Atmospheric aerosols influence the climate via cloud
droplet nucleation and can facilitate the long-range transport of harmful
pollutants. The lifetime of such aerosols can therefore determine their
environmental impact. Fatty acids are found in organic aerosol emissions
with oleic acid, an unsaturated fatty acid, being a large contributor to
cooking emissions. As a surfactant, oleic acid can self-organise into
nanostructured lamellar bilayers with its sodium salt, and this
self-organisation can influence reaction kinetics. We developed a kinetic
multi-layer model-based description of decay data we obtained from
laboratory experiments of the ozonolysis of coated films of this
self-organised system, demonstrating a decreased diffusivity for both oleic
acid and ozone due to lamellar bilayer formation. Diffusivity was further
inhibited by a viscous oligomer product forming in the surface layers of the
film. Our results indicate that nanostructure formation can increase the reactive half-life of oleic acid by an order of days at typical indoor and
outdoor atmospheric ozone concentrations. We are now able to place
nanostructure formation in an atmospherically meaningful and quantifiable
context. These results have implications for the transport of harmful
pollutants and the climate.
Introduction
Atmospheric aerosols represent a large uncertainty when considering their
impact on human-made climate change
(Boucher et al., 2013) and can
be associated with poor air quality in urban areas
(Chan
and Yao, 2008; Kulmala et al., 2021; Li et al., 2018; Molina, 2021). The
organic fraction of atmospheric aerosols includes a diverse range of
molecules with differing functionalities, varying with season and
environment
(Jimenez
et al., 2009; T. Wang et al., 2020).
Cooking emissions are key contributors to urban aerosols
(Ots
et al., 2016; Vicente et al., 2021). Fatty acids are a well-established set
of marker compounds used to track cooking emissions due to their relatively
high abundance (Q. Wang
et al., 2020; Zhao et al., 2015). In particular, oleic acid, an unsaturated fatty acid, has been used to follow the ageing of cooking aerosols
(Q. Wang et al., 2020). The
lifetime of oleic acid in the atmosphere is longer compared with laboratory
predictions (days compared to hours)
(e.g.
Pfrang et al., 2011; Robinson et al., 2006; Rudich et al., 2007; Wang and
Yu, 2021). This is a long-standing discrepancy and suggests that some
physical process is inhibiting the ageing of such aerosols. There is also
field evidence of a difference in atmospheric lifetime between oleic acid
and its trans isomer (elaidic acid), suggesting that the confirmation of the molecule (i.e. how the molecules organise themselves) plays a role in
inhibiting reactivity (Wang and Yu, 2021). For these reasons, oleic acid has
been the compound of choice for laboratory studies into aerosol
heterogeneous oxidation
(Gallimore
et al., 2017; King et al., 2020; Milsom et al., 2021b, a; Pfrang et al.,
2017; Woden et al., 2021; Zahardis and Petrucci, 2007; Berkemeier et al., 2021).
The phase state and viscosity of atmospheric aerosols can impact on
heterogeneous processes such as reactive gas and water uptake
(Reid et al., 2018; Shiraiwa
et al., 2011). Field measurements have shown that semi-solid-phase formation takes place in the atmosphere (Virtanen et
al., 2010) and that phase state can vary between night and day as well as
with organic mass fraction
(Slade et al., 2019). The
long-range transport of harmful polycyclic aromatic hydrocarbons (PAHs) has
been linked with particle phase state and the formation of a viscous organic
layer, protecting the aerosol's potentially harmful contents
(Mu et al.,
2018; Shrivastava et al., 2017). Viscous-phase formation is therefore a plausible explanation for the persistence of organic aerosol components in
the atmosphere.
As a surfactant, oleic acid can self-organise into a range of
nanostructures, known as lyotropic liquid crystals (LLCs), with its sodium
salt and water
(Mele et al.,
2018; Seddon et al., 1990). The viscosity and the diffusion of small
molecules through these phases can vary significantly
(Mezzenga et al., 2005;
Zabara and Mezzenga, 2014). Some nanostructures, such as the lamellar phase,
have highly anisotropic diffusivities resulting in substantially higher
diffusivity parallel to the lamellar bilayer compared to the perpendicular
direction (Lindblom and Orädd, 1994). This
nanostructure formation has been studied in levitated droplets and in coated
quartz capillaries
(Pfrang et al.,
2017; Seddon et al., 2016), where the self-organisation of this proxy system
decreased the reactivity of oleic acid by approximately an order of
magnitude (Milsom et al., 2021b).
The nanostructure studied here is the lamellar phase. This consists of a
bilayer of surfactant molecules with their alkyl tails directed inwards. The
lamellar phase studied here is anhydrous, with no water between the bilayers (see the cartoon in Fig. 1b). This lamellar phase is liquid crystalline, as opposed to the crystalline lamellar phase observed previously in levitated
particles (Milsom et al., 2021a). This is due to the lack of characteristic
wide-angle X-ray scattering (WAXS) peaks returned by these samples (Milsom
et al., 2021b), characteristic of the crystalline form of this lamellar
phase (Tandon et al., 2001; Milsom et al., 2021a). Liquid oleic acid does
exhibit some order via the formation of dimers. This has been observed in
the literature, and we have previously confirmed this experimentally (Iwahashi et al., 1991; Milsom et al., 2021b).
(a) A schematic representation of the small-angle X-ray scattering
(SAXS) and Raman spectroscopy experiments. (b) The lamellar phase formed by
oleic acid and sodium oleate.
