Glass-forming properties of 3-methylbutane-1,2,3-tricarboxylic acid and its mixtures with water and pinonic acid

Article abstract of DOI:10.1021/jp505910w

3-Methylbutane-1,2,3-tricarboxylic acid (3-MBTCA) is an atmospheric oxidation product of α-pinene and has been identified as the most relevant tracer compound for atmospheric terpene secondary organic aerosol (SOA) particles. Little is known, however, of its physicochemical properties such as water solubility and phase state (e.g., liquid, crystalline, glassy). To gain knowledge, we synthesized 3-MBCTA from methyl 2-methylpropanoate and dimethyl maleate via a Michael addition and subsequent hydrolysis with 78% overall yield. It was found that 3-MBTCA transforms into anhydrides upon melting at T _m = 426 ± 1 K, thus preventing a determination of the glass transition temperature T_g by differential scanning calorimetry (DSC) through melting and subsequent cooling. Therefore, we designed the novel technique MARBLES (metastable aerosol by low temperature evaporation of solvent) for transferring a substance into a glassy state without heating. In MARBLES an aqueous solution is atomized into wet aerosol particles that are subsequently dried in several diffusion dryers resulting in glass formation of the residual particles for several solutes. The glassy aerosol particles are collected in an impactor until enough mass has accumulated that the sample's T_g can be determined by DSC. Using this method, the glass transition temperature of 3-MBTCA was found to be T_g ≈ 305 ± 2 K. Moreover, we have determined the glass transition T_g' of the maximal freeze-concentrated aqueous solution of 3-MBTCA, and T_g of mixtures of 3-MBTCA with water and pinonic acid. The latter data indicate a dependence of T_g upon the atomic oxygen-to-carbon ratio of the mixture, with implications for parametrizing the glass-forming behavior of α-pinene SOA particles in the atmosphere.

Full text of DOI:10.1021/jp505910w

Article
pubs.acs.org/JPCA
Glass-Forming Properties of 3‑Methylbutane-1,2,3-tricarboxylic Acid
and Its Mixtures with Water and Pinonic Acid
Hans P. Dette,†,‡ Mian Qi,†,‡ David C. Schroder,† Adelheid Godt,†,‡ and Thomas Koop*,†,‡
̈
‡
†Faculty of Chemistry and Center for Molecular Materials, Bielefeld University, Universitatsstraße 25, D-33615 Bielefeld, Germany
̈
S
* Supporting Information
ABSTRACT: 3-Methylbutane-1,2,3-tricarboxylic acid (3-
MBTCA) is an atmospheric oxidation product of α-pinene
and has been identified as the most relevant tracer compound
for atmospheric terpene secondary organic aerosol (SOA)
particles. Little is known, however, of its physicochemical
properties such as water solubility and phase state (e.g., liquid,
crystalline, glassy). To gain knowledge, we synthesized 3-
MBCTA from methyl 2-methylpropanoate and dimethyl
maleate via a Michael addition and subsequent hydrolysis
with 78% overall yield. It was found that 3-MBTCA transforms
into anhydrides upon melting at Tm = 426 1 K, thus preventing a determination of the glass transition temperature Tg by
differential scanning calorimetry (DSC) through melting and subsequent cooling. Therefore, we designed the novel technique
MARBLES (metastable aerosol by low temperature evaporation of solvent) for transferring a substance into a glassy state without
heating. In MARBLES an aqueous solution is atomized into wet aerosol particles that are subsequently dried in several diffusion
dryers resulting in glass formation of the residual particles for several solutes. The glassy aerosol particles are collected in an
impactor until enough mass has accumulated that the sample’s Tg can be determined by DSC. Using this method, the glass
transition temperature of 3-MBTCA was found to be Tg ≈ 305 2 K. Moreover, we have determined the glass transition Tg′ of
the maximal freeze-concentrated aqueous solution of 3-MBTCA, and Tg of mixtures of 3-MBTCA with water and pinonic acid.
The latter data indicate a dependence of Tg upon the atomic oxygen-to-carbon ratio of the mixture, with implications for
parametrizing the glass-forming behavior of α-pinene SOA particles in the atmosphere.
ketones, aldehydes, different mono- and disaccharides, and
others.10,16−18 In contrast to most inorganic aerosol particles,
which crystallize readily upon drying,19 the organic aerosol
particles rather tend to form liquid particles instead of crystals
due to being a mixture of compounds.20
1. INTRODUCTION
Aerosols influence various processes in Earth’s atmosphere. For
example, they directly affect Earth’s albedo through scattering
and absorption of incoming sunlight, which in turn depend
upon the aerosol particles’ properties such as phase state, size,
chemical composition, and absorption cross section.1 Aerosols
further influence the atmosphere’s radiative transfer indirectly
by serving as cloud condensation nuclei (CCN) or ice nuclei
(IN), partly depending on the pre-existing aerosol particles’
phase state.2−4 For example, liquid particles of water-soluble
compounds may act as CCN for liquid water clouds and solid
particles, both crystalline and amorphous ones, may act as IN
for ice clouds5−8 The magnitude of their influence is little
understood,9 in part because aerosol properties such as phase
state are often variable and not well-known. This is particularly
true for organic aerosols that make up for about half of the
aerosol mass in the atmosphere.10−13
Most organic molecules are released into the atmosphere as
volatile organic compounds from which secondary organic
aerosol (SOA) particles may arise through atmospheric
oxidation and subsequent condensation.10,13−15 The organic
compounds found in organic aerosols often show moderate to
good water solubility due to their high oxidation state. For
example, atmospheric organic aerosols may consist of
compounds such as mono-, di-, or tricaboxylic acids, polyols,
SOA particles arising from the oxidation of biogenic
precursors and subsequent condensation consist of hundreds
of different organic compounds, but a number of key
components have been identified. For example, in boreal
forests α-pinene is one of the most important biogenic SOA
precursors.10,14,21 Experiments have shown that the gas-phase
oxidation of α-pinene proceeds via pinonic acid, a semivolatile
compound, to the second-generation oxidation product 3-
methylbutane-1,2,3-tricarboxylic acid (3-MBTCA), a low-
volatile compound found predominantly in the aerosol phase,
Figure 1.13,14,22−27 Not much is known about the phys-
icochemical properties and the phase state of pinonic acid and
3-MBCTA, neither as pure components nor in mixtures. This
knowledge is, however, particularly important, because it has
been suggested that SOA particles in boreal forests may exist in
an amorphous solid, i.e., glassy state at room temperature under
Received: June 13, 2014
Revised: August 6, 2014
Published: August 6, 2014
© 2014 American Chemical Society
7024
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
changes in heat-unstable samples. These potential risks apply to
heat drying of bulk samples as well as to spray drying, in which
the sample solution is first sprayed into an aerosol before being
heated by mixing with a hot gas.43,44
To circumvent these problems, we have developed the
MARBLES method (metastable aerosol by low temperature
evaporation of solvent), which is described herein. MARBLES
is based on diffusion drying of an aerosol at room temperature
and has three principal advantages: relatively fast drying on the
order of seconds to minutes, a reduced likelihood of nucleation
of crystalline phases in the small, typically micrometer- to
submicrometer-sized liquid aerosol droplets, and avoidance of
chemical changes of organic substances by operating at room
temperature. Below, we will show the functionality of the
method using different substances with known Tg before we
apply it to 3-MBTCA and its mixtures with water and pinonic
acid.
