JOURNAL OF GEOPHYSICAL RESEARCH, VOL 107, NO. D19, 8111, doi:10.1029/2001JD000900, 2002
Aerosol formation during PARFORCE: Ternary nucleation of
H2SO4, NH3, and H2O
M. Kulmala,1 P. Korhonen,1,2 I. Napari,1 A. Karlsson,3 H. Berresheim,4
and C. D. O’Dowd1,5,6
Received 29 May 2001; revised 17 September 2001; accepted 17 September 2001; published 25 September 2002.
[1] A new version of a ternary nucleation (sulphuric acid-ammonia-water) model based
on classical nucleation theory, but with an improved ability to predict nucleation rates over
a larger temperature range (258–303 K) compared with previous work, is presented. The
modeled nucleation rates are given as a function of temperature and ambient acid and
ammonia concentrations. For the first time the predicted ternary nucleation rates are
compared to the observed particle production rates using measured ambient sulphuric acid
and ammonia concentrations as input data. The ambient gas concentrations were measured
simultaneously to aerosol formation rates during the 1999 New Particle Formation and
Fate in the Coastal Environment (PARFORCE) coastal field campaign at Mace Head.
According to the results, daytime ambient acid and ammonia concentrations were
significantly higher than required by model calculations to induce the formation of new
particles by homogeneous ternary nucleation. However, binary nucleation of sulphuric
acid-water molecules is not able to predict new particle formation since the binary
nucleation rate is far too small. We conclude that all particle formation events observed at
coastal sites can be initiated by ternary nucleation of sulphuric acid, ammonia, and water
vapor. However, related studies illustrate that ambient sulphuric acid concentrations are,
nevertheless, insufficient to explain observed rapid growth of particles from 1 to 3 nm
INDEX TERMS: 0305 Atmospheric
sizes which can be detected by current instrumentation.
Composition and Structure: Aerosols and particles (0345, 4801); 0365 Atmospheric Composition and
Structure: Troposphere—composition and chemistry; 0322 Atmospheric Composition and Structure:
Constituent sources and sinks; KEYWORDS: water, sulfuric acid, ammonia, nucleation
Citation: Kulmala, M., P. Korhonen, I. Napari, A. Karlsson, H. Berresheim, and C. D. O’Dowd, Aerosol formation during
PARFORCE: Ternary nucleation of H2SO4, NH3, and H2O, J. Geophys. Res., 107(D19), 8111, doi:10.1029/2001JD000900, 2002.
1. Introduction
[2] The formation and growth of atmospheric aerosols
have been studied experimentally and theoretically for some
time. Bursts of recently formed particles have been
observed in several regions around the world, for example,
in the free troposphere [Weber et al., 1999], in the marine
boundary layer [Covert et al., 1992], at coastal sites
[O’Dowd et al., 1998], in the vicinity of evaporating clouds
[Clarke et al., 1998], in Arctic areas [Wiedensohler et al.,
1996], in Antarctic areas [O’Dowd et al., 1997], in urban
1
Department of Physical Sciences, University of Helsinki, Helsinki,
Finland.
2
Air Quality Research, Finnish Meteorological Institute, Helsinki,
Finland.
3
Institutet för Tillämpad Miljöforskning, Air Pollution Laboratory,
Stockholm University, Sweden.
4
German Weather Service, Meteorological Observatory, Hohenpeissenberg, Germany.
5
National University of Ireland, Galway, Ireland.
6
Centre for Marine and Atmospheric Sciences, University of Sunderland, Sunderland, UK.
Copyright 2002 by the American Geophysical Union.
0148-0227/02/2001JD000900$09.00
PAR
areas and in stack plumes [Kerminen and Wexler, 1994],
and in boreal forests [Kulmala et al., 1998a; Mäkelä et al.,
1997]. Several nucleation mechanisms have been proposed
to explain this particle production, along with meteorological-related nucleation enhancement processes such as turbulent fluctuations, waves and mixing [Easter and Peters,
1994; Nilsson and Kulmala, 1998].
[3] It is generally accepted that the new particle formation in the atmosphere occurs via homogeneous heteromolecular nucleation process, in which two or more vapor
species form new stable particles. Typically, the formation
of atmospheric aerosols is attributed to binary nucleation of
water and sulphuric acid [Kulmala et al., 1998b]; however,
the binary theory is able to predict the nucleation rates only
at some extreme conditions of low temperatures, high
relative humidities, small preexisting-existing aerosol concentrations and at high sulphuric acid concentrations. It has
been shown in field measurements that there exist situations
where new particle formation cannot be explained with this
nucleation route alone [Covert et al., 1992; Hoppel et al.,
1994; Mäkelä et al., 1997; O’Dowd et al., 1999]. Since the
presence of ammonia in the aerosol particles considerably
decreases the vapor pressure of sulphuric acid above the
solution surface [e.g., Scott and Cattell, 1979] it has been
15 - 1
PAR
15 - 2
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
suggested that ammonia (NH3) forms new particles with
sulphuric acid [e.g., Scott and Cattell, 1979] or with
sulphuric acid and water [e.g., Coffmann and Hegg,
1995]. The recently developed ternary nucleation model
[Korhonen et al., 1999; Coffmann and Hegg, 1995] of
water-sulphuric acid-ammonia gives significantly higher
nucleation rates than those associated with binary nucleation and thus predicts nucleation under typical tropospheric
sulphuric acid concentrations (105 – 107 cm3) [Weber et
al., 1999] and ammonia mixing ratios (of the order of 10
ppt). Although using different thermodynamical data (equilibrium vapor pressures, surface tension) both models
[Korhonen et al., 1999; Coffmann and Hegg, 1995] give
reasonable agreement in nucleation rate predictions.
[4] In a coupled nucleation, coagulation and condensation study of aerosol evolution, Kulmala et al. [2000]
utilized the Korhonen et al. [1999] model to elucidate if
ternary nucleation mechanism could explain the production of new particles in the atmosphere and they found
that, under normal atmospheric conditions, ternary nucleation had the potential to form large amount of thermodynamically stable clusters of the order of 1 nm. However,
these clusters could not grow to detectable sizes of 3 nm
without condensation of an additional species since the
majority of the sulphuric acid was consumed in forming
stable clusters.
[5] In this study, we have further developed a temperature
dependent version of the recent ternary nucleation model
[Korhonen et al., 1999]. In the earlier version, the temperature range was very limited and the predictions were
mainly valid only at 298.15 K, while in the new version,
the thermodynamic part has been revised (see section 2).
The new model is used to predict ternary nucleation rates
under a wider range of atmospheric conditions (section 3).
[6] In particular, this study focuses on particle formation
events observed frequently in the coastal atmosphere during
the New Particle Formation and Fate in the Coastal Environment (PARFORCE) experiment [O’Dowd et al., 2002a,
2002b]. PARFORCE was a dedicated new particle field
experiment located at Mace Head, Ireland and comprised
two field campaigns. During the second campaign in 1999,
sulphuric acid and ammonia concentrations as well as
temperature and relative humidity were measured, thus
providing all necessary input parameters to the model.