In the present work, we develop a model description of self-organised oleic
acid ozonolysis and apply this both to kinetic data of the lamellar phase
presented recently by Milsom et al. (2021b) and also to new liquid-phase oleic acid ozonolysis data measured by Raman microscopy. We determine the effect on particle diffusivity of both
nanostructure formation and the formation of a later-stage reaction product, which congregates in the surface region of the film. We then predict the
impact on the atmospheric lifetime of such films, linking this to the
discrepancy between measured and predicted atmospheric lifetimes for oleic
acid (Robinson et al., 2006;
Rudich et al., 2007).
Methodology
In this study, oleic acid refers to both oleic acid and sodium oleate as
they are both constituents of the lamellar-phase bilayer. Pure oleic acid is referred to as liquid oleic acid. Oleic acid and sodium oleate represent the
conjugate acid and conjugate base form of the same common organic aerosol
component and are expected to be present together in intermediate pH ranges
(oleic acid pKa is ca. 5.0).
For the present study we made liquid oleic acid capillary coatings, exposed
them to ozone and followed the kinetics by Raman microscopy. The coatings
were prepared by the following method (compare
Milsom et al., 2021b): oleic acid (90 % purity, Sigma-Aldrich) was
dissolved as a 10 wt % solution in methanol. A 70 µL aliquot of
this mixture was passed up and down a quartz capillary tube (Capillary Tube
Supplies Ltd., UK, 1.5±0.25 mm diameter, wall thickness 0.010 mm)
embedded in a metal holder until the methanol had evaporated from the
solution, aided by passing room-temperature condensed air through the
capillary.
The Raman microscopy and ozonolysis experiment is based on the set-up detailed in Milsom et al. (2021b). The set-up is summarised along with the small-angle X-ray scattering (SAXS)
experiment used to measure kinetics in the lamellar phase in Fig. 1a. A
long-working-distance objective lens (0.42 numerical aperture) was used to
focus a 532 nm laser on the capillary. The minimum spot diameter was
∼1.5µm, and the laser power emitted was between 20 and 50 mW. Oxygen (BOC, 99.5 %) was passed through a pen-ray ozoniser
(Ultraviolet Products Ltd, Cambridge, UK) to produce ozone. The ozone
concentration was calibrated offline by UV spectroscopy using the absorption
cross section for ozone at 254 nm ((1.14±0.07)×10-17cm2) (Mauersberger et al., 1986). A concentration of 77±5 ppm was used for comparison with the lamellar-phase kinetics presented by Milsom et al. (2021b).
Four datasets from the same coated capillary were selected from Milsom et al. (2021b) for the following reasons. (i) They are from different sections of the same coated capillary – the experimental
conditions are exactly the same (77±5 ppm ozone, dry oxygen–ozone flow). (ii) They have thicknesses <5µm – atmospherically relevant. (iii) They are complete decays – more constraint on the model
fit as the reaction was followed to completion. All error bars from these
data points are derived from the uncertainty in the measured scattered X-ray intensity. The ozone concentration used was much higher than that found in
the atmosphere due to the major time limitations associated with synchrotron
beamtime experiments.
The experimental data were modelled following the approach of the kinetic
multi-layer model of aerosol surface and bulk chemistry (KM-SUB;
Shiraiwa et al., 2010) based on
the Pöschl–Rudich–Ammann (PRA) framework (Pöschl et al., 2007). An oleic
acid ozonolysis reaction scheme was chosen where oligomer formation,
viscosity and diffusivity were explicitly treated
(compare Hosny et al.,
2016). Our model uses a flat film geometry. The model description and the
reaction scheme used are presented in Sect. S1 in the Supplement.
The model was written in the Python programming language. A series of ordinary differential equations (ODEs) describes the change in concentration for each model component in each model bulk and surface layer over time. These ODEs
were integrated using the SciPy solve_ivp solver with a backward differentiation formula (BDF) for stiff ODE solving (Virtanen et al., 2020).
Parameters associated with reaction rate constants, Henry's law coefficient
and the gas uptake coefficient for ozone into the organic phase were set to
values used in the previous oleic acid ozonolysis literature (all model parameters are summarised in Sect. S2 in the Supplement).
The diffusion of model components throughout the film was allowed to vary
with composition. It is known that self-organised phase formation affects
viscosity and diffusivity
(Mezzenga et al., 2005;
Zabara and Mezzenga, 2014). Therefore, determining the effect of particle
diffusivity on reactivity is a key focus of this study. A Vignes-type
diffusion regime was employed to account for the effect of composition on
molecular diffusivity
(Alpert
et al., 2019; Vignes, 1966; S. Zhou et al., 2019). The diffusion of model
components was dependent on the fraction of lamellar oleic acid as well as
dimer and trimer oligomers formed by oleic acid ozonolysis
(Lee
et al., 2012; Zahardis et al., 2006).
The Vignes-type diffusion parameterisation is outlined in Eqs. (1) and (2).
1DY,i=(DY,lam)1-fdi,i-ftri,i×(DY,di)fdi,i×(DY,tri)ftri,i2DX,i=(DX,lam)1-fdi,i-ftri,i×(DX,di)fdi,i×(DX,tri)ftri,iY in Eq. (1) refers to oleic acid and 9-carbon monomer products, the
diffusion of which in each model layer (i) was treated the same under the
assumption that monomer diffusion is strongly linked to oleic acid
diffusion. X corresponds to the reactive gas, in this case ozone. The
fraction of dimer (fdi,i) and trimer (ftri,i) in model layer
i was
used to represent layer composition. The diffusion coefficients of
components in the lamellar (DY,lam and DX,lam), dimer (DY,di
and DX,di) and trimer (DY,tri and DX,tri) media were allowed to
vary during the model fitting.