The paper is structured as follows: Section 2 presents the
synthesis and characterization of 3-MBTCA. Section 3
describes experimental details of the DSC measurements and
the new experimental method MARBLES as well as several
validation experiments with reference substances. In section 4,
the results of glass transition measurements of pure 3-MBTCA
and its mixtures with water and pinonic acid are presented as
well as a discussion of atmospheric implications, before we
finish with a conclusion and a brief outlook in section 5.
Figure 1. Structures and properties of marker compounds during the
atmospheric oxidation of α-pinene.
dry conditions.28,29 A semisolid or solid state of organic aerosol
particles has important atmospheric implications such as
delayed gas-to-aerosol partitioning.27,30−33 Moreover, such
particles may also have a longer chemical lifetime as the
oxidation can take place only on the surface of glassy particles
because of inhibited diffusion in the bulk of a particle.31,34−38
The glass-forming properties of key compounds of oxygenated
organic aerosol such as 3-MBTCA have not been investigated
experimentally until now.29 Because of its atmospheric
relevance, we synthesized 3-MBTCA and conducted a
physicochemical characterization of 3-MBTCA in the pure
state and in mixtures with water and pinonic acid.
For the estimation of whether a SOA compound such as 3-
MBTCA may form a glass in the atmosphere, it is essential to
determine its glass transition temperature Tg. The most
common approach for the determination of Tg is to transfer
a sample into a glass by first melting it by heating and
subsequently cooling it to temperatures sufficiently below the
estimated Tg, which typically is about 0.7 times its melting
temperature.29 Then the sample is reheated at a constant
heating rate and the glass transition is detected by means of
differential scanning calorimetry (DSC) through the associated
change in heat capacity.39,40 The approach described here
works for many substances, but it requires two important
preconditions to be fulfilled: first, the substance must remain
chemically intact upon melting and, second, the substance must
not recrystallize upon cooling but instead remain in a
supercooled metastable liquid state until the glass transition is
reached. However, in particular, sugars or di- and tricarboxylic
acids including 3-MBTCA undergo chemical changes upon
melting. Hence alternative methods for glass formation must be
chosen, for example, by drying a solution of the substance of
interest.41 The most common approaches are freeze-drying or
simple heating. In freeze-drying it is difficult to ensure complete
dehydration of a glassy sample, because of the rather small
water diffusion coefficient in glassy samples at low temper-
ature42 and, thus, freeze-drying a sample may require a
significant amount of time. Drying by heating may enhance
the likelihood of crystallization and bears the risk of chemical
2. SYNTHESIS AND CHARACTERIZATION OF
3-MBTCA
To obtain 3-MBTCA, we first pursued the route outlined by
Szmigielski,25 i.e., the deprotonation of dimethyl succinate,
substitution reaction between the resulting enolate and methyl
2-bromo-2-methylpropanoate, and finally ester hydrolysis.
However, in our hands the substitution reaction gave a very
low conversion of about 26% and side products. Therefore, we
developed an alternative method (Scheme 1): first deprotona-
tion of methyl 2-methylpropanoate (1), second Michael
addition of the resulting enolate to dimethyl maleate (2)45
resulting in the trimethyl ester (3) of 3-MBTCA, and third
Scheme 1. Synthesis of 3-MBTCA
7025
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
ester hydrolysis yielding the desired 3-MBTCA. This simple
three-step synthesis provided pure 3-MBTCA in 78% yield
from very cheap starting compounds. Details of the preparation
are given in the Supporting Information.
of solid impurities such as dust in most aerosol droplets that
might trigger crystal nucleation. And third, the entire procedure
(dissolution and drying) can be performed at room temper-
ature, thus eliminating the occurrence of chemical changes
which are often found when bulk samples are melted (see
below).
The melting point of 3-MBTCA was determined with DSC
measurements (see below) by heating a few crystals (about 5.1
mg) at a heating rate of 10 K min−1 to give Tm = 426 1 K, in
good agreement with literature data of Tm = 427−428 K.46
Furthermore, we determined the solubility of 3-MBTCA in
water at 293.15 K by drying about 0.3 or 1 g of a saturated
solution and weighing the resulting residue, yielding a solubility
value of about 320 g of 3-MBTCA per liter of water,
corresponding to a solute mass fraction of wsol ≈ 0.24
0.02. The water activity and density of an aqueous solution with
a mass fraction of w = 0.2 were determined to be aw = 0.982
and ρ = 1.05(8) g cm−3 using a water activity meter (Decagon
model series 3 TE) and a density meter (Anton Paar, DMA
4500 M).
Figure 2a outlines the working principle of the MARBLES
method. A dilute aqueous solution of the substance of interest
3. THERMAL CHARACTERIZATION TECHNIQUE AND
SAMPLE PREPARATION WITH MARBLES
3.1. Differential Scanning Calorimetry. Differential
scanning calorimetry (DSC) is a standard technique for
studying phase transition processes such as crystallization,
melting, and glass formation. For the experiments described
below we used a differential scanning heat-flow calorimeter
(Q100, TA Instruments) that was calibrated comprehensively
using a series of nine calibration standards at cooling and
heating rates between −10 and +10 K min−1.47 The resulting
temperature accuracy of the calorimeter is about 0.2 at 1 and
0.4 K at 10 K min−1. In a typical measurement, 1−10 mg of a
sample was placed in an aluminum pan (thereafter sealed
hermetically), which was then transferred into the calorimeter.
Measurements were performed with an empty pan as a
reference. The glass transition temperature Tg of a sample was
determined in the heating mode at 10 K min−1 using the onset
of the glass transition signal, i.e., the intersection between the
extrapolated baseline and a line of maximum slope in the heat
flow signal, following the convention of Angell.39 We note that
because of small ambiguities in the evaluation of these slopes
and the repeatability of the experimentally determined Tg
values across multiple samples, we estimate the accuracy of
the glass transition temperatures given below to be about 2 K.
3.2. The MARBLES Method. We have developed and
tested a novel method for preparing glassy samples of water-
soluble substances which we term MARBLES for “metastable
aerosol by low temperature evaporation of solvent”. With the
MARBLES method an aqueous solution of a compound is
sprayed into an aerosol followed by drying of the aerosol
droplets and collection of the dry particles. This is similar to
classical spray drying, but MARBLES avoids heating of the
sample and requires only a small amount of material for
preparing samples that can be analyzed by DSC. MARBLES has
several key advantages. First, drying of aqueous bulk samples or
droplets with a diameter of several micrometers or more can be
time-consuming, because of the small water diffusivity in glassy
materials.42,48 However, under the same conditions drying of
submicrometer aerosol particles (see images of collected
particles in Figure S3 in the Supporting Information) can be
achieved on the time scale of seconds to minutes.29,34,49,50 This
is made use of in MARBLES. Second, the crystallization
probability of the solute compound is strongly reduced in
aerosol particles, because of the smaller volume and the absence
Figure 2. (a) Working principle of the MARBLES system: wet aerosol
generation from an aqueous solution in an atomizer, aerosol drying in
diffusion dryer, and dry aerosol collection in an impactor. (b) Detailed
sketch of the MARBLES system. A nitrogen source (NS) is used to
drive an atomizer (A) for aerosol generation, the drying section
consists of several diffusion dryers (DD), a hygrometer for relative
humidity measurement (RH) with a sensor (S1) in the aerosol flow,
and a sensor (S2) in the dry glovebox (GB). Dried particles are
collected in an impactor (IM) and residual aerosol particles are filtered
out (F). The overall aerosol flow rate is measured with a flow meter
(FM). The exhaust line (EL) is protected from room air humidity by
another diffusion dryer. Samples are hermetically sealed in a sealing
unit (SU), before they are transferred from the glovebox via a pass-
through transfer chamber (TC) to the DSC (not shown).