Typically, in this environment, new particle formation events
are observed on a daily basis and occur under conditions of
low tide and solar radiation. Peak new particle concentrations often exceed 106 cm3. A full description of these
events is given by O’Dowd et al. [2002a, 2002b]. In this
paper we compare modeled ternary nucleation rates with
observed particle production rates (section 4) and also with
modeled binary nucleation rates. The results are discussed in
section 5, and the conclusions are given in section 6.
2. Theory for the Ternary Nucleation of Stable
H2O-NH3-H2SO4 Clusters
2.1. Cluster Formation
[7] In this section, we focus on the formation of new
particles via nucleation of stable H2O-NH3-H2SO4 clusters.
The ternary water-ammonia-sulphuric acid solution nuclei
are assumed to be in the liquid phase. The nucleation rate of
stable water-ammonia-sulphuric acid clusters (J ) is obtained
from
!
G*
J ¼ C exp :
kT
ð1Þ
where C is a kinetic factor. In this study the minimum work
for the critical nucleus formation is determined using the so
called revised classical theory. The minimized Gibbs free
energy change (in the limits of the capillarity approximation) is obtained from
4
2
G* ¼ pr* ss=a ;
3
ð2Þ
where r* is the critical radius of the cluster and ss/a is the
surface tension. The composition of the critical nucleus is
obtained from the following equations [Arstila et al., 1999]:
P1
2ss;a n1
P2
2ss;a n2
¼ kT ln
¼ ...
þ
þ
kT ln
Ps;1
r
P
r
s;2
Pi
2ss;a ni
¼0
þ
¼ kT ln
Ps;i
r
ð3Þ
where Pi is the ambient partial vapor pressure of species i,
Ps,i is the equilibrium vapor pressure of species i above the
flat solution surface, r is the radius of the cluster and ni is
the partial molecular volume of species i. When the critical
cluster is formed from water, ammonia and sulphuric acid
equation (3) becomes:
Pw
Ph2so4
nw ln
¼ 0;
nh2so4 ln
Ps;w
Ps;h2so4
Pw
Pnh3
nnh3 ln
nw ln
¼ 0:
Ps;w
Ps;nh3
ð4Þ
Here the subscript w refers to water, h2so4 to sulphuric acid
and nh3 to ammonia. From equation (4) one can solve the
composition of the critical cluster by numerical iteration.
When the composition of the cluster is known, the critical
radius is obtained from the Kelvin equation
r* ¼
2ss;a ni
:
Pi
kT ln
Ps;i
ð5Þ
In this context i refers either water or ammonia or sulphuric
acid. We have used recently presented rigorous kinetic
factor for the ternary system [Arstila et al., 1999]. In order
to solve the radius and composition of the critical cluster for
the systems presented above, one needs surface tension
(surface free energy), density of the solution and equilibrium vapor pressures of the various species above the flat
solution surface. When the classical nucleation theory is
used, the thermodynamical properties of the nucleus are
assumed to be those of the bulk substance in question.
2.2. Thermodynamical Model
[8] For the nucleation rate calculations, one needs various
thermodynamical data. When the classical nucleation theory
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
is used, the equilibrium vapor pressures of water, ammonia
and sulphuric acid above the flat solution surface, molecular
volumes of the species and the surface tension of the
solution is required.
[9] In the used thermodynamical model the equilibrium
vapor pressures above the flat solution surface are calculated from the equations [see, e.g., Bassett and Seinfeld,
1983; Kim et al., 1993]:
Ps;w
¼ aw ;
Pe;w
Ps;h2so4 ¼
Ps;nh3 ¼
g32h;so4 m2h mso4
;
Kh2so4
ð6Þ
g2nh4;oh mnh4 moh
;
Knh3 Knh4 aw
where Pe,w is the equilibrium vapor pressure of water above
the flat surface of pure substance, aw is the water (liquid
phase) activity, mi is the molality of species i (moles per kg
of pure water), Ki is the thermodynamical constant and gi,j is
the mean molal activity coefficient of ion i, j pair for the
multicomponent solution. In this context the subscript nh3
refers to ammonia, h2so4 to sulphuric acid, h to hydrogen
ions, oh to hydroxyl ions, nh4 to ammonium ions, and so4
to sulphate ions.
[10] In order to calculate the equilibrium vapor pressures
from the above equation (6) one has to solve the composition of the multicomponent solution. In the following the
dissociation of sulphuric acid to hydrogen (H+) and bisulphate ions (HSO
4 ) in the solution is assumed to be
complete. The dissociation of bisulphate ion to hydrogen
and sulphate (SO42) ions is described by the following
equation [e.g., Bassett and Seinfeld, 1983; Clegg et al.,
1994]:
gh gso4 mh mso4 g32h;so4 mh mso4
¼ 2
¼ Khso4 :
ghso4 mhso4
gh;hso4 mhso4
ð7Þ
Here gi is the molal activity coefficient of ion species i. The
concentrations of sulphate, bisulphate ammonia, hydroxyl
and hydrogen ions in the solution can be calculated with
numerical iteration from the following set of equations (for
further details, see Korhonen et al. [1998a]):
mhso4 ¼ ð1 aÞmh2so4 ;
mso4 ¼ amh2so4 ;
mnh4 ¼
m0nh3
g2nh4;oh Kw
1þ 2
gh;oh gnh3 Knh4
moh ¼
!;
ð8Þ
Kw aw
;
g2h;oh mh
mh þ mnh4 mhso4 2mso4 moh ¼ 0:
0
Here mnh3
is the total amount of ammonium species
(ammonia and ammonium ions) in the solution, gnh3 is
the molal activity coefficient for the aqueous ammonia, and
a is the degree of bisulphate dissociation, which is obtained
PAR
15 - 3
to the used form by combining the equation (7) and above
equations for bisulphate and sulphate ions:
a¼
g2h;hso4 Khso4
g32h;so4 mh þ g2h;hso4 Khso4
:
ð9Þ
[11] The used set of equations (8) and (9) depends now
on the given concentrations of total ammonia and sulphuric
acid, hydrogen ions (which is the variable for the iteration),
various thermodynamical constants, water activity for the
multicomponent solution and the mean molal activity coefficients.
[12] In this model the thermodynamical constants and
their temperature dependence for water, ammonia and
ammonium ions is obtained from Kim et al. [1993] and
for bisulphate ion from Clegg and Brimblecombe [1995].
The thermodynamic constant for sulphuric acid at 298.15 K
is obtained by fitting on the vapor pressure data presented
by Marti et al. [1997] and the temperature dependence for
the constant is obtained by using the partial molal enthalpies
and the partial molal heat capacities presented by Bassett
and Seinfeld [1983]. In practice the thermodynamic constant
for sulphuric acid has been divided by a factor of 60. The
choice has been done in this way in order to be able to
obtain agreement with experimental nucleation rates at the
binary limit [see Korhonen et al., 1999].