The diffusion of the dimer and trimer products was treated using a power law relationship via a scaling factor (fdiff) in line with an oleic acid
ozonolysis modelling study that focussed on viscosity data measurements
(Hosny et al., 2016). We
adapted this parameterisation to define oligomer diffusivity rather than
viscosity.
Dtri=DdiMdiMtrifdiffDtri and Ddi are the diffusion coefficients of the trimer and
dimer, respectively. Ddi was allowed to vary during the model fitting
procedure but was not itself made to be composition-dependent. We found that
the model was particularly intensive to diffusivity in the dimer (Figs. 6b
and S3e in the Supplement). This therefore did not justify adding additional parameters
and computational resources required to evolve dimer diffusivity. Mdi and Mtri are the respective molecular masses of the dimer and trimer
products.
The model output was fitted to experimental data using a differential
evolution (DE) global optimisation algorithm (Storn and Price,
1997). An initial population of candidate parameter sets was created by
Latin hypercube sampling of the parameter space (McKay et al., 1979). This was carried out in
parallel, similar to the procedure described by Berkemeier et al. (2017), who used Monte
Carlo sampling to initialise their candidate parameter sets. The DE algorithm is a
popular one for finding the global minimum of a loss function used to
evaluate model fitness, which in this case was the mean-squared error of the
model fit. This fitting procedure was implemented using the DE method in the
optimise module of the SciPy package
(Virtanen et al., 2020).
Twenty computer processor cores were used for parallelisation of the differential evolution algorithm. The model was optimised to all the datasets simultaneously, analogous to the method recently employed by Berkemeier et al. (2021). The
loss function from each experimental fit was weighted according to the
number of data points fitted to. Separate model optimisations to each
individual dataset were carried out in order to find a range of optimised
parameter values.
The sensitivity of the model to the varied parameters was investigated using
an elementary effects algorithm via the method of Morris implementation of the SALibPython package (Campolongo et al., 2007; Herman
and Usher, 2017; Morris, 1991). The total loss rate of oleic acid after 50 % has reacted was used as the output variable. Normalised sensitivity
coefficients for each varied parameter were then obtained by measuring
changes in the total loss rate of oleic acid with changes in each model
parameter.
Model sensitivity to ozone solubility and surface accommodation coefficient
was also explored. We found that varying the Henry's law constant by 1 order of magnitude more and less than that used here caused little change in
the model output (Fig. S3b). The accommodation coefficient was also varied
with some impact observed on the model output (Fig. S3d). Without
experimental constraint on these parameters and the surface desorption
lifetime of ozone, all of which are associated with surface and bulk uptake
(Shiraiwa et al., 2010), a range of optimum parameter combinations is
possible. We therefore decided to hold these parameters to plausible values
from the modelling literature (see Table S1 in the Supplement), highlighting the more
significant effect of diffusion in this system.
We found that the model was somewhat insensitive to the branching ratio
between the volatile nonanal and other monomer 9-carbon products (Fig. S3a). The value used in this study (0.454) agrees with experimentally
determined yields (Hearn and Smith, 2004).
Results and discussion
The experimental decays, derived from SAXS peak areas, are a direct measure
of oleic acid decay in the lamellar phase (Fig. 2). The half-lives of the
self-organised films are significantly longer than that of liquid oleic acid
(Table 1). This suggests a significant diffusion limitation to the reaction
due to the formation of this viscous self-organised phase. The half-lives of
the self-organised films are also thickness-dependent. Both observations are consistent with previous work on self-organised oleic acid (Milsom et al.,
2021a, b, Pfrang et al., 2017).
Kinetic decay plots of normalised lamellar-phase oleic acid concentration ([OA]Lam/[OA]Lam,0) as a function of time
(experimental data from Milsom et al., 2021b); model predictions are based
on the optimised model parameters determined by fitting all the data
simultaneously. Individual fits to each dataset are also presented. Film
thicknesses are displayed in each legend. The grey-shaded regions represent the range of model outputs using parameter sets optimised from each
individual fit.
The half-life of the films used in this study (to the nearest
minute). Taken from individual model fits to the data.
Film thickness (µm)Half-life (min)0.59 (Lam.)∼110.91–1.66 (Lam.)∼18–220.6–0.9∗ (Liq.)∼1–2
∗ The range of half-lives from model outputs presented in Fig. 3. Lam.: lamellar-phase oleic acid; Liq.: liquid oleic acid.
Forcing the model to fit the data without considering composition-dependent diffusion results in a significantly poorer quality of fit (Fig. S3c). In addition
to this, we selected composition-dependent diffusion for a few reasons: (i)
there is experimental evidence that aggregates form on the outside of
levitated particles of the crystalline form of this proxy (Milsom et al., 2021a) – this aggregate is likely to be viscous. (ii) The kinetic decay from a much thicker portion of the same film (∼73µm)
effectively stopped by the end of the experiment (Milsom et al., 2021b) –
a secondary factor, such as a more viscous outer crust, may be inhibiting the reaction. (iii) The viscosity of oleic acid particles is known to increase during ozonolysis (Hosny et al., 2016). This increased viscosity is believed to affect the
diffusivity of trace gases, such as ozone, through the condensed phase
(Shiraiwa et al., 2011). It follows that, as the composition of the film changes, so could the diffusivity of ozone.