is atomized into a wet aerosol by an aerosol generator. The
aerosol is dried by diffusion drying before the aerosol particles
are collected in an impactor. A detailed functional scheme of
the MARBLES setup is shown in Figure 2b. Aerosol particles
are produced with an atomizer (A) by nitrogen purge gas from
a pressurized nitrogen source (NS), following the general
working principle of the constant output atomizer design of Liu
and Lee.51 The aerosol flow is directed through a drying section
consisting of three homemade diffusion dryers (DD), each 80
cm long. Each diffusion dryer consists of an outer tube made
7026
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
from poly(methyl methacrylate) with an inner diameter of 8.9
cm and an inner tube made of stainless steel wire screen about
2.5 cm in diameter. The annular space between the inner and
the outer tube is filled with silica gel drying pearls that are too
big to penetrate the wire screen (Silica Gel Orange, Roth,
diameter 2−5 mm, specific surface area ∼750 m2 g−1, water
absorbance ∼42 wt % at 80% RH). The aerosol stream is
directed through the inner wire-screen tube. While passing
through the drying section, water evaporates from the aerosol
droplets and diffuses through the wire screen to the annular
space where it is absorbed by the silica gel. The dried aerosol
then passes the first of two sensors (S1) of a hygrometer (RH;
Ahlborn Almeo 2590), which determines the relative humidity
of the gas phase of the aerosol. Thereafter, the aerosol particles
are collected using a homemade impactor (IM) by accelerating
the aerosol stream through a conical glass nozzle (orifice
diameter: 1 mm) and directing it onto an impaction plate. In
our setup, the impaction plate consists of the bottom piece of a
DSC aluminum pan (TA Instruments, hermetic aluminum pan
900793-901), which can be used directly for the subsequent
DSC measurements, thus avoiding further sample treatment.
Those aerosol particles that pass the impactor without being
collected are captured in a HEPA (high efficiency particulate
airfilter) filter (F) downstream of the impactor. The filtered gas
is directed through a flow meter (FM; Yokogawa Rotameter,
model RGC2) for determining the gas volume flow rate
through the entire setup. All components are connected either
by stainless steel Swagelock fittings and tubing or by plastic
tubing. The components RH, IM, F, and FM are placed inside a
dry glovebox (GB) about 310 L in volume (dimensions W × D
× H: 95 × 50 × 65 cm). During the experiments the relative
humidity inside the glovebox is continuously measured with a
second sensor (S2) of the hygrometer. Subsequent to sample
collection, the aluminum pan is hermetically sealed using a TA
Instruments sample encapsulation crimper press (sealing unit,
SU) inside the glovebox. Thereafter, the sample pan can be
handled in laboratory air for transferring it into the differential
scanning calorimeter for further analysis, without running the
risk of sample contamination by air.
3.3. Setup Optimization and Test Measurements. To
evaluate the working principle and the effectiveness of the
MARBLES method, we performed experiments with four
reference substances. After numerous tests it proved most
practicable to work with a solute mass fraction of w = 0.02 for
the aqueous solutions feeding the atomizer and an impactor
nozzle orifice diameter of 1 mm.
The aerosol drying efficiency depends on the residence time
of a particular aerosol particle in the drying section and,
therefore, on both the total length of the drying section and the
aerosol flow rate. Because the volume flow rate is fixed at
∼1900 sccm due to the operation requirements of the aerosol
generator, optimization was achieved by varying the length of
the drying section. The relative humidity (RH) of the aerosol
gas phase was measured after the drying section and was used
as one indicator for a completed aerosol drying process. When
a short drying section consisting of a single diffusion dryer tube
only is used, the outflow revealed measurable RH values,
indicating insufficient drying. However, with a drying section
consisting of three diffusion dryer tubes, the monitored RH was
always 0% 0.1% at a flow rate of ∼1900 sccm. This length-to-
flow-rate ratio corresponds to a residence time of an aerosol
particle in the drying section of about 37 s. An additional
increase of the drying section length did not lead to higher
values of Tg in the collected and analyzed samples, which
indirectly confirms that the chosen length is sufficient to result
in the lowest possible water content, see detailed discussion
below.
For the DSC analysis the aluminum pan with the collected
aerosol sample was removed from the impactor, sealed
hermetically, and then transferred to the DSC. To avoid
water uptake by the collected sample during the sealing of the
DSC pans with the crimper press, this sealing was done inside a
dry glovebox in which the RH was always held below 1%. The
effect of water uptake from ambient humidity during sealing on
the measured Tg is demonstrated with Tg data of trehalose
obtained from the DSC measurements for five different sample
preparation cases (Table 1): Sealing the sample inside the dry
Table 1. Tg of Trehalose and 3-MBTCA Samples after
Drying and Exposure to Different Humidity Conditions
(Dry Glovebox and Ambient Humidity) for Varying Time
a
Periods before Sealing
sealing conditions
time before sealing [s]
Tg [K] ( 2 K)
Trehalose
dry glovebox
dry glovebox
ambient
0
369
368
362
348
340
60
0
ambient
30
ambient
60
3-MBTCA
dry glovebox
ambient
0
0
305
303
290
284
281
ambient
15
30
45
ambient
ambient
a
Because ambient humidity varies, all experiments with one substance
were done on the same day. A “time before sealing” of 0 s indicates
immediate sealing, but we note that the sealing itself takes a few
seconds.
glovebox where no significant water uptake may either occur
immediately (0 s) or occur after waiting for 60 s; immediately
sealing the sample in ambient air (0 s) or sealing the sample
after exposure to ambient air for 30 s and for 60 s. The glass
transition temperature Tg decreases strongly with increasing
waiting time before sealing in ambient air and, thus, with
increasing exposure time of the sample to the ambient
atmosphere. Even if the sample is sealed in ambient air without
delay (0 s), the sealing itself takes about 3−5 s and the
measured Tg is lower by about 7 K than that of a sample that is
sealed inside the dry glovebox. Obviously the 3−5 s needed for
sealing is sufficient to adulterate the measured Tg value. Sealing
the sample inside the glovebox after a 60 s waiting time did not
significantly affect the glass transition temperature, proving that
no water is taken up by the sample inside the glovebox. Similar
results were obtained with 3-MBTCA (Table 1).
For validating the drying effectiveness of the MARBLES
method we determined the glass transition temperatures of
glucose, sucrose, trehalose, and citric acid. The saccharides were
chosen because they are well-studied glass-forming substances,
and citric acid was used because it is structurally similar to 3-
MBTCA and, thus, it may act as a surrogate in test experiments.
In Table 2 we show the Tg values of samples vitrified with the
MARBLES setup. All substances were collected under dry
conditions for 20 min (resulting in a total mass of up to 5 mg)
7027
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
been transferred into the glassy state by the MARBLES method
was cooled from 293 to 263 K (blue), heated from 263 to 463
K (green), cooled again to 263 K (magenta), and finally heated
to 463 K (red). In the first heating cycle, the transition from a
glass to a supercooled liquid occurred as a step in the heat flow
signal (1) at 331 K. On further heating, the supercooled liquid
first crystallized (2) and then melted (3). In the second cooling
and heating cycle again a glass transition was detected (4a and
4b), but now at a lower temperature than in the first heating
cycle (309 K). We conclude from this observation that the
sucrose underwent chemical changes during the first melting,
and that the Tg determined in the second heating cycle has to
be assigned to another compound or a compound mixture. In
contrast, the glass transition temperature obtained from the
sample vitrified at room temperature by the MARBLES method
suggests that the sample indeed consists of pure glassy sucrose
and, hence, the measured value may represent the “real” Tg of
sucrose. In the results section below we show more examples
for this kind of behavior and also provide evidence that it is the
thermal treatment leading to chemical changes and, thus, a
reduced Tg value, and not the MARBLES method that leads to
an “artificially enhanced” Tg value.