[13] The water activity for the multicomponent solution is
calculated according to Stelson et al. [1983] and the mean
molal activity coefficients by using the mixing rule presented originally by Kusik and Meissner [1978].
[14] The used methods for the calculation of the water
activity and the mean molal activity coefficients requires
water activities and activity coefficients for the binary
(aqueous) solutions of sulphuric acid (separately for HHSO4
and H2SO4), ammonium bisulphate (NH4HSO4), ammonium sulphate ((NH4)2SO4) and ammonium hydroxide
(NH4OH).
[15] The mixing rule needs the binary activity coefficients
for water - HHSO4 and water-H2SO4, each isolated in the
solution. This means that there would exist only bisulphate
or sulphate ions in the aqueous solution in question. In this
context we have followed the work of Jacobson et al. [1996]
and used the model for the thermodynamic properties of
aqueous sulphuric acid presented by Clegg and Brimblecombe [1995]. The model gives mixed activity coefficients
for the aqueous sulphate and bisulphate at the temperature
range 200 to 328 K from which we separate the required
binary activity coefficients numerically.
[16] The temperature dependent binary activity coefficient for the aqueous ammonium bisulphate and sulphate
are obtained by using the binary activity coefficient at
298.15 K, the relative apparent molal enthalpy, and the
apparent molal heat capacity [see Harned and Owen, 1958;
Jacobson et al., 1996].
[17] The binary activity coefficient for aqueous ammonium bisulphate and sulphate at 298.15 K is obtained by
using Gibbs-Duhem equation and the data for the water
activity presented by Tang and Munkelwitz [1994]. The
used equations are presented by Korhonen et al. [1998b].
[18] The temperature dependency of the binary activity
coefficient of aqueous ammonium sulphate is obtained from
PAR
15 - 4
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
Figure 1. The partial derivatives of the relative apparent molal enthalpy and the apparent molal heat
capacity with respect to the molality of the solution.
Bassett and Seinfeld [1983]. If the temperature or the
solution concentration goes out of the valid ranges during
the model calculations, numerical extrapolation is used.
[19] The temperature dependent binary activity coefficient (gnh4,hso4) and water activity (aw) for water-ammonium
bisulphate solution is obtained by using the following
equations [see also Jacobson et al., 1996]:
@fcp
Mw m2 TL @fL
þ TC
;
1000R* T0 @m
@m
TL
@fL
0
¼ ln gnh4;hso4 þ
f þm
2R*T0 L
@m
@f
TC
cp
f þm
þ f0cp ;
þ
2R* cp
@m
ln aw ¼ ln a0w ln gnh4;hso4
ð10Þ
where aw0 is the water activity for the system at 298.15 K,
g0nh4,hso4 is the mean binary activity coefficient at 298.15 K,
fL is the relative apparent molal enthalpy, fcp is the
apparent molal heat capacity, m is molality of the solution
0
is the apparent molal heat
(mol salt per kg of water), fcp
capacity at infinite dilution, Mw is the molar mass of water,
R* is the ideal gas constant, T0 is the reference temperature
(298.15 K), TL = (T0/T ) 1 and TC =1 + ln(T0/T ) (T0/T ).
Here T is temperature in Kelvin degrees.
[20] Unfortunately, the relative apparent molal enthalpy
and the apparent molal heat capacity for ammonium bisulphate are not available at literature for the needed wide
temperature and concentration ranges. Instead we have
obtained them by fitting and using the following equations
[see Jacobson et al., 1996]:
fL ¼ U1 m1=2 þ U2 m þ U3 m3=2 þ . . . ;
fcp ¼ f0cp V1 m1=2 þ V2 m þ V3 m3=2 þ . . . ;
ð11Þ
where Ui and Vi are the fitting coefficients. From equation
(11) we also get the required partial derivatives:
@fL U1 1=2
3U3 1=2
¼
m
m þ ...;
þ U2 þ
@m
2
2
@fcp V1 1=2
3V3 1=2
¼ m
m þ ...
þ V2 þ
@m
2
2
ð12Þ
The fitting coefficients are obtained by using the water
activity equation (10) and equation (12). First the water
activities for the system (in temperatures from 253.15 to
303.15 K and molalities up to 270 mol/kg) are derived from
the model of Clegg et al. [1998] (http://www.hpc1.uea.
ac.uk/ ~e770/aim.html). With these water activities the values
for the partial derivatives of the relative apparent molal
enthalpy and the apparent molal heat capacity in the different
molalities are obtained by fitting on the above water activity
equation. Then the partial derivatives are fitted on equation
(12) from which the fitting coefficients are finally obtained.
Figure 1 presents the used fitting points (circles) and the
obtained fittings (solid line) for the partial derivatives. The
obtained fitting coefficients are shown in Table 1.
[21] Due to the lack of published data for the molal binary
activity coefficient for ammonium hydroxide, we must
Table 1. Fitting Coefficients for the Relative Apparent Molal
Enthalpy fL and the Apparent Molal Heat Capacity fcp
i
Ui
Vi
1
2
3
4
5
6
7
10548.498
2882.2061
405.91066
30.659585
1.0641795
8.2138068E-4
5.17477249E-7
540.11296
157.79882
23.676279
1.88999554
6.87856641E-2a
5.72195606E-5
3.80497946E-8
a
Read E-x as 10x.
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
calculate it based on other physical properties. In this model
the molal activity coefficient is obtained by calculating first
the equilibrium ammonium vapor pressure and water activity
from the equations presented by Koutrakis and AurianBlajeni [1993]. The molal binary activity coefficient for
ammonium hydroxide is then obtained by using the equilibrium ammonia vapor pressure, the water activity and the
thermodynamical constants presented by Kim et al. [1993].
[22] The molal activity coefficient for aqueous ammonia
is also obtained by using the data presented by Koutrakis
and Aurian-Blajeni [1993]. Although this assumption is
made only because of missing data, it does not appear to
cause any major errors. We have reached this conclusion for
two reasons: first, there exist only very small amounts of
ammonia in the solutions relevant in the these ternary
nucleation studies and second, the used activity coefficient
is very near unity in the studied cases (i.e., the coefficient
could be replaced with 1 without any major changes in the
results).
[23] The calculation of the temperature-dependent binary
water activities in the model is done with the model
presented, e.g., by Harned and Owen [1958] or by Jacobson et al. [1996]. The model needs the water activity for the
binary solution at 298.15 K, the relative apparent molal
enthalpy and the apparent molal heat capacity.
[24] The required binary water activities at 298.15 K for
water-HHSO4 and water-H2SO4 are calculated using the
Gibbs-Duhem equation to the binary activity coefficients
given by the model of Clegg and Brimblecombe [1995].
The apparent molal enthalpy and the apparent molal heat
capacity for these species are also obtained by fitting the
results given by the same model at the temperature range
200 to 328 K.