Diffusion behaviour
The optimised model parameters from simultaneous fitting of all datasets
returned good fits for the experimental data measured at 0.59 and 0.98 µm film thickness (Fig. 2). The 0.91 and 1.66 µm
films returned poorer fits than the other films (see the succeeding discussion). A summary of the optimised diffusion parameters is presented in Table 2 and
those from the individual fits are presented in Table S2 in the Supplement.
Optimised diffusion parameters allowed to vary during model
optimisation. The full set of model parameters is available in
Sect. S2 in the Supplement.
ParameterDescriptionValuea/cm2s-1Rangeb/cm2s-1DdimerBulk diffusion coefficient of the dimer1.03×10-121.03×10-12–9.49×10-10DtrimerBulk diffusion coefficient of the trimer2.07×10-132.07×10-13–1.90×10-10DX,lamBulk diffusion coefficient of O3 in the lamellar phase3.35×10-121.13×10-12–8.78×10-8DY,lamBulk diffusion coefficient of oleic acid and monomer products in the lamellar phase2.81×10-127.32×10-13–2.81×10-12DX,diBulk diffusion coefficient of O3 in the dimer4.66×10-92.14×10-9–7.34×10-9DY,diBulk diffusion coefficient of oleic acid and monomer products in the dimer8.85×10-115.03×10-11–9.93×10-11DX,triBulk diffusion coefficient of O3 in the trimer1.49×10-121.49×10-12–9.76×10-9DY,triBulk diffusion coefficient of oleic acid and monomer products in the trimer8.16×10-111.24×10-11–9.88×10-11
a From simultaneous fitting to all experimental
datasets in Fig. 2. b From individual fits to each experimental dataset in Fig. 2.
Ozone diffusion in the lamellar phase (DX,lam=3.35×10-12cm2s-1) is consistent with the diffusion of a reactive
gas through a highly viscous matrix (Shiraiwa et
al., 2011a). The spacing between lamellar alkyl chains in this system is
4.41 Å (Milsom et al., 2021b), which is close to
the molecular diameter of ozone of 4 Å used here
(Pfrang
et al., 2010; Shiraiwa et al., 2010). It has been suggested that the shorter
spacing between fatty acid tails on a particle surface could provide steric
hindrance to diffusing ozone molecules, limiting access to the carbon–carbon double bond
(Hearn et al.,
2005; Vieceli et al., 2004). The anhydrous lamellar phase, being viscous and
with closely packed alkyl chains, is likely to present extra steric
hindrance compared to surface monolayers because this effect would prevail
throughout the film regardless of the orientation of the lamellae relative
to the substrate.
Ozone diffusion in the dimer is faster than in the lamellar phase. This is consistent with a steric hindrance argument. The assumed unordered nature of
the dimer product suggests that ozone can diffuse past these molecules more
easily compared with diffusion through the restricted bilayers formed by the
lamellar-phase oleic acid. By contrast, diffusion through the trimer product is slower than in the dimer and lamellar phase. The trimer product in this
model represents all the higher-order oligomers that can be formed during
oleic acid ozonolysis, contributing to an increase in particle viscosity
(Hosny et al., 2016).
The diffusivity of oleic acid is low in the lamellar phase (DY,lam=2.81×10-12cm2s-1) compared to ∼1.53×10-9cm2s-1 for pure liquid oleic acid
(DY,liq) based on its viscosity at 293.15 K
(Sagdeev et al., 2019).
Experimentally determined surfactant lateral diffusion coefficients in
hydrated lamellar bilayers are at least 4 orders of magnitude higher than our model optimisation returned (∼10-8–10-6cm2s-1) (Lindblom and Orädd, 1994;
Lindblom and Wennerström, 1977). Note that these experimental
determinations were on hydrated lamellar phases, which are expected to be
less viscous than the anhydrous lamellar phase studied here due to water
acting as a plasticiser. The model does not deconvolute directionally
dependent diffusion through the lamellar phase because no bilayer
orientation was observed: 2-D SAXS patterns obtained for these samples did
not exhibit any alignment of the lamellar phase, and lamellae were randomly oriented relative to the substrate (see Fig. S2 in the Supplement), though
there is qualitative evidence of some degree of parallel orientation at the
surface (Milsom et al., 2021b).
The diffusivity of oleic acid in the trimer product is within an order of
magnitude of the lamellar-phase oleic acid diffusivity. After an increase in diffusivity going from the lamellar phase to the dimer phase (DY,di=8.85×10-11cm2s-1), oleic acid diffusivity
decreases in the trimer (DY,tri=8.16×10-11cm2s-1).
The diffusion constants returned from the optimised model are within the
range expected for a semi-solid system (Shiraiwa et al., 2011). However,
there remains significant uncertainty about the true value of some parameters (notably DX,lam; see Table 2). Thus, we caution against over-interpretation
of the absolute values, but we are reasonably confident in the general trends presented here.