Table 2. Glass Transition Temperatures of Reference
Compounds Determined with MARBLES and
Corresponding Literature Values
Tg
Vitrification procedure and
DSC heating rate
[K min−1
substance
sucrose
[K]
literature references
]
331 MARBLES
10
319 freeze-drying63
335 freeze-drying64
297 MARBLES
10
5
glucose
10
10
10
10
10
10
10
295 melting and cooling63
286 MARBLES
citric acid
284 melting and cooling65
283 melting and cooling66
369 MARBLES
368 freeze-drying63
365 freeze-drying67
373 melting and cooling64
trehalose
1, 10, and 20
5
385 heating of dihydrate at reduced
1, 10, and 20
pressure67
388 freeze-drying68
10
10
388 heating of dihydrate at ambient and
reduced pressure69
and then analyzed by DSC. Also shown for comparison are Tg
values of the same substances in which vitrification was
achieved by established alternative methods. Given the
variability of Tg values reported in the literature, our results
agree well with the literature values, providing further support
for the applicability of our method for determining Tg of water-
soluble organics.
4. THERMAL CHARACTERIZATION OF 3-MBTCA AND
MIXTURES THEREOF
4.1. Tg of 3-MBTCA. 3-MBTCA was dissolved in water (w
= 0.02) and transferred into the glassy state through the
MARBLES process as described above. The collected aerosol
sample was then investigated by DSC, Figure 4a. The sample
was first cooled from room temperature to 233 K at 10 K min−1
and then heated at 10 K min−1 to 463 K (green line), before the
cooling/heating cycle was repeated (red line). The glass
transition (1) was observed at Tg = 305 2 K, which is close to
the value predicted from Tm by the Boyer−Beaman relationship
Tg,pred ≈ 0.7Tm ≈ 299 K.29 Upon further heating, crystallization
(2) of the sample occurred at about 345 K before melting (3)
commenced at Tm = 426 K, in agreement with a literature value
of Tm = 427−428 K.46 In the second cooling cycle only a liquid-
to-glass transition was observed and in the second heating cycle
a glass transition (4) was detected at about 259 K, i.e.,
significantly below Tg = 305 K of the first cycle. The latter is a
clear indication of a chemical change upon melting in the first
cycle, and it is in agreement with the fact that no crystallization
occurred in the second heating cycle (gray line). In
independent experiments we found that upon melting of 3-
MBTCA a mixture of the anhydrides 4 and 5 is formed
(Scheme 2; see Supporting Informatin for product identi-
fication by NMR spectroscopy, mass spectrometry, and IR
spectroscopy). Hence, we assign the glass transition temper-
ature Tg = 259 K to a mixture of the two anhydrides, residual 3-
MBTCA, and water. When a crystalline bulk sample of 3-
MBTCA is melted, cooled, and reheated (blue line in Figure
4b), the same low glass transition temperature (4) is observed
(compare with the red line, which is reproduced from Figure
4a). This measurement confirms that it is not possible to
determine Tg of 3-MBTCA by classical means (heating and
cooling) and emphasizes the usefulness of the MARBLES
approach taken here. Note that citric acid, a compound that had
been used to test the MARBLES approach (Table 2), is
structurally similar to 3-MBTCA. Interestingly, citric acid is
chemically stable during melting. Chemical changes of citric
acid occurred at around 450 K, i.e., well above the melting point
A typical DSC measurement of a heat unstable substance is
shown in Figure 3. In this experiment a sucrose sample that had
Figure 3. DSC thermogram with two consecutive cooling/heating
cycles of the heat-unstable substance sucrose. The first cooling/heating
cycle is marked in blue/green, and the second cycle in magenta/red.
Numbers indicate the glass transition during the first heating cycle (1),
subsequent crystallization (2), and melting/decomposition (3). The
glass transition in the second cooling (4a) and heating cycle (4b) is
observed at a lower temperature (see inset).
7028
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
Therefore, instead of starting with an aqueous solution, we
prepared dry glassy samples of 3-MBTCA by the MARBLES
method. We then exposed these bulk samples (2−3 mg) to
ambient humidity such that water uptake could occur as
described above. The amount of water taken up by these bulk
samples was determined by weighing the sample before and
after the exposure to humidity. Only samples with water mass
fractions of 0.01 and 0.02 (corresponding to exposure times of
1 and 5 min, respectively) could be prepared because longer
exposure times did not lead to further water uptake but instead
to crystallization of 3-MBTCA in the sample pans. The Tg
values for 3-MBTCA mass fractions of w = 0.99 and w = 0.98
are 289 2 and 272 2 K, respectively. The change of Tg as a
function of the water content of a sample can be estimated
quite generally with the semiempirical Gordon−Taylor
equation:52
1
kGT
1
w1T +
w2T
g2
g1
T =
g
w1 +
w2
kGT
Here, Tg is the glass transition temperature of the binary
mixture, Tg1 and Tg2 are the glass transition temperatures of the
two pure compounds, here of water and of solute, w1 and w2 are
the mass fractions of the two compounds in the mixture, and
k
GT is the Gordon−Taylor constant. Because Tg of water is very
low (Tg1 = 136 K), even small amounts of water can
significantly decrease Tg of the mixture, as found for 3-
MBTCA (cf. Table 1). The strong decrease of the water/3-
MBTCA mixtures’ Tg with only small water content suggests a
relatively large Gordon−Taylor constant kGT, and fitting the
Gordon−Taylor equation to the available data points results in
kGT ≈ 9. We note, however, that this value can be regarded as a
rough estimate only due to the limited available data and the
associated uncertainty.
Another important characteristic temperature of aqueous
glassy materials is the glass transition of the maximally freeze-
concentrated aqueous solution, commonly termed Tg′.40,53,54
We determined Tg′ of 3-MBTCA by studying the phase
transitions observed in inverse emulsion samples as described
in previous studies.40 For these measurements an aqueous 3-
MBTCA solution (w = 0.2) was emulsified by mixing one part
of a 7 wt % solution of SPAN60 (Sigma) in a 1:1 mixture of
methylcyclopentane/methylcyclohexane (Merck) with one part
of the aqueous solution. This mixture was stirred with a
homogenizer (IKA Ultra-Turrax T25 basic) for 10 min at
20 000 rpm to yield an inverse emulsion containing aqueous
droplets in the micrometer size range. Analysis of such a sample
by DSC is shown in Figure 5. Upon cooling at a rate of 10 K
min−1 (red line) homogeneous ice nucleation occurred at Th =
229 K, in agreement with a value of 229.5 K predicted by water-
activity-based ice nucleation theory55,56 using the experimen-
tally determined water activity of the aqueous 3-MBTCA
solution of aw = 0.982 at room temperature and a value of
Δaw,hom = 0.325 for homogeneous ice nucleation in micro-
meter-sized droplets.57 Below Th no further transitions were
observed upon cooling to 153 K. Upon heating at 10 K min−1
(magenta line) a small exothermic signal was observed above a
temperature of 228 K (see also inset), most likely resulting
from ice crystal growth that was inhibited during cooling and at
lower temperature due to the high viscosity of the freeze-
concentrated aqueous 3-MBCTA matrix. At 246 K heating was
stopped and tempering of the sample commenced by slowly
Figure 4. DSC thermograms of 3-MBTCA samples. (a) 3-MBCTA
sample collected after the MARBLES process. The first cooling/
heating cycle is marked in green, and the second one is marked in red.