[25] The binary water activities for the water-ammonium
bisulphate and water-ammonium sulphate solutions at
298.15 K are obtained from Tang and Munkelwitz [1994].
The apparent molal enthalpy and the apparent molal heat
capacity for these species are obtained as described in the
context of the binary activity coefficients for the species.
[26] The binary water activity for water ammonium
hydroxide solution is obtained from Koutrakis and
Aurian-Blajeni [1993]. The used surface tensions and densities are given by Korhonen et al. [1999].
3. Model Predictions Under Atmospheric
Conditions
[27] The developed model has been used to account for
nucleation rates under different ambient conditions and also
to study if ternary nucleation is possible for the reported
coastal particle formation events at Mace Head during the
PARFORCE campaign. Around 300 model runs have been
performed.
[28] In general, the results of the model study suggest that
nucleation of water-ammonia-sulphuric acid clusters occurs
significantly easier than the nucleation of water-sulphuric
acid clusters in the atmospheric conditions. The results also
suggest that the composition of the critical clusters in the
atmospheric conditions is typically more acidic than that of
water-ammonium bisulphate solution. The formed H2ONH3-H2SO4 clusters are very small and highly concentrated
in sulphuric acid and ammonia.
PAR
15 - 5
[29] The nucleation rates as a function of temperature,
ammonia mixing ratios, and sulphuric acid concentrations
are given in Table 2 for relative humidities of 52% and 90%.
The results indicate that the nucleation rate does not
increase significantly from 52% up to 90% for the ternary
system. Presumably, above 52% relative humidity, there is
sufficient water vapor available so that this is not a limiting
factor in the process. The lack of dependence may also be
caused by the absence of an explicit hydration correction in
the present model; however, the effect of hydrates is
implicitly taken account since the thermodynamic equilibrium constant for the dissociation of sulphuric acid is
adjusted to achieve better agreement with experimental
results at the water-sulphuric acid limit, and, in this limit,
hydrates are included. Owing to the scarcity of the experimental data, we can only estimate the accuracy of the
model for the full ternary system. In addition to the
application of classical nucleation theory, which relies on
the capillarity approximation, the most important source of
error is the uncertainties in the equilibrium vapor pressures
in the thermodynamical model. Due to these inherent
problems we expect the error bars in the nucleation rate
values to be of 3 –5 orders of magnitude.
[30] The nucleation rate is a strong function of sulphuric
acid concentration, ammonia mixing ratio and temperature.
In Figure 2, the ternary nucleation rate is presented as a
function of ammonia mixing ratio at different temperatures
and sulphuric acid concentrations. The effect of ammonia in
enhancing nucleation rates at low ammonia mixing ratios is
pronounced compared to higher ammonia mixing ratios,
where the ammonia effect will saturate. For example, if the
ammonia mixing ratio increases from 1 pptv to 5 pptv, the
nucleation rate increases several orders of magnitude while
a further increase from 5 pptv to 25 pptv results in about 1
order of magnitude increase.
[31] The ternary and binary nucleation rates as a function
of sulphuric acid concentration are presented in Figure 3.
Sulphuric acid is the key molecule in ternary nucleation and
even small sulphuric acid concentrations are enough to form
new stable clusters. For example, at 258.15 K, only 104
molecules cm3 are needed for significant nucleation, i. e. at
least 106 cm3s1 (see also Table 2). If the acid concentration is increased by one order of magnitude, a typical
increase in nucleation rate is 6 orders of magnitude,
although this dependency reduces at high nucleation rates
and increases at smaller nucleation rates. By comparison,
the binary nucleation rate is significantly smaller than the
ternary nucleation rate. Even at ammonia mixing ratio of 1
pptv, the ternary nucleation rate is over 10 orders of
magnitude higher than the binary nucleation rate, and to
match the binary nucleation rate with that predicted from
the ternary system, the sulphuric acid concentration must be
50 times higher compared to the ternary system.
[32] In Figure 4 the nucleation rate as a function of
temperature is presented. When the temperature is decreased
the nucleation is increased significantly, e. g. the nucleation
rate of 1 cm3s1 is obtained with much smaller acid
concentrations when ammonia mixing ratio is constant (like
5 ppt). At acid concentrations 104 cm3, 105 cm3, and 106
cm3 the corresponding temperatures required for the
nucleation rate of 1 cm3s1 are around 270 K, 280 K
and 290 K, respectively.
PAR
15 - 6
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
Table 2. Nucleation Rate J As a Function of Temperature,
Ammonia mixing ratio, and Sulphuric Acid Concentration at RH =
90%a
T (K)
NH3
( ppt)
H2SO4
(m3)
J (cm3s1)
RH = 90%
258.15
5.0
1.E10
0.20E8
258.15
5.0
1.E11
0.11E11
258.15
25.0
1.E10
0.38E9
263.15
0.2
1.E10
0.10E2
263.15
1.0
1.E10
0.11E7
263.15
5.0
1.E10
0.30E7
263.15
25.0
1.E10
0.19E8
268.15
1.0
1.E10
0.12E2
268.15
1.0
1.E11
0.12E8
268.15
5.0
1.E10
0.10E5
268.15
5.0
1.E11
0.51E9
268.15
5.0
1.E12
0.12E12
268.15
25.0
1.E10
0.34E5
268.15
25.0
1.E11
0.40E9
268.15
25.0
1.E12
0.42E12
268.15
25.0
1.E13
0.60E13
273.15
0.2
1.E11
0.49E-10
273.15
1.0
1.E10
0.25E-9
273.15
5.0
1.E10
0.11E-1
273.15
5.0
1.E11
0.55E6
273.15
25.0
1.E10
0.17E3
273.15
25.0
1.E11
0.20E8
273.15
25.0
1.E12
0.28E11
278.15
0.2
1.E12
0.11E-10
278.15
0.2
1.E13
0.33E4
278.15
0.2
1.E14
0.10E10
278.15
1.0
1.E11
0.15E-8
278.15
5.0
1.E11
0.98E1
278.15
5.0
1.E12
0.17E9
278.15
25.0
1.E11
0.95E4
278.15
25.0
1.E12
0.20E10
278.15
25.0
1.E13
0.13E13
283.15
0.2
1.E13
0.15E-7
283.15
1.0
1.E12
0.93E-7
283.15
5.0
1.E11
0.30E-8
283.15
5.0
1.E12
0.25E5
283.15
25.0
1.E11
0.64E0
283.15
25.0
1.E12
0.23E8
283.15
25.0
1.E13
0.46E12
288.15
0.2
1.E14
0.21E-3
288.15
0.2
1.E15
0.63E6
288.15
1.0
1.E13
0.11E-2
288.15
1.0
1.E14
0.55E7
288.15
1.0
1.E15
0.68E11
288.15
5.0
1.E12
0.48E-3
288.15
5.0
1.E13
0.95E8
288.15
5.0
1.E14
0.20E11
288.15
25.0
1.E12
0.31E5
288.15
25.0
1.E13
0.85E10
288.15
25.0
1.E14
0.95E13
288.15
125.