Differences between the model and data may arise for a number of reasons
associated with the experiment: (i) there is an uncertainty associated with
the film thickness measurement; in particular, the 0.91 and 0.98 µm films are similar when considering their quoted thickness uncertainties
representing 1 standard deviation (0.03 µm) (Milsom et al., 2021b). (ii) If the film structure
changes over time exposed to ozone, which has been observed under a
microscope (Hung and Tang, 2010), the surface area available for
ozone uptake may also change. This change in surface structure is not
considered in the model since it would require an experimental determination
of the surface roughness not possible using SAXS. (iii) The film may have been slightly thicker on one side of the capillary compared to the other:
this technique required the X-ray beam to pass through both sides of a
coated quartz capillary. Given that film thickness affects reaction
kinetics, a difference in film thickness between both sides could impact the
experimental result. (iv) The film thickness could have varied over the part
of the film illuminated by the X-ray beam: the beam was ∼320µm×400µm in size, and therefore the film thickness is an average of the illuminated area. These arguments could account for the
range of fitted model parameters when each dataset was fitted separately
(see Table 1).
We can rule out any variation in sample environment because all these
datasets were obtained at different positions along the same capillary
during the same ozonolysis experiment. Thus, we are confident that the film
structure and morphology must have some impact on reactivity on this
thickness scale.
Viscous phases have been demonstrated to drive ozonolysis chemistry down a
specific route, influencing the product distribution (Z. Zhou et al., 2019).
This is certainly possible for this viscous oleic acid system. However, there does not currently exist a product identification and distribution study for
self-organised oleic acid. This is a motivation for future work in order to
constrain this new aspect of the oleic acid–ozone heterogeneous system.
Following the mixing rule presented by Hosny et al. (2016), the final
viscosity of these self-organised films was ∼1800mPas, this close to the experimental region reported for ozonised liquid oleic acid
particles (∼1200–1400 mPas), reported as a lower estimate
(Hosny et al., 2016).
In order to contrast liquid and nanostructured oleic acid kinetic decays,
ozonolysis of liquid oleic coated inside a quartz capillary was carried out
and followed by Raman microscopy – a technique previously used to follow
oleic acid reaction kinetics
(King et al., 2004;
Pfrang et al., 2017). We then applied the optimised model to these
experimental data, replacing the diffusivity of ozone and oleic acid in the
lamellar phase (DX,lam and DY,lam, respectively) with values for
diffusion through liquid-phase oleic acid (DX,liq and DY,liq, respectively). For the diffusion of ozone in liquid oleic acid, we used the
value from previous modelling studies on oleic acid ozonolysis
(Pfrang
et al., 2010; Shiraiwa et al., 2010).
Encouragingly, the optimised model returned a reasonable fit to ozonolysis
decay data obtained by Raman spectroscopy on a film coated with pure oleic
acid in the liquid state (Fig. 3). This film was prepared in the same way as
the semi-solid films, and therefore it is not unreasonable to assume a similar film thickness. We varied the modelled film thickness and found that a range
of film thicknesses (0.6–0.9 µm) fitted best to these data. Note
that these data are noisier than those derived from SAXS. The concentration
evolution of the model components from the fit with a 0.9 µm film
thickness is presented in the Supplement (Fig. S1).
Kinetic decay plot of the ozonolysis of liquid oleic acid measured by Raman microscopy. Model output for two film thicknesses (0.6 and 0.9 µm) with liquid oleic acid diffusion parameters (DY,liq=1.53×10-9cm2s-1, DX,liq=1.00×10-5cm2s-1, replacing DY,lam and DX,lam).
Experimental [O]3=77±5 ppm.
Spatial and temporal evolution of composition and diffusion
The spatial and temporal evolution of ozone concentration is consistent with
a bulk diffusion-limited reaction. The concentration of ozone in the
majority of the film bulk does not exceed ∼1 % of the
concentration in the surface layers (Fig. 4a – effectively no ozone near
the film substrate). The steep ozone concentration gradient developed during
the reaction is illustrated by the log scale in Fig. 3a.
Spatially and temporally resolved concentration evolution of ozone
(a – log-scale concentration), oleic acid (b), dimer (c) and trimer (d) model components during ozonolysis for a 0.98 µm film – d: the
distance from the film–substrate interface. Contours illustrate the change in concentration gradient over time for the non-reactive gas species.
[O3]=77 ppm.
Diffusion of ozone through regions of higher viscosity is expected to be
slower, and the formation of a crust in the surface layers of the film, consisting of the viscous trimer product, inhibits the diffusion of ozone
through the particle (Figs. 4d and 5). The formation of a surface
crust has been postulated in the literature
(Pfrang
et al., 2011; Zhou et al., 2013), and direct experimental evidence of surface product aggregation has recently been presented in a similar proxy
(Milsom et al., 2021a).
The evolution of ozone diffusivity throughout a 0.98 µm
film during ozonolysis. [O3]=77 ppm. d: the distance from the film–substrate interface.
Similarly, an oleic acid concentration gradient also develops during the
reaction (Fig. 4b). This gradient is not as steep as the one observed for
ozone but is still noteworthy. Surface crust formation is the source of
increasing diffusive inhibition during the reaction and is therefore a key factor inhibiting the oleic acid ozonolysis kinetics for this system.
The atmospheric implications of this diffusive inhibition, caused by the
initial phase state and crust formation, are explored in Sect. 3.4.
Kinetic regime analysis
The model output was most sensitive to ozone and oleic acid diffusivity,
highlighting that film phase heavily influences its lifetime (Fig. 6b).
From this analysis and the concentration profiles (Fig. 4), we can conclude that the reaction is limited both by bulk oleic acid diffusion to the
reaction region and by the diffusivity of ozone through the film –
illustrated by the concentration gradients observed for both components
(Fig. 4a and b). The model was least sensitive to diffusion
coefficients of ozone and oleic acid in the dimer.