(b) Comparison of thermograms of a glassy and a crystalline bulk
sample of 3-MBCTA. The blue line marks the heating/cooling cycle of
the crystalline sample; the red one is reproduced from panel (a). For
details, see text.
Scheme 2. Melting of 3-MBTCA Forming the Anhydrides 4
and 5
1
of 426 K. IR and H NMR spectroscopic data indicate olefin
and anhydride formation.
4.2. Tg of Water/3-MBTCA Mixtures. The glass transition
temperatures of aqueous solutions of 3-MBTCA are not
accessible because of the relatively low water solubility of 3-
MBTCA with a solute mass fraction of wsol ≈ 0.24; see above.
7029
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
Figure 5. DSC thermogram of an emulsified aqueous solution (w =
0.2) of 3-MBTCA. The red line shows the first cooling, the magenta
line shows the heating above Tg′, the orange line shows the tempering
with a lower cooling rate, and the following cooling and heating are
shown by the blue and green lines. Th indicates homogeneous ice
nucleation, Tm the ice melting, and Tg′ the glass transition of the
maximally freeze concentrated sample. For details, see text.
Figure 6. Tg of pinonic acid/3-MBTCA mixtures (black circles) as a
function of the 3-MBTCA mass fraction. Also shown are Tg values
from the second heating cycle (after chemical changes, blue circles).
The red and green circles indicate the measured Tg of pure 3-MBTCA
and the proposed Tg of pinonic acid, respectively. The latter results
from an extrapolation of the black line obtained from a Gordon−
Taylor-fit to the data. The red and pink shadowed areas indicate the 1σ
and 2σ confidence intervals of the fit.
cooling the sample to 233 K at 1 K min−1 (orange line) before
it was cooled further to 153 K at 10 K min−1 (blue line).
Finally, the sample was reheated to 303 K at 10 K min−1 (green
line) This time a clear glass transition was observed at 230 K
(see also inset), which we assign to the glass transition of the
maximally freeze-concentrated solution with a value of Tg′ =
230 2 K.
is almost linear (kGT ≈ 1), thus allowing a relatively stable
extrapolation of the Tg curve into the non-glass-forming
concentration range at wdry(3-MBTCA) < 0.3. Moreover, the
Tg of pinonic acid might be estimated by extrapolating the
Gordon−Taylor line to wdry(3-MBTCA) = 0, yielding a Tg of
pinonic acid of about 221 K. We note, however, that the
Gordon−Taylor approach may result in curved lines for kGT
values that are significantly smaller or larger than 1. The pink
and red shaded areas indicate, respectively, the 1σ and 2σ
confidence intervals of the fitting procedure, representing the
very conservative estimates of Tg in the non-glass-forming
region.
A Tg of pinonic acid of about 221 K suggests a low or
moderate viscosity of supercooled liquid pinonic acid at 293 K,
which is too low to prevent crystallization of pinonic acid
during the MARBLES procedure, in keeping with experimental
observations; see above. Therefore, determining the Tg of
pinonic acid requires a modification of MARBLES to allow for
particle collection and/or drying at low temperatures.
Also shown in Figure 6 are the Tg values determined in a
second cooling and heating cycle. As noted above, 3-MBTCA is
converted into the anhydrides 4 and 5 upon melting. Therefore,
these Tg values represent the glass transition of mixtures of
anhydrides 4 and 5, water, 3-MBTCA, and pinonic acid. With
an increasing amount of pinonic acid, the data approach the Tg
proposed for pure pinonic acid. This finding is in agreement
with the assumption that pinonic acid is less prone to forming
an anhydride.
4.3. Tg of Pinonic Acid/3-MBTCA Mixtures. α-Pinene
SOA particles consist of different α-pinene oxidation
products.13,14,27 Therefore, we studied Tg of mixtures of the
intermediate oxidation product pinonic acid (O:C ratio of 0.3;
Figure 1) with the second-generation oxidation product 3-
MBTCA (O:C ratio of 0.75). For this purpose we dissolved
different mixtures of the two acids with varying mass ratios at
an overall mass fraction of about 0.02 in water and then applied
the MARBLES procedure. The collected aerosol samples were
then analyzed by DSC by first cooling the samples at 10 K
min−1 to 133 K and then reheating them at 10 K min−1 to 433
K. The resulting Tg values are shown as a function of 3-
MBTCA dry mass fraction in Figure 6. Clearly, the higher the
dry mass fraction of 3-MBTCA the higher is the Tg of the
mixture, finally approaching the Tg of pure 3-MBTCA for
wdry(3-MBTCA) = 1 in agreement with the Gordon−Taylor
equation.29,52 At lower dry mass fractions wdry(3-MBTCA) <
0.3, i.e., at higher pinonic acid mass fractions, no Tg could be
determined owing to crystallization of pinonic acid, in
agreement with results of experiments on pure pinonic acid.
To describe the Tg of pinonic acid/3-MBTCA mixtures, we
fitted the available Tg data to the Gordon−Taylor equation
described above, resulting in the black solid line. The fitted line
7030
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
Finally, in Figure 7 we show the Tg values of several key
products of α-pinene SOA oxidation as a function of their O:C
pinene. Several physical and chemical properties of 3-MBTCA
were determined such as its solubility in water at room
temperature, its melting temperature Tm, its glass transition
temperature Tg, and the glass transition temperature of the
maximally freeze-concentrated solution Tg′. Furthermore, we
have determined the glass-forming properties of 3-MBTCA in
mixtures with water and pinonic acid. In particular, the results
of pinonic acid/3-MBTCA mixtures suggest a dependence of
Tg upon the O:C ratio for a set of compounds that originate
from the same precursor along an oxidation pathway. Such a
relation may prove helpful for modeling the effects of phase
state on the properties and behavior of atmospheric SOA
particles.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details of the 3-MBTCA synthesis and of experiments on
anhydride formation and associated analytical data, including IR
and NMR spectra and TEM images of particles after drying
with MARBLES, are presented. This material is available free of
charge via the Internet at http://pubs.acs.org.
Figure 7. Tg of α-pinene oxidation products as a function of the
atomic O:C ratio of the compound or compound mixture. The red
data points are from Figure 6, the Tg values for α-/β-pinene and
pinonaldehyde were estimated from the Boyer−Beaman relationship
(Koop et al.29). The dashed black line is a fit to the red data points
only, and the solid black line includes all data points with the resulting
fit parameters given in the figure.
AUTHOR INFORMATION
Corresponding Author
*T. Koop. E-mail: thomas.koop@uni-bielefeld.de.
Author Contributions
The study was conceived by H.P.D., A.G., and T.K., the
synthesis was developed and conducted by H.P.D., D.C.S.,
M.Q., and A.G., the anhydrides were characterized by M.Q.,
D.C.S., and A.G., the setup was developed and built by H.P.D.
and T.K., the experiments were performed by H.P.D., the
results were analyzed by H.P.D. and T.K. and discussed by all
authors. The paper was written by H.P.D. and T.K. with
contributions from all authors. All authors have given approval
to the final version of the manuscript.