1.E11
0.50E-1
288.15
125.
1.E12
0.54E8
288.15
125.
1.E13
0.15E12
293.15
0.2
1.E14
0.51E-13
293.15
0.2
1.E15
0.29E2
293.15
1.0
1.E14
0.61E2
293.15
1.0
1.E15
0.84E9
293.15
5.0
1.E13
0.15E3
293.15
25.0
1.E12
0.64E0
293.15
25.0
1.E13
0.18E9
293.15
125.
1.E11
0.25E-9
293.15
125.
1.E12
0.42E5
293.15
125.
1.E13
0.13E11
298.15
0.2
1.E15
0.50E-3
298.15
1.0
1.E15
0.46E5
298.15
5.0
1.E13
0.30E-9
298.15
25.0
1.E13
0.13E7
a
For comparison, some values for the nucleation rate at RH
shown in the last column.
RH = 52%
0.92E7
...
0.23E9
...
...
0.45E7
0.18E8
...
...
...
0.67E9
...
...
0.56E9
0.10E12
...
...
0.10E-7
0.95E-1
0.10E7
0.27E4
0.20E8
0.21E11
...
0.23E4
...
0.69E-4
0.58E-2
...
0.84E5
0.14E10
...
0.13E-5
...
0.22E-6
0.18E6
0.39E2
0.22E8
0.20E12
0.63E-3
...
0.18E1
...
...
0.11E0
...
...
0.18E5
0.97E10
...
...
...
...
...
...
0.10E4
...
...
0.35E1
0.24E9
...
...
...
0.31E-5
0.70E5
...
0.44E7
= 52% are
[33] The ternary system of sulphuric acid-ammonia-water
clusters is thermodynamically very complex. Although we
are able to predict nucleation rates as a function of temperature, one should keep in mind that all predictions are
preliminary as far as there are enough experimental data
to compare. All existing laboratory data [Ball et al., 1999;
D. Hanson, personal communication, 2000] so far are still
preliminary. However, the model results presented here
agree qualitatively with those aforementioned data,
although the small size of the critical nucleus (typically
20 molecules in our calculations with 8 H2SO4, 5 NH3, and
7 H2O molecules) casts some doubt on the reliability of
nucleation prediction by classical nucleation theory based
on capillarity approximation.
4. Comparison of Model Predictions to Coastal
Observations
[34] During the PARFORCE coastal particle formation
events, sulphuric acid and ammonia concentrations were
measured simultaneously during new particle production
events. In clear air, significant particle production (>105
cm3s1) was observed for sulphuric aid gas phase concentration as low as 2 106 molecules cm3 and ammonia
mixing ratios of the order of 25 ppt, while under polluted
conditions, particle production was observed for sulphuric
acid concentrations as high as 1.2 108 molecules cm3
and ammonia mixing ratios of the order of 1000 ppt
[O’Dowd et al., 2002b; Berresheim et al., 2002; O’Dowd
and Hämeri, 2000]. Particle formation events were
observed on 90% of the campaign days and occurred in
all air masses. A complete classification of these coastal
particle formation events and the environmental conditions
(e.g., temperature, humidity, sulphuric acid, ammonia, etc.)
under which these occurred is given by O’Dowd et al.
[2002b].
[35] Using the experimental gas phase concentration data
as input, together with RH and T, ternary nucleation rates
have been calculated for these conditions. For all PARFORCE conditions, typical of a mid-latitude coastal environment, under which particle production events are
observed, ternary nucleation of sulphuric acid, water and
ammonia particles is predicted to occur according to classical ternary nucleation theory (see Figure 5).
[36] Typically, in all cases, the sulphuric acid concentration is over 2 106 cm3 and ammonia mixing ratio over
20 ppt. Even in extreme cases when sulphuric acid concentrations are around 106 cm3 and ammonia mixing ratio
only 4 ppt, the ternary nucleation is predicted to occur since
the temperature is around 285 K. Furthermore, ternary
nucleation is predicted under most daytime conditions
reported during the campaign even when no new particles
are observed. By comparison, the binary nucleation rate is
too small under these conditions. For example, the binary
nucleation rate under the most favorable conditions (T =
288.15 K, RH = 90%, and sulphuric acid concentration 3 108 molecules cm3) is around 1 104 cm3 s1. Such a
nucleation rate is much too low to produce the concentrations of particles observed in the coastal environment.
[37] Six Case Study particle production days have been
selected for more detailed investigations [see O’Dowd et al.,
2002b]. In general, the particle production events observed
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
PAR
Figure 2. The ternary nucleation rate as a function of ammonia mixing ratio.
Figure 3. The ternary and binary nucleation rates as a function of sulphuric acid concentration.
15 - 7
PAR
15 - 8
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
Figure 4. The ternary nucleation rate as a function of temperature.
at Mace Head could be categorized into three different
primary event types.
[38] Type I events were typical of clean air arriving at the
station from the west and south west, having passed over
one tidal region, approximately 100 m directly upwind of
the sampling location, presumed to the source of the
particles precursors [see O’Dowd et al., 2002b]. Under
these conditions, rapid growth of new particles up to >3
nm is observed, with few of these having sufficient time to
grow to 10 nm. It should be noted that under Type I
conditions, for typical wind speeds of 5 m/s, particles grow
from stable embryos to detectable 3 nm particles, and larger,
in 25 seconds.
[39] By comparison, Type II events were typical of clean
marine air having passed over multiple tidal source regions
up to 10– 20 km upwind, including the local region responsible for Type I events. Type III events were characterized
as those occurring in polluted air masses and having
advected over source regions 1– 2 km upwind. The primary
differences seen between the three types are as follows:
Type II events generally produce more particles than Type I
events [see O’Dowd et al., 2002b] with concentrations
reaching more than 106 cm3, compared to 2 – 3 105
cm3 for Type I events. During the polluted events, concentrations are typically of the same magnitude as Type I
events; however, the nucleation mode peak has grown to
almost 10 nm, compared to 3– 5 in the Type I event. The six
cases cover two of each event types. For the Type I and II
events since the measurement point was a few tens of
meters from the perceived source region, and using the rate
of change of particle concentration between 3 and 10 nm on
a 1 Hz time bases, we could derive an estimate for the
source rate of 3 nm particles [O’Dowd et al., 2002b].
Average formation rates for these cases were between 104
and 105 cm3 s1, with peak source rates reaching 8 105
cm3 s1. These rates for 3 nm particles are tabulated along
with temperature, ammonia and sulphuric acid as well as
Figure 5. Threshold for significant ternary nucleation (106
cm3s1). The line is for T = 288 K and RH = 90%.
PARFORCE data are marked with a square.