(a) A “kinetic cube” plot (described by Berkemeier et al., 2013) of surface-to-total-loss ratio (STLR), bulk mixing parameter (BMP) and bulk saturation ratio (BSR) for a model run
at 77 ppm ozone and 0.98 µm film thickness. The black arrow
illustrates the movement from the mass transfer to the reaction–diffusion kinetic regimes described by Berkemeier et al. (2013). (b) A summary of the
normalised sensitivity coefficients for each varied model parameter.
Further analysis using a method described by Berkemeier et al. (2013) for multi-layer model outputs demonstrates the evolution of the kinetic regime as ozonolysis
proceeds (Fig. 6a). The surface-to-total-loss ratio (STLR) observed throughout the reaction is close to zero, suggesting that reactant loss is not a
surface-dominated process. The bulk mixing parameter (BMP) starts at
∼0.18 and decreases with time. This is a measure of how well-mixed the particle is in terms of both the reactive gas- and condensed-phase reactants – a value of one is well-mixed. The film therefore starts poorly
mixed and becomes less well-mixed as the reaction progresses. After an
initial transient phase, the bulk saturation ratio (BSR) increases steadily
over time. This reflects the supply of the reactive gas to the film, which
is inhibited by viscous product formation and the viscous lamellar phase.
For an appreciable amount of time the reaction regime lies between a
mass-transfer and reaction–diffusion regime, illustrating the importance of both bulk diffusion and accommodation parameters at different times during
the reaction (Fig. 6a). The transient nature of the kinetic regime
demonstrates the added insight obtained through this more explicit
description. Limiting cases based on a resistor model do not account for
changes in kinetic regime (Worsnop et al., 2002).
This kind of analysis demonstrates the power of spatially and temporally
resolved kinetic modelling, enabling us to present a more nuanced picture of
the kinetic regimes underpinning this reaction.
Atmospheric implications
There is a known discrepancy between laboratory-determined and field-based
lifetimes of fatty acids, such as oleic acid
(Robinson et al., 2006;
Rudich et al., 2007), and there is evidence fatty acid confirmation could affect atmospheric lifetime (Wang and Yu, 2021). In order to demonstrate the
potential impact self-organisation has on the atmospheric lifetime of such
organic coatings, our model was run with a film thickness range of 0.50–1.50 µm and an ozone concentration range of 10–150 ppb, covering
pristine (∼10 ppb), typical (20–40 ppb) and polluted
(>40 ppb) ozone concentrations in the urban and indoor
environment (Fig. 7) (Weschler, 2000).
Plots of film half-life as a function of ozone concentration
([O3]) and film thickness. (a) Model runs using parameters for liquid oleic acid (DY,liq=1.53×10-9cm2s-1,
DX,liq=1.00×10-5cm2s-1); (b) model runs
using the optimised parameters for lamellar-phase (nanostructured) oleic acid (DY,lam=2.81×10-12cm2s-1,
DX,lam=3.35×10-12cm2s-1); (c) resulting
increase in half-life due to nanostructure formation. Contours in each plot represent lines of constant half-life.
Taking a 1 µm film as an example, where the model agrees best with the experiment (Fig. 2c), the half-life increases from
∼1 to 10 d when moving from the liquid to the
nanostructured (lamellar) state at ∼30 ppb ozone
concentration (Fig. 7c). Such an increase in the atmospheric lifetime of
the organic film has implications for the persistence of organic matter in
such particles.
These predictions are most likely an estimate of half-life, especially for
the thicker films; the model over-predicts the experiment at 1.66 µm
(Fig. 2d). Phase changes can occur with changes in relative humidity (RH)
(Pfrang et al.,
2017; Seddon et al., 2016). This particular system is stable below 55 %
RH, above which the anhydrous lamellar phase can break down into inverse
micelles which are thought to be less viscous. Atmospheric humidity is
variable. Therefore, any phase transition to a less-viscous phase could
enhance ozone uptake and promote a faster reaction, decreasing the
half-life. The effects of different molecular arrangements are very challenging to determine experimentally (compare Milsom et al., 2021b).
An increased organic film lifetime also has direct implications for the lifetime of other particle constituents. Organic particulate matter can contain a range of chemical species, many of which are harmful to human
health (Chan and Yao, 2008). The long-range
transport of carcinogenic PAHs has been linked to particle phase state and the formation of a semi-solid organic coating on PAH-containing particles,
increasing the risk of ill health (Mu et al.,
2018; Shrivastava et al., 2017). Our model predictions show that the
semi-solidification of this atmospheric aerosol proxy can increase the
lifetime of the organic film substantially. Moreover, the formation of a
surface layer of high-molecular-weight products (represented as the trimer
in the model) forms an extra diffusional barrier to oxidants such as ozone.
This extension of atmospheric lifetime implies a slower rate of particle
oxidation. The degree of oxygenation, measured by the O:C ratio, is linked
to aerosol hygroscopicity
(Wu et al.,
2016). Therefore, it is possible that the inhibition of particle oxidation
by the formation of this semi-solid phase could have an impact on the cloud
condensation nucleus (CCN) ability of the particle. The increased lifetime
of oleic acid, and therefore the 9-carbon products included in this model,
suggests that surface-active material can persist for longer times in the
atmosphere in a semi-solid organic film. Two of the 9-carbon primary oleic
acid ozonolysis products, azelaic acid and nonanoic acid, may also be
surface-active under certain conditions (King et
al., 2009; Tuckermann, 2007, but also compare
King et al., 2020). Such surface-active
material has the potential to alter aerosol hygroscopicity by decreasing the
surface tension of aqueous droplets, affecting the aerosol's ability to act
as a CCN (Ovadnevaite et al., 2017). The link
between clouds and aerosols is clear, and any process affecting the ability of an aerosol to act as a CCN can have an impact on the climate
(Boucher et al., 2013).