■
ratio, a property that is often used to characterize multi-
component SOA particles using aerosol mass spectrometry and
also during modeling of SOA oxidation.13,58,59 The exper-
imental data reveal a clear and almost linear increase of Tg with
increasing O:C ratio. This seems to be in contradiction with an
analysis in our previous work on the general behavior of Tg in
organic molecules: A compilation of data of many very different
organic molecules did not show a clear trend of Tg as a function
of O:C. Instead, the molar mass dependence of Tg turned out
to be much more important.29 This analysis was a rather
general one on a compilation of diverse compounds, however,
and an O:C dependence of Tg may hold for specific sets of
chemically related compounds only. One such example is the
oxidation products of α-pinene shown in Figure 7, starting from
the pinonaldehyde toward 3-MBTCA. Similar trends of Tg have
been suggested for oxidation products identified in SOA
derived from specific precursors such as n-heptadecane and
naphthalene.60 Hence, we suggest that an O:C dependence of
Tg may thus facilitate the prediction of the SOA particles’
viscosity and phase state and facilitate modeling their effects on
gas-to-particle partitioning and the heterogeneous chemistry of
SOA particles.61,62
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank Y. Hertle for taking the TEM images, S. Gunthe for
help with the operation of the atomizer and helpful comments
on the MARBLES setup, and U. Poschl for providing the
atomizer. H.P.D. acknowledges a Nachwuchsfonds stipend
from Bielefeld University.
■
̈
ABBREVIATIONS
■
CCN
DSC
IN
cloud condensation nuclei
differential scanning calorimetry
ice nuclei
MARBLES metastable aerosol by low temperature evaporation
of solvent
3-MBTCA 3-methylbutane-1,2,3-tricarboxylic acid
O:C ratio oxygen-to-carbon atom ratio of an organic
substance
5. SUMMARY AND CONCLUSION
In this study we have established the MARBLES method based
on aerosol diffusion drying as a simple and fast approach to
form glassy samples from aqueous solutions. This method is
particularly suited for substances that undergo chemical
changes upon melting or heating. With MARBLES an aerosol
collection time of about 20 min was found to be sufficient for
obtaining about 2−5 mg of sample material as is required for a
DSC measurement of the glass transition temperature.
RH
relative humidity
standard cubic centimeter
secondary organic aerosol
sccm
SOA
REFERENCES
■
(1) Haywood, J.; Boucher, O. Estimates of the Direct and Indirect
Radiative Forcing due to Tropospheric Aerosols: A Review. Rev.
Geophys. 2000, 38, 513.
We were able to accomplish an efficient synthesis for 3-
MBTCA, which is a second-generation oxidation product of α-
7031
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
(2) Albrecht, B. A. Aerosols, Cloud Microphysics, and Fractional
Cloudiness. Science 1989, 245, 1227−1230.
(20) Marcolli, C.; Luo, B.; Peter, T. Mixing of the Organic Aerosol
Fractions: Liquids as the Thermodynamically Stable Phases. J. Phys.
Chem. A 2004, 108, 2216−2224.
(21) Guenther, A.; Hewitt, C. N.; Erickson, D.; Fall, R.; Geron, C.;
Graedel, T.; Harley, P.; Klinger, L.; Lerdau, M.; Mckay, W. A.; et al. A
Global Model of Natural Volatile Organic Compound Emissions. J.
Geophys. Res. 1995, 100, 8873.
(3) Lohmann, U.; Feichter, J. Global Indirect Aerosol Effects: A
Review. Atmos. Chem. Phys. 2005, 5, 715−737.
(4) Koop, T.; Mahowald, N. Atmospheric Science: The Seeds of Ice
in Clouds. Nature 2013, 498, 302−303.
(5) Murray, B. J.; Wilson, T. W.; Dobbie, S.; Cui, Z.; Al-Jumur, S. M.
(22) Muller, L.; Reinnig, M.-C.; Naumann, K. H.; Saathoff, H.;
̈
R. K.; Mohler, O.; Schnaiter, M.; Wagner, R.; Benz, S.; Niemand, M.;
̈
Mentel, T. F.; Donahue, N. M.; Hoffmann, T. Formation of 3-Methyl-
1,2,3-Butanetricarboxylic Acid via Gas Phase Oxidation of Pinonic
Acid − a Mass Spectrometric Study of SOA Aging. Atmos. Chem. Phys.
2012, 12, 1483−1496.
et al. Heterogeneous Nucleation of Ice Particles on Glassy Aerosols
under Cirrus Conditions. Nat. Geosci. 2010, 3, 233−237.
(6) Andreae, M. O.; Rosenfeld, D. Aerosol−cloud−precipitation
Interactions. Part 1. The Nature and Sources of Cloud-Active
Aerosols. Earth-Sci. Rev. 2008, 89, 13−41.
(23) Zhang, Y. Y.; Muller, L.; Winterhalter, R.; Moortgat, G. K.;
̈
Hoffmann, T.; Poschl, U. Seasonal Cycle and Temperature Depend-
̈
(7) Hoose, C.; Mohler, O. Heterogeneous Ice Nucleation on
̈
ence of Pinene Oxidation Products, Dicarboxylic Acids and Nitro-
phenols in Fine and Coarse Air Particulate Matter. Atmos. Chem. Phys.
2010, 10, 7859−7873.
Atmospheric Aerosols: A Review of Results from Laboratory
Experiments. Atmos. Chem. Phys. 2012, 12, 9817−9854.
(8) Abbatt, J. P. D.; Benz, S.; Cziczo, D. J.; Kanji, Z.; Lohmann, U.;
́ ́
(24) Yasmeen, F.; Szmigielski, R.; Vermeylen, R.; Gomez-Gonzalez,
Mohler, O. Solid Ammonium Sulfate Aerosols as Ice Nuclei: A
̈
Y.; Surratt, J. D.; Chan, A. W. H.; Seinfeld, J. H.; Maenhaut, W.;
Claeys, M. Mass Spectrometric Characterization of Isomeric
Terpenoic Acids from the Oxidation of A-Pinene, B-Pinene, D-
Limonene, and Δ3-Carene in Fine Forest Aerosol. J. Mass Spectrom.
2011, 46, 425−442.
Pathway for Cirrus Cloud Formation. Science 2006, 313, 1770−1773.
(9) Stocker, T. K.; Qin, D.; Plattner, G.-K.; Tignor, M.; Allen, S. K.;
Boschung, J.; Nauels, A.; Xia, Y.; Bex, V.; Midgley, P. M. IPCC, 2013:
Climate Change 2013: The Physical Science Basis. Contribution of
Working Group I to the Fifth Assessment Report of the Inter-
governmental Panel on Climate Change; Cambridge University Press:
Cambridge, United Kingdom and New York, NY, USA, 2013; p 1535.
(10) Kanakidou, M.; Seinfeld, J. H.; Pandis, S. N.; Barnes, I.;
Dentener, F. J.; Facchini, M. C.; Van Dingenen, R.; Ervens, B.; Nenes,
a.; Nielsen, C. J.; et al. Organic Aerosol and Global Climate Modelling:
A Review. Atmos. Chem. Phys. 2005, 5, 1053−1123.
́ ́
(25) Szmigielski, R.; Surratt, J. D.; Gomez-Gonzalez, Y.; Van der
Veken, P.; Kourtchev, I.; Vermeylen, R.; Blockhuys, F.; Jaoui, M.;
Kleindienst, T. E.; Lewandowski, M.; et al. 3-Methyl-1,2,3-Butane-
tricarboxylic Acid: An Atmospheric Tracer for Terpene Secondary
Organic Aerosol. Geophys. Res. Lett. 2007, 34, L24811.
(26) Eddingsaas, N. C.; Loza, C. L.; Yee, L. D.; Chan, M.; Schilling,
K. A.; Chhabra, P. S.; Seinfeld, J. H.; Wennberg, P. O. A-Pinene
Photooxidation under Controlled Chemical Conditions − Part 2: SOA
Yield and Composition in Low- and High-NOx Environments. Atmos.
Chem. Phys. 2012, 12, 7413−7427.