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
PAR
15 - 9
Table 3. Ternary Nucleation During the Case Daysa
Case
JD
T
NH3
H2SO4
FRmean
A
B
C
D
E
F
157
163
164
165
175
176
285
285
286
286
289
288
100
25
20
25
900
1300
1 – 1.5 107
0.2 107
0.6 – 0.8 107
0.8 – 1.2107
3 – 12107
6 – 8107
5 103
1 105
1 105
2 105
not applicable
not applicable
FRpeak
4 104
7 105
8 105
8 105
not applicable
not applicable
Model
Type
>1010
>1010
>1010
>1010
>1010
>1010
II
II
I
I
III
III
a
FR is the formation rate of 3 nm particles; mean and peak values are tabulated. Temperature T is expressed in kelvins, NH3 mixing ratio in ppt, and
H2SO4 concentration in cm3. FRmean, FRpeak, and model prediction are expressed in cm3s1. JD is Julian day.
modeled nucleation rates in Table 3. It should be noted that
due to the distance from the source region (1 – 2 km) and
their bigger size, it is not possible to calculate particle
source rates for Type III events. By comparison, the
model-predicted nucleation rates for each of these events
were greater than 1010 cm3 s1 (i.e., at least 5 orders of
magnitude above the observed particle formation rates).
These case study model results indicate that ternary nucleation is highly probable in all of the selected cases (see
Table 3).
5. Discussion
[40] The previous sections compared theoretical ternary
nucleation rates for sulphuric acid, ammonia, and water for
a range of common atmospheric conditions. The results
suggest that significant ternary nucleation occur under all
conditions and concentrations of sulphuric acid and ammonia measured during PARFORCE campaign. However, in
practice, the presence of new aerosol particles are infrequently observed in the atmosphere. The reason for this
inconsistency is that current day aerosol instrumentation can
only detect particles as small as 3 nm while nucleation is
said to occur when stable embryos, typically of the order of
1 nm in size, are produced. In other words, we cannot
measure aerosol nucleation but we can only detect it sometimes when there is sufficient vapor to grow nucleated
particles to detectable sizes.
[41] The particle source rates presented above, and
reported in more detail by O’Dowd et al. [2002b], are for
3 nm particles and are not, as such, nucleation rates since
coagulation significantly reduces the total concentration of
particles between 1 and 3 nm over the evolution, even over
the short time scale involved. Actual nucleation rates are
expected to be somewhat higher, depending on the preexisting-existing coagulation sink available. In an associated
study utilizing an aerosol dynamics model to reproduce the
observed particle concentrations [Pirjola et al., 2002], it
was found that the required nucleation rates were between
3 105 cm3 s1 (for a condensable vapor of 1 1010 cm3)
and 5 107 cm3 s1 (for a condensable vapor concentration of 5109 cm3). These nucleation rates are significantly smaller than those predicted by the ternary model.
[42] Even accounting for the difference between the
experimental sources rates and the estimated scaling of 3
nm source rates to actual nucleation rates (illustrated
above), the theoretical nucleation rates are much higher
than sulphuric acid and ammonia concentrations, and with
these nucleation rates the vapor concentrations will
decrease very rapidly. Therefore the modeled nucleation
rate should be considered to be an instantaneous rate. On
the other hand, the real nucleation rate could be somewhat
smaller and it is very probably kinetically limited. Since,
the nucleation rates are predicted using the observed
ambient sulphuric acid and ammonia concentrations, there
is no need to take into account the competition between
condensation and nucleation.
[43] Given the predicted nucleation rates, ternary nucleation is able to predict the formation of new 1 nm particles,
but it is not able to predict the formation of 3 nm particles
since most, if not all, of the sulphuric acid is depleted. Even
if not depleted the sulphuric acid concentration is still too
small to grow particles from 1 nm to 3 nm.
[44] It should be also noted that the sulphuric acid
concentrations required for the ternary nucleation in the
above cases is still too small to be able to explain the growth
of 1 nm particles to 3 nm size [see Kulmala et al., 2000;
O’Dowd et al., 1999]. This implies that some additional,
nonsulphuric acid vapors are required for condensation
growth.
[45] The requirement for an additional vapor source to
grow stable clusters to observable sizes is corroborated by
detailed condensation and aerosol dynamics studies that
illustrate that the minimum condensable vapor concentration is > 109 cm3 (i.e., more than 2 orders of magnitude
higher than the measured sulphuric acid concentration
[Pirjola et al., 2002].
[46] In all the aforementioned coastal particle production
events, particle growth occurs very rapidly [O’Dowd et al.,
2002b; Pirjola et al., 2002] with growth rates greater than
100 nm/h [e.g., Dal Maso et al., 2002]. For the case with the
highest measured sulphuric acid concentration, 1.1 108
cm3, and the lowest corresponding growth rate of around
100 nm/h [Kulmala et al., 2001; Dal Maso et al., 2002], the
sulphuric acid concentration can account for only about
10% of the growth.
[47] From a number of theoretical [this study; Pirjola et
al., 2002] and analytical [Dal Maso et al., 2002] approaches,
it is clear that while sulphuric acid is very likely to form
stable sulphate clusters, it is incapable of producing new
detectable 3 nm particles and an additional source of condensable vapor is required. This is further corroborated
through experimental measurements of hygroscopic growth
on 8 nm particles (it was not possible to conduct hygroscopic
growth experiments on smaller particles). Väkevä et al.
[2002] demonstrate that the hygroscopic growth factors of
these 8 nm particles are characteristic of insoluble material,
although during periods of peak sulphuric acid production,
sulphuric acid can increase their solubility. Furthermore, xray dispersive analysis of 6 nm particles indicate the pres-
PAR
15 - 10
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
ence of predominantly iodine, a derivative of biogenic
halocarbon emissions, in all particles, with some presence
of sulphur [Mäkelä et al., 2002].
[48] With the given experimental and theoretical results
from studies in these coastal nucleation evens, it appears
that particle production requires first the production of
stable clusters via ternary nucleation of sulphuric acid,
ammonia and water, and secondly, biogenic iodine to
condense and grow the clusters into detectable particle
sizes. The biogenic iodine is thought to result from the
photolysis of CH2I2, emitted from the coastal biota [Carpenter et al., 1999], in the presence of ozone. This reaction
produces IO which, along with CH2I2, is observed to posses
a tidal cycle in concentration [Alicke et al., 1999]. It should
be noted, however, that subsequent laboratory studies
[Hoffmann et al., 2001] illustrated that massive amounts
of particles could be produced from the photolysis of CH2I2
in the presence of ozone and it is suggested that iodine
oxides could also participate in nucleation, as well as
condensation. Further theoretical development is required
to explore whether or not iodine oxides contribute to the
nucleation process or whether the current estimate that
ternary nucleation of sulphuric acid, water and ammonia
is the most likely nucleation mechanism for coastal particle
cluster production.
6. Conclusions
[49] In the present study, we have developed a new
temperature dependent model for ternary nucleation. In
the model calculations, we have shown that ternary nucleation is highly probable for very minor sulphuric acid
concentrations (of less than 105 – 106 cm3) and ammonia
mixing ratios of around 1 pptv. The model itself, however,
contains several problems: Besides those known for classical nucleation theory, the ternary system has additional
extremely complex thermodynamics, therefore the development and improvement of ternary nucleation theory will
continue in the future.