With cooking aerosols accounting for up to 10 % of reported PM2.5
emissions in the UK
(Ots et al.,
2016) and fatty acids being major contributors to cooking emissions also in
other regions such as China
(Q. Wang et al., 2020), it
follows that these effects would most likely be observed in the urban
environment.
Conclusions
The effect of the aerosol-phase state continues to be a key topic for the atmospheric aerosol community. In this study, a multi-layer kinetic model
was fitted to experimental data collected during ozonolysis of oleic acid
coatings in a self-organised semi-solid state.
A key advantage of this particular co-ordinated model–experiment approach is that all fitted experimental data were from samples exposed to exactly the
same conditions in the same sample environment. Therefore, differences
between model fits and experimental data are most likely originating from
variations in film structure and morphology rather than experimental
conditions, thus minimising uncertainties associated with other kinetic
techniques.
The increase in atmospheric lifetime of this proxy from hours to days is
consistent with field measurements of oleic acid demonstrating a
much-extended atmospheric lifetime in comparison to laboratory measurements.
Future work should focus on constraining film viscosity and diffusivity
experimentally and studying the effect of lamellar anisotropy on reaction
kinetics. Kinetic experiments on a highly aligned lamellar phase compared
with a randomly oriented lamellar phase would provide a key insight into the
role a bilayer of surfactant molecules could have in hindering the uptake of
trace gases to a film or particle.
We are now able to place nanostructure formation in an atmospherically
meaningful and quantifiable context, thus establishing a clear pathway to
determining the impact nanostructure formation could have on the atmospheric
lifetime of organic aerosol emissions.
Code availability
The model code was produced using MultilayerPy (version 1.0), which is available at 10.5281/zenodo.6411189 (Milsom et al., 2022a). The model code for this study is included in the data deposit and will be included in the next minor release of MultilayerPy (version 1.1) with instructions on how to run it.
Data availability
The underlying experimental and model data are available at 10.5281/zenodo.6421940 (Milsom et al., 2022b).
The supplement related to this article is available online at: https://doi.org/10.5194/acp-22-4895-2022-supplement.
Author contributions
AM wrote the initial draft of the manuscript, wrote the model in Python and
carried out the analysis and interpretation. CP contributed to the
interpretation of the results and contributed to the manuscript. AMS
contributed to the manuscript and discussion. ADW set up and supported the
Raman microscopy experiment on I22 at DLS. All the authors were involved in the Raman microscopy experiment.
Competing interests
The contact author has declared that neither they nor their co-authors have any competing interests.
Disclaimer
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Acknowledgements
Adam Milsom was funded by the NERC SCENARIO DTP (NE/L002566/1) and NERC grant (NE/T00732X/1) and was supported by the NERC CENTA DTP. This work was carried out with the support of the Diamond Light Source (DLS), instrument I22
(proposal SM21663). The authors are grateful to the Central Laser Facility
for access to key equipment for the Raman work carried out simultaneously
with the DLS beamtime experiments. Nick Terrill (DLS), Andy Smith (DLS) and
Tim Snow (DLS) are acknowledged for their support during the beamtime. The computations described in this paper were performed using the University of
Birmingham's BlueBEAR HPC service, which provides a high-performance computing service to the university's research community.
Financial support
This research has been supported by the Natural Environment Research Council (grant nos. NE/L002566/1 and NE/T00732X/1).
Review statement
This paper was edited by Markus Ammann and reviewed by two anonymous referees.
ReferencesAlpert, P. A., Arroyo, P. C., Dou, J., Krieger, U. K., Steimer, S. S., Förster, J. D., Ditas, F., Pöhlker, C., Rossignol, S., Passananti, M., Perrier, S., George, C., Shiraiwa, M., Berkemeier, T., Watts, B., and Ammann, M.: Visualizing reaction and diffusion in xanthan gum aerosol particles exposed to ozone, Phys. Chem. Chem. Phys., 21, 20613–20627, 10.1039/c9cp03731d, 2019.Berkemeier, T., Huisman, A. J., Ammann, M., Shiraiwa, M., Koop, T., and Pöschl, U.: Kinetic regimes and limiting cases of gas uptake and heterogeneous reactions in atmospheric aerosols and clouds: a general classification scheme, Atmos. Chem. Phys., 13, 6663–6686, 10.5194/acp-13-6663-2013, 2013.Berkemeier, T., Ammann, M., Krieger, U. K., Peter, T., Spichtinger, P., Pöschl, U., Shiraiwa, M., and Huisman, A. J.: Technical note: Monte Carlo genetic algorithm (MCGA) for model analysis of multiphase chemical kinetics to determine transport and reaction rate coefficients using multiple experimental data sets, Atmos. Chem. Phys., 17, 8021–8029, 10.5194/acp-17-8021-2017, 2017.Berkemeier, T., Mishra, A., Mattei, C., Huisman, A. J., Krieger, U. K., and Pöschl, U.: Ozonolysis of Oleic Acid Aerosol Revisited: Multiphase Chemical Kinetics and Reaction Mechanisms, ACS Earth Space Chem., 5, 3313–3323, 10.1021/acsearthspacechem.1c00232, 2021.