(11) Murphy, D. M.; Cziczo, D. J.; Froyd, K. D.; Hudson, P. K.;
Matthew, B. M.; Middlebrook, A. M.; Peltier, R. E.; Sullivan, A.;
Thomson, D. S.; Weber, R. J. Single-Particle Mass Spectrometry of
Tropospheric Aerosol Particles. J. Geophys. Res. 2006, 111, D23S32.
(12) Zhang, Q.; Jimenez, J. L.; Canagaratna, M. R.; Allan, J. D.; Coe,
H.; Ulbrich, I.; Alfarra, M. R.; Takami, A.; Middlebrook, a. M.; Sun, Y.
L.; et al. Ubiquity and Dominance of Oxygenated Species in Organic
Aerosols in Anthropogenically-Influenced Northern Hemisphere
Midlatitudes. Geophys. Res. Lett. 2007, 34, L13801.
(13) Jimenez, J. L.; Canagaratna, M. R.; Donahue, N. M.; Prevot, A S.
H.; Zhang, Q.; Kroll, J. H.; DeCarlo, P. F.; Allan, J. D.; Coe, H.; Ng, N.
L.; et al. Evolution of Organic Aerosols in the Atmosphere. Science
2009, 326, 1525−1529.
(14) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.;
Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.;
Goldstein, A. H.; et al. The Formation, Properties and Impact of
Secondary Organic Aerosol: Current and Emerging Issues. Atmos.
Chem. Phys. 2009, 9, 5155−5236.
(15) Kroll, J. H.; Seinfeld, J. H. Chemistry of Secondary Organic
Aerosol: Formation and Evolution of Low-Volatility Organics in the
Atmosphere. Atmos. Environ. 2008, 42, 3593−3624.
(16) Decesari, S.; Fuzzi, S.; Facchini, M. C.; Mircea, M.; Emblico, L.;
Cavalli, F.; Maenhaut, W.; Chi, X.; Schkolnik, G.; Falkovich, A.; et al.
Characterization of the Organic Composition of Aerosols from
(27) Donahue, N. M.; Henry, K. M.; Mentel, T. F.; Kiendler-Scharr,
A.; Spindler, C.; Bohn, B.; Brauers, T.; Dorn, H. P.; Fuchs, H.;
Tillmann, R.; et al. Aging of Biogenic Secondary Organic Aerosol via
Gas-Phase OH Radical Reactions. Proc. Natl. Acad. Sci. U. S. A. 2012,
109, 13503−13508.
(28) Virtanen, A.; Joutsensaari, J.; Koop, T.; Kannosto, J.; Yli-Pirila,
̈
P.; Leskinen, J.; Makela, J. M.; Holopainen, J. K.; Poschl, U.; Kulmala,
̈
̈
̈
M.; et al. An Amorphous Solid State of Biogenic Secondary Organic
Aerosol Particles. Nature 2010, 467, 824−827.
(29) Koop, T.; Bookhold, J.; Shiraiwa, M.; Poschl, U. Glass
̈
Transition and Phase State of Organic Compounds: Dependency on
Molecular Properties and Implications for Secondary Organic Aerosols
in the Atmosphere. Phys. Chem. Chem. Phys. 2011, 13, 19238−19255.
́
(30) Zelenyuk, A.; Imre, D.; Beranek, J.; Abramson, E.; Wilson, J.;
Shrivastava, M. Synergy between Secondary Organic Aerosols and
Long-Range Transport of Polycyclic Aromatic Hydrocarbons. Environ.
Sci. Technol. 2012, 46, 12459−12466.
(31) Perraud, V.; Bruns, E. A.; Ezell, M. J.; Johnson, S. N.; Yu, Y.;
Alexander, M. L.; Zelenyuk, A.; Imre, D.; Chang, W. L.; Dabdub, D.;
et al. Nonequilibrium Atmospheric Secondary Organic Aerosol
Formation and Growth. Proc. Natl. Acad. Sci. U. S. A. 2012, 109,
2836−2841.
̂
Rondonia, Brazil, during the LBA-SMOCC 2002 Experiment and Its
Representation through Model Compounds. Atmos. Chem. Phys. 2006,
6, 375−402.
(32) Shiraiwa, M.; Seinfeld, J. H. Equilibration Timescale of
Atmospheric Secondary Organic Aerosol Partitioning. Geophys. Res.
Lett. 2012, 39, L24801.
(17) Rogge, W. F.; Mazurek, M. A.; Hildemann, L. M.; Cass, G. R.;
Simoneit, B. R. T. Quantification of Urban Organic Aerosols at a
Molecular Level: Identification, Abundance and Seasonal Variation.
Atmos. Environ. Part A. Gen. Top. 1993, 27, 1309−1330.
(18) Simoneit, B. R. T.; Elias, V. O.; Kobayashi, M.; Kawamura, K.;
Rushdi, A. I.; Medeiros, P. M.; Rogge, W. F.; Didyk, B. M.
SugarsDominant Water-Soluble Organic Compounds in Soils and
Characterization as Tracers in Atmospheric Particulate Matter.
Environ. Sci. Technol. 2004, 38, 5939−5949.
́
(33) Vaden, T. D.; Imre, D.; Beranek, J.; Shrivastava, M.; Zelenyuk,
A. Evaporation Kinetics and Phase of Laboratory and Ambient
Secondary Organic Aerosol. Proc. Natl. Acad. Sci. U. S. A. 2011, 108,
2190−2195.
(34) Shiraiwa, M.; Ammann, M.; Koop, T.; Poschl, U. Gas Uptake
̈
and Chemical Aging of Semisolid Organic Aerosol Particles. Proc. Natl.
Acad. Sci. U. S. A. 2011, 108, 11003−11008.
(35) Berkemeier, T.; Huisman, A. J.; Ammann, M.; Shiraiwa, M.;
(19) Martin, S. T. Phase Transitions of Aqueous Atmospheric
Particles. Chem. Rev. 2000, 100, 3403−3454.
Koop, T.; Poschl, U. Kinetic Regimes and Limiting Cases of Gas
̈
Uptake and Heterogeneous Reactions in Atmospheric Aerosols and
7032
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033

The Journal of Physical Chemistry A
Article
Clouds: A General Classification Scheme. Atmos. Chem. Phys. 2013,
13, 6663−6686.
(36) You, Y.; Renbaum-Wolff, L.; Carreras-Sospedra, M.; Hanna, S.
J.; Hiranuma, N.; Kamal, S.; Smith, M. L.; Zhang, X.; Weber, R. J.;
Shilling, J. E.; et al. Images Reveal That Atmospheric Particles Can
Undergo Liquid-Liquid Phase Separations. Proc. Natl. Acad. Sci. U. S.
A. 2012, 109, 13188−13193.
(58) Aiken, A. C.; DeCarlo, P. F.; Kroll, J. H.; Worsnop, D. R.;
Huffman, J. A.; Docherty, K. S.; Ulbrich, I. M.; Mohr, C.; Kimmel, J.
R.; Sueper, D.; et al. O/C and OM/OC Ratios of Primary, Secondary,
and Ambient Organic Aerosols with High-Resolution Time-of-Flight
Aerosol Mass Spectrometry. Environ. Sci. Technol. 2008, 42, 4478−
4485.
(59) Donahue, N. M.; Kroll, J. H.; Pandis, S. N.; Robinson, A. L. A
Two-Dimensional Volatility Basis Set − Part 2: Diagnostics of
Organic-Aerosol Evolution. Atmos. Chem. Phys. 2012, 12, 615−634.
(60) Saukko, E.; Lambe, A. T.; Massoli, P.; Koop, T.; Wright, J. P.;
Croasdale, D. R.; Pedernera, D. A.; Onasch, T. B.; Laaksonen, A.;
Davidovits, P.; et al. Humidity-Dependent Phase State of SOA
Particles from Biogenic and Anthropogenic Precursors. Atmos. Chem.
Phys. 2012, 12, 7517−7529.
(37) Kuwata, M.; Martin, S. T. Phase of Atmospheric Secondary
Organic Material Affects Its Reactivity. Proc. Natl. Acad. Sci. U. S. A.
2012, 109, 17354−17359.
(38) Shiraiwa, M.; Pfrang, C.; Koop, T.; Poschl, U. Kinetic Multi-
̈
Layer Model of Gas-Particle Interactions in Aerosols and Clouds (KM-
GAP): Linking Condensation, Evaporation and Chemical Reactions of
Organics, Oxidants and Water. Atmos. Chem. Phys. 2012, 12, 2777−
2794.
(61) Robinson, E. S.; Saleh, R.; Donahue, N. M. Organic Aerosol
Mixing Observed by Single-Particle Mass Spectrometry. J. Phys. Chem.
A 2013, 117, 13935−13945.
(39) Angell, C. A. Liquid Fragility and the Glass Transition in Water
and Aqueous Solutions. Chem. Rev. 2002, 102, 2627−2650.
(40) Zobrist, B.; Marcolli, C.; Pedernera, D. A.; Koop, T. Do
Atmospheric Aerosols Form Glasses? Atmos. Chem. Phys. 2008, 8,
5221−5244.
(62) Zaveri, R. A.; Easter, R. C.; Shilling, J. E.; Seinfeld, J. H.
Modeling Kinetic Partitioning of Secondary Organic Aerosol and Size
Distribution Dynamics: Representing Effects of Volatility, Phase State,
and Particle-Phase Reaction. Atmos. Chem. Phys. 2014, 14, 5153−5181.
(63) Simperler, A.; Kornherr, A.; Chopra, R.; Bonnet, P. A.; Jones,
W.; Motherwell, W. D. S.; Zifferer, G. Glass Transition Temperature
of Glucose, Sucrose, and Trehalose: An Experimental and in Silico
Study. J. Phys. Chem. B 2006, 110, 19678−19684.
(41) Roberts, C. J.; Debenedetti, P. G. Engineering Pharmaceutical
Stability with Amorphous Solids. AIChE J. 2002, 48, 1140−1144.
(42) Zobrist, B.; Soonsin, V.; Luo, B. P.; Krieger, U. K.; Marcolli, C.;
Peter, T.; Koop, T. Ultra-Slow Water Diffusion in Aqueous Sucrose
Glasses. Phys. Chem. Chem. Phys. 2011, 13, 3514−3526.
(43) Singh, S. A Review on Spray Drying: Emerging Technology in
Food Industry. Int. J. Appl. Eng. Technol. 2014, 4, 1−8.
(44) Cal, K.; Sollohub, K. Spray Drying Technique. I: Hardware and
Process Parameters. J. Pharm. Sci. 2010, 99, 575−586.
(45) Yamaguchi, M.; Tsukamoto, M.; Hirao, I. The Michael Reaction
of Simple Ester Enolates to α,β-Unsaturated Esters. Chem. Lett. 1984,
375−376.
(64) Roos, Y. Melting and Glass Transitions of Low Molecular
Weight Carbohydrates. Carbohydr. Res. 1993, 238, 39−48.
(65) Lu, Q.; Zografi, G. Properties of Citric Acid at the Glass
Transition. J. Pharm. Sci. 1997, 86, 1374−1378.
(66) Timko, R. J.; Lordi, N. G. Thermal Characterization of Citric
Acid Solid Dispersions with Benzoic Acid and Phenobarbital. J. Pharm.
Sci. 1979, 68, 601−605.
(67) Crowe, L. M.; Reid, D. S.; Crowe, J. H. Is Trehalose Special for
Preserving Dry Biomaterials? Biophys. J. 1996, 71, 2087−2093.
(68) Miller, D.; Pablo, J. de; Corti, H. Thermophysical Properties of
Trehalose and Its Concentrated Aqueous Solutions. Pharm. Res. 1997,
14, 578−590.
(69) Ding, S.-P.; Fan, J.; Green, J. L.; Lu, Q.; Sanchez, E.; Angell, C.
A. Vitrification of Trehalose by Water Loss from Its Crystalline
Dihydrate. J. Therm. Anal. 1996, 47, 1391−1405.
̈
(46) Alder, K.; Soll, H.; Soll, H. Uber Die Indirekte Substituierende
̈
̈
Addition. Justus Liebigs Ann. Chem. 1949, 565, 73−99.
(47) Riechers, B.; Wittbracht, F.; Hutten, A.; Koop, T. The
Homogeneous Ice Nucleation Rate of Water Droplets Produced in a
Microfluidic Device and the Role of Temperature Uncertainty. Phys.
Chem. Chem. Phys. 2013, 15, 5873−5887.
̈
(48) Tong, H.-J.; Reid, J. P.; Bones, D. L.; Luo, B. P.; Krieger, U. K.
Measurements of the timescales for the mass transfer of water in glassy
aerosol at low relative humidity and ambient temperature. Atmos.
Chem. Phys. 2011, 11, 4739−4754.
(49) Price, H. C.; Murray, B. J.; Mattsson, J.; O’Sullivan, D.; Wilson,
T. W.; Baustian, K. J.; Benning, L. G. Quantifying water diffusion in
high-viscosity and glassy aqueous solutions using a Raman isotope
tracer method. Atmos. Chem. Phys. 2014, 14, 3817−3830.
(50) Berkemeier, T.; Shiraiwa, M.; Poschl, U.; Koop, T. Competition
̈
between water uptake and ice nucleation by glassy organic aerosol
particles. Atmos. Chem. Phys. Discuss. 2014, 14, 16451−16492.
(51) Liu, B. Y.; Lee, K. W. An Aerosol Generator of High Stability.
Am. Ind. Hyg. Assoc. J. 1975, 36, 861−865.
(52) Gordon, B. M.; Taylor, J. S. Ideal Copolymers and the Second-
Order Transition of Synthetic Rubbers. I. Non-Crystalline Copoly-
mers. J. Appl. Chem. 1952, 2, 493−500.
(53) Roos, Y. H. Glass Transition Temperature and Its Relevance in
Food Processing. Annu. Rev. Food Sci. Technol. 2010, 1, 469−496.
(54) Corti, H. R.; Angell, C. A.; Auffret, T.; Levine, H.; Buera, M. P.;
Reid, D. S.; Roos, Y. H.; Slade, L. Empirical and Theoretical Models of
Equilibrium and Non-Equilibrium Transition Temperatures of
Supplemented Phase Diagrams in Aqueous Systems (IUPAC
Technical Report). Pure Appl. Chem. 2010, 82, 1065−1097.
(55) Koop, T.; Luo, B.; Tsias, A.; Peter, T. Water Activity as the
Determinant for Homogeneous Ice Nucleation in Aqueous Solutions.
Nature 2000, 406, 611−614.
(56) Koop, T. Homogeneous Ice Nucleation in Water and Aqueous
Solutions. Z. Phys. Chem. 2004, 218, 1231−1258.
(57) Koop, T.; Zobrist, B. Parameterizations for Ice Nucleation in
Biological and Atmospheric Systems. Phys. Chem. Chem. Phys. 2009,
11, 10839−10850.
7033
dx.doi.org/10.1021/jp505910w | J. Phys. Chem. A 2014, 118, 7024−7033