[50] During the PARFORCE campaign, coastal particle
formation events are not correlated with peak sulphuric acid
concentrations and are more connected to low tide conditions, and presumably biogenic emissions from these tidal
regions [O’Dowd et al., 2002; Berresheim et al., 2002].
Nevertheless, it is possible that the coastal particle formation events are still driven by sulphuric acid nucleation since
the ternary nucleation model indicates that new particle
formation will occur almost continuously for the given
conditions, but there is never enough sulphuric acid to grow
the particles to observable sizes.
[51] The growth to observable sizes occurs then via
condensation of other vapors emitted from open seashore.
This finding supports the idea of decoupling between
nucleation and condensation growth [see Kulmala et al.,
2000]. Nevertheless, we do not have a direct proof of this
phenomenon since by using the present state-of-art instrumentation it is impossible to determine the composition of
particles with diameters between 1 and 5 nm. On the other
hand, binary nucleation of sulphuric acid and water could
be ruled out as an explanation as it is not able to replicate
observed particle production events, or approach near the
required nucleation rates. Although it was recently hypothe-
sized that iodine oxides could also be implicated in these
coastal particle formation events, there is insufficient understanding or knowledge of that system to evaluate whether or
not it is more probable than the ternary sulphuric-acid-water
nucleation system, which, on the basis of the current theory
given here, is viewed as the system we have most confidence in with respect to particle nucleation.
References
Alicke, B., K. Hebestreit, J. Stutz, and U. Platt, Iodine oxide in the marine
boundary layer, Nature, 397, 572 – 573, 1999.
Arstila, H., P. Korhonen, and M. Kulmala, Ternary nucleation: Kinetics and
application to water-ammonia-hydrochlorid acid system, J. Aerosol Sci.,
30, 131 – 138, 1999.
Ball, S. M., D. R. Hanson, F. L. Eisele, and P. H. McMurry, Laboratory
studies of particle nucleation: Initial results for H2SO4, H2O, and NH3
vapors, J. Geophys. Res., 104, 23,709 – 23,718, 1999.
Bassett, M., and J. H. Seinfeld, Atmospheric equilibrium model of sulfate
and nitrate aerosols, Atmos. Environ., 17, 2237 – 2252, 1983.
Berresheim, H., T. Elste, H. G. Tremmel, C. D. O’Dowd, A. G. Allen, H.-C.
Hansson, K. Rossman, M. Dal Maso, J. M. Mäkelä, and M. Kulmala,
Gas-aerosol relationships of OH, H2SO4 and MSA: Observations in the
coastal marine boundary layer at Mace Head, Ireland, J. Geophys. Res.,
107, 10.1029/2000JD000229, in press, 2002.
Carpenter, L. J., W. T. Sturges, S. A. Penkett, P. Liss, A. Alicke, K. Hebestreit, and U. Platt, Short-lived alkyl iodides and bromides at Mace
Head, Ireland: Links to biogenic sources, J. Geophys. Res., 104, 1679 –
1689, 1999.
Clarke, A. D., et al., Particle production in the remote marine atmosphere:
Cloud outflow and subsidence during ACE 1, J. Geophys. Res., 103,
16,397 – 16,409, 1998.
Clegg, S. L., and P. Brimblecombe, Application of a multicomponent thermodynamic model to activities and thermal properties of 0 – 40 mol kg1
aqueous sulfuric acid from <200 to 328 K, J. Chem. Eng. Data, 40, 43 –
64, 1995.
Clegg, S. L., J. A. Rard, and K. S. Pizer, Thermodynamic properties of 0 – 6
mol kg1 aqueous sulfuric acid from 273.15 to 328.15 K, J. Chem. Soc.
Faraday Trans., 90, 1875 – 1894, 1994.
Clegg, S. L., P. Brimblecombe, and A. S. Wexler, A thermodynamic model
of the system H-NH4 -SO4 -NO3 -H2O at tropospheric temperatures,
J. Phys. Chem. A, 102, 2137 – 2152, 1998.
Coffmann, D. J., and D. A. Hegg, A preliminary study of the effect of
ammonia on particle nucleation in the marine boundary layer, J. Geophys.
Res., 100, 7147 – 7160, 1995.
Covert, D. S., V. N. Kapustin, P. K. Quinn, and T. S. Bates, New particle
formation in the marine boundary layer, J. Geophys. Res., 97, 20,581 –
20,589, 1992.
Dal Maso, M., M. Kulmala, J. M. Mäkelä, P. Aalto, and C. D. O’Dowd,
Condensation and coagulation sinks and formation of nucleation mode
particles in coastal and boreal forest boundary layers, J. Geophys. Res.,
107, 10.1029/2001JD001053, 2002.
Easter, R. C., and L. K. Peters, Binary homogeneous nucleation: Temperature
and relative humidity fluctuations, nonlinearity, and aspects of new particle
production in the atmosphere, J. Appl. Meteorol., 33, 775 – 784, 1994.
Harned, H. S., and B. B. Owen, The Physical Chemistry of Electrolytic
Solutions, Van Nostrand, New York, 1958.
Hoffmann, T., C. D. O’Dowd, and J. H. Seinfeld, Iodide oxide homogeneous nucleation: An explanation for coastal new particle production,
Geophys. Res. Lett., 28, 1949 – 1952, 2001.
Hoppel, W. A., G. M. Frick, and J. W. Fitzgerald, Marine boundary layer
measurements of new particle formation and the effects nonprecipitating
clouds have on aerosol size distribution, J. Geophys. Res., 99, 14,443 –
14,495, 1994.
Jacobson, M. Z., A. Tabazadeh, and R. P. Turco, Simulating equilibrium
within aerosols and nonequilibrium between gases and aerosols, J. Geophys. Res., 101, 9079 – 9091, 1996.
Kerminen, V.-M., and A. S. Wexler, Post-fog nucleation of H2SO4-H2O
particles in smog, Atmos. Environ., 28, 2399 – 2406, 1994.
Kim, Y. P., J. H. Seinfeld, and P. Saxena, Atmospheric gas-aerosol equilibrium, I, Thermodynamic model, Aerosol Sci. Tech., 19, 157 – 182,
1993.
Korhonen, P., A. Laaksonen, Y. Viisanen, and M. Kulmala, The homogeneous heteromolecular nucleation of sulfuric acid and ammonia in the
atmosphere, Rep. Ser. Aerosol Sci., 41, 110 – 115, 1998a.
Korhonen, P., A. Laaksonen, E. Batris, and Y. Viisanen, Thermodynamics
for highly concentrated water-ammonium sulfate solutions, J. Aerosol
Sci., 29, S379 – S380, 1998b.
Korhonen, P., M. Kulmala, A. Laaksonen, Y. Viisanen, R. McGraw, and
KULMALA ET AL.: PARFORCE—TERNARY NUCLEATION
J. H. Seinfeld, Ternary nucleation of H2SO4, NH3 and H2O in the atmosphere, J. Geophys. Res., 104, 26,349 – 26,353, 1999.
Koutrakis, P., and B. Aurian-Blajeni, Measurement of partial vapor pressure
of ammonia over acid ammonium sulfate solutions by an integral method,
J. Geophys. Res., 98, 2941 – 2948, 1993.
Kulmala, M., A. Toivonen, J. M. Mäkelä, and A. Laaksonen, Analysis of
the growth of nucleation mode particles observed in boreal forest, Tellus,
Ser. B, 50, 449 – 462, 1998a.
Kulmala, M., A. Laaksonen, and L. Pirjola, Parameterizations for sulphuric
acid/water nucleation rates, J. Geophys. Res., 103, 8301 – 8308, 1998b.
Kulmala, M., L. Pirjola, and J. M. Mäkelä, Stable sulphate clusters as a
source of new atmospheric particles, Nature, 404, 66 – 69, 2000.
Kulmala, M., M. Dal Maso, J. M. Mäkelä, L. Pirjola, M. Väkevä, P. Aalto,
P. Miikkulainen, K. Hämeri, and C. D. O’Dowd, On the formation,
growth and composition of nucleation mode particles, Tellus, Ser. B,
53, 479 – 490, 2001.
Kusik, C. L., and H. P. Meissner, Electrolytic activity coefficients in inorganic processing, AIChE Symp. Ser., 173, 14 – 20, 1978.
Mäkelä, J. M., P. Aalto, V. Jokinen, T. Pohja, A. Nissinen, S. Palmroth,
T. Markkanen, K. Seitsonen, H. Lihavainen, and M. Kulmala, Observations of ultrafine aerosol particle formation and growth in boreal forest,
Geophys. Res. Lett., 24, 1219 – 1222, 1997.
Mäkelä, J. M., T. Hoffmann, C. Holzke, M. Väkevä, T. Suni, T. Mattila,
P. P. Aalto, U. Tapper, E. I. Kauppinen, and C. D. O’Dowd, Biogenic
iodine emissions and identification of end-products in coastal ultrafine
particles during nucleation bursts, J. Geophys. Res., 107, doi:10.1029/
2001JD000580, in press, 2002.
Marti, J. J., A. Jeffersson, X. P. Cai, C. Richert, P. H. McMurry, and
F. Eisele, H2SO4 vapor pressure of sulfuric acid and ammonia sulfate
solutions, J. Geophys. Res., 102, 2725 – 2735, 1997.
Nilsson, E. D., and M. Kulmala, The potential for atmospheric mixing
processes to enhance the binary nucleation rate, J. Geophys. Res., 103,
1381 – 1389, 1998.
O’Dowd, C. D., New Particle Formation and Fate in the Coastal Environment (PARFORCE) objectives and initial achievements, Rep. Ser. Aerosol Sci., 48, 5 – 10, 2000.
O’Dowd, C. D., B. Davison, J. A. Lowe, M. H. Smith, R. M. Harrison, and
C. N. Hewitt, Biogenic sulphur emissions and inferred sulphate CCN
concentrations in and around Antarctica, J. Geophys. Res., 102,
12,839 – 12,854, 1997.
O’Dowd, C. D., M. Geever, M. K. Hill, S. G. Jennings, and M. H. Smith,
New particle formation: Spatial scales and nucleation rates in the coastal
environment, Geophys. Res. Lett., 25, 1661 – 1664, 1998.
O’Dowd, C. D., et al., On the photochemical production of new particles in
the coastal boundary layer, Geophys. Res. Lett., 26, 1707 – 1710, 1999.
O’Dowd, C. D., et al., A dedicated study of New Particle Formation and
Fate in the Coastal Environment (PARFORCE): Overview of objectives
PAR
15 - 11
and achievements, J. Geophys. Res., 107, doi:10.1029/2001JD000555, in
press, 2002a.
O’Dowd, C. D., et al., Coastal new particle formation: Environmental
conditions and aerosol physicochemical characteristics during nucleation
bursts, J. Geophys. Res., 107, doi:10.1029/2000JD000206, in press,
2002b.
Pirjola, L., C. D. O’Dowd, and M. Kulmala, A model prediction of the yield
of cloud condensation nuclei from coastal nucleation events, J. Geophys.
Res., 107, 8098, doi:10.1029/2000JD000213, 2002.
Scott, W. D., and F. C. R. Cattell, Vapor pressure of ammonium sulfates,
Atmos. Env., 13, 307 – 317, 1979.
Stelson, A. W., M. E. Bassett, and J. H. Seinfeld, Thermodynamic equilibrium properties of aqueous solutions of nitrate, sulfate and ammonium,
in Chemistry of Particles, Fogs and Rain, edited by J. L. Durham, pp. 1 –
52, Butterworth-Heinemann, Woburn, Mass., 1983.
Tang, I. N., and H. R. Munkelwitz, Water activities, densities, and refractive
indices of aqueous sulfates and sodium nitrate droplets of atmospheric
importance, J. Geophys. Res., 99, 18,801 – 18,808, 1994.
Väkevä, M., K. Hämeri, and P. P. Aalto, Hygroscopic properties of nucleation mode and Aitken mode particles during and outside nucleation bursts
in west coast of Ireland, J. Geophys. Res., 107, 8104, doi:10.1029/
2000JD000176, 2002.
Weast, R. C., (Ed.), CRC Handbook of Chemistry and Physics, 74th ed.,
CRC Press, Boca Raton, Fla., 1994.
Weber, R. J., P. H. McMurry, R. L. Mauldin, D. J. Tanner, F. L. Eisele,
A. D. Clarke, and V. Kapustin, New particle formation in the remote
troposphere: A comparison of observations at various sites, Geophys.
Res. Lett., 26, 307 – 310, 1999.
Wiedensohler, A., D. S. Covert, E. Swietlicki, P. Aalto, J. Heintzenberg,
and C. Leck, Occurrence of an ultrafine particle mode less than 20 nm in
diameter in the marine boundary layer during Arctic summer and autumn,
Tellus, Ser. B, 48, 213 – 222, 1996.
H. Berresheim, German Weather Service, Meteorological Observatory,
Albin-Schwaiger-Weg 10, D-82383 Hohenpeissenberg, Germany. (harald.
[email protected])
C. D. O’Dowd, Department of Physics, National University of Ireland,
University Road, Galway, UK. ([email protected])
A. Karlsson, ITM Air Pollution Laboratory, Stockholm University, SE-106
91 Stockholm, Sweden. ([email protected])
P. Korhonen, M. Kulmala, and I. Napari, Laboratory of Aerosol and
Environmental Physics, Department of Physical Sciences, University of
Helsinki, P.O. Box 64, FIN-00014 Helsinki, Finland. (pekka.korhonen@
fmi.fi; [email protected]; [email protected])