Boucher, O., Randall, D., Artaxo, P., Bretherton, C., Feingold, G., Forster, P., Kerminen, V.-M., Kondo, Y., Liao, H., Lohmann, U., Rasch, P., Satheesh, S. K., Sherwood, S., Stevens, B., and Zhang, X. Y.: Clouds and Aerosols, in: Climate Change 2013 – The Physical Science Basis, edited by: Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, 571–658, ISBN 978-1-107-05799-1, 2013.Campolongo, F., Cariboni, J., and Saltelli, A.: An effective screening design for sensitivity analysis of large models, Environ. Modell. Softw., 22, 1509–1518, 10.1016/j.envsoft.2006.10.004, 2007.Chan, C. K. and Yao, X.: Air pollution in mega cities in China, Atmos. Environ., 42, 1–42, 10.1016/j.atmosenv.2007.09.003, 2008.Gallimore, P. J., Griffiths, P. T., Pope, F. D., Reid, J. P., and Kalberer, M.: Comprehensive modeling study of ozonolysis of oleic acid aerosol based on real-time, online measurements of aerosol composition, J. Geophys. Res., 122, 4364–4377, 10.1002/2016JD026221, 2017.Hearn, J. D. and Smith, G. D.: Kinetics and product studies for ozonolysis reactions of organic particles using aerosol CIMS, J. Phys. Chem. A, 108, 10019–10029, 10.1021/jp0404145, 2004.Hearn, J. D., Smith, G. D., and Lovett, A. J.: Ozonolysis of oleic acid particles: evidence for a surface reaction and secondary reactions involving Criegee intermediates, Phys. Chem. Chem. Phys., 7, 501–511, 10.1039/b414472d, 2005.Herman, J. and Usher, W.: SALib: An open-source Python library for Sensitivity Analysis, J. Open Source Softw., 2, 97, 10.21105/joss.00097, 2017.Hosny, N. A., Fitzgerald, C., Vyšniauskas, A., Athanasiadis, A., Berkemeier, T., Uygur, N., Pöschl, U., Shiraiwa, M., Kalberer, M., Pope, F. D., and Kuimova, M. K.: Direct imaging of changes in aerosol particle viscosity upon hydration and chemical aging, Chem. Sci., 7, 1357–1367, 10.1039/c5sc02959g, 2016.Hung, H. and Tang, C.: Effects of Temperature and Physical State on Heterogeneous Oxidation of Oleic Acid Droplets with Ozone, J. Phys. Chem. A, 114, 13104–13112, 10.1021/jp105042w, 2010.Iwahashi, M., Kato, T., Horiuchi, T., Sakurai, I., and Suzuki, M.: Temperature dependence of molecular conformation and liquid structure of cis-9-octadecenoic acid, J. Phys. Chem., 95, 445–451, 10.1021/j100154a078, 1991.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., Aiken, A. C., Docherty, K. S., Ulbrich, I. M., Grieshop, A. P., Robinson, A. L., Duplissy, J., Smith, J. D., Wilson, K. R., Lanz, V. A., Hueglin, C., Sun, Y. L., Tian, J., Laaksonen, A., Raatikainen, T., Rautiainen, J., Vaattovaara, P., Ehn, M., Kulmala, M., Tomlinson, J. M., Collins, D. R., Cubison, M. J., Dunlea, J., Huffman, J. A., Onasch, T. B., Alfarra, M. R., Williams, P. I., Bower, K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demerjian, K., Salcedo, D., Cottrell, L., Griffin, R., Takami, A., Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K., Kimmel, J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M., Williams, L. R., Wood, E. C., Middlebrook, A. M., Kolb, C. E., Baltensperger, U., and Worsnop, D. R.: Evolution of Organic Aerosols in the Atmosphere, Science, 326, 1525–1529, 10.1126/science.1180353, 2009.King, M. D., Thompson, K. C., and Ward, A. D.: Laser tweezers raman study of optically trapped aerosol droplets of seawater and oleic acid reacting with ozone: Implications for cloud-droplet properties, J. Am. Chem. Soc., 126, 16710–16711, 10.1021/ja044717o, 2004.King, M. D., Rennie, A. R., Thompson, K. C., Fisher, F. N., Dong, C. C., Thomas, R. K., Pfrang, C., and Hughes, A. V.: Oxidation of oleic acid at the air-water interface and its potential effects on cloud critical supersaturations, Phys. Chem. Chem. Phys., 11, 7699–7707, 10.1039/b906517b, 2009.King, M. D., Jones, S. H., Lucas, C. O. M., Thompson, K. C., Rennie, A. R., Ward, A. D., Marks, A. A., Fisher, F. N., Pfrang, C., Hughes, A. V., and Campbell, R. A.: The reaction of oleic acid monolayers with gas-phase ozone at the air water interface: The effect of sub-phase viscosity, and inert secondary components, Phys. Chem. Chem. Phys., 22, 28032–28044, 10.1039/d0cp03934a, 2020.Kulmala, M., Dada, L., Daellenbach, K. R., Yan, C., Stolzenburg, D., Kontkanen, J., Ezhova, E., Hakala, S., Tuovinen, S., Kokkonen, T. V., Kurppa, M., Cai, R., Zhou, Y., Yin, R., Baalbaki, R., Chan, T., Chu, B., Deng, C., Fu, Y., Ge, M., He, H., Heikkