A. S. and Clegg, S. L.: Atmospheric aerosol models for systems
including the ions H+, NH4+, Na+, SO42-,
NO3-,Cl-, Br-, and H2O, J. Geophys. Res.-Atmos., 107,
4207, 10.1029/2001jd000451, 2002.Xu, L., Guo, H. Y., Boyd, C. M., Klein, M., Bougiatioti, A., Cerully, K. M., Hite, J. R., Isaacman-VanWertz, G., Kreisberg, N. M., Knote, C., Olson, K., Koss, A., Goldstein, A. H., Hering, S. V., de Gouw, J., Baumann, K., Lee, S. H., Nenes, A., Weber, R. J., and Ng, N. L.: Effects of anthropogenic emissions on aerosol formation from
isoprene and monoterpenes in the southeastern United States, P.
Natl. Acad. Sci. USA, 112, 37–42, 10.1073/pnas.1417609112, 2015.
Yatavelli, R. L. N., Mohr, C., Stark, H., Day, D. A., Thompson, S. L., Lopez-Hilfiker, F. D., Campuzano-Jost, P., Palm, B. B., Vogel, A. L., Hoffmann, T., Heikkinen, L., Aijala, M., Ng, N. L., Kimmel, J. R., Canagaratna, M. R., Ehn, M., Junninen, H., Cubison, M. J., Petaja, T., Kulmala, M., Jayne, J. T., Worsnop, D. R., and Jimenez, J. L.: Estimating the contribution of organic acids to
northern hemispheric continental organic aerosol, Geophys. Res.
Lett., 42, 6084–6090, 10.1002/2015gl064650, 2015.You, Y., Renbaum-Wolff, L., and Bertram, A. K.: Liquid–liquid phase separation in particles containing organics mixed with ammonium sulfate, ammonium bisulfate, ammonium nitrate or sodium chloride, Atmos. Chem. Phys., 13, 11723–11734, 10.5194/acp-13-11723-2013, 2013.Zhou, W., Wang, Q., Zhao, X., Xu, W., Chen, C., Du, W., Zhao, J., Canonaco, F., Prévôt, A. S. H., Fu, P., Wang, Z., Worsnop, D. R., and Sun, Y.: Characterization and source apportionment of organic aerosol at 260 m on a meteorological tower in Beijing, China, Atmos. Chem. Phys., 18, 3951–3968, 10.5194/acp-18-3951-2018, 2018.Zuend, A. and Seinfeld, J. H.: Modeling the gas-particle partitioning of secondary organic aerosol: the importance of liquid-liquid phase separation, Atmos. Chem. Phys., 12, 3857–3882, 10.5194/acp-12-3857-2012, 2012.Zuend, A., Marcolli, C., Luo, B. P., and Peter, T.: A thermodynamic model of mixed organic-inorganic aerosols to predict activity coefficients, Atmos. Chem. Phys., 8, 4559–4593, 10.5194/acp-8-4559-2008, 2008.Zuend, A., Marcolli, C., Booth, A. M., Lienhard, D. M., Soonsin, V., Krieger, U. K., Topping, D. O., McFiggans, G., Peter, T., and Seinfeld, J. H.: New and extended parameterization of the thermodynamic model AIOMFAC: calculation of activity coefficients for organic-inorganic mixtures containing carboxyl, hydroxyl, carbonyl, ether, ester, alkenyl, alkyl, and aromatic functional groups, Atmos. Chem. Phys., 11, 9155–9206, 10.5194/acp-11-9155-2011, 2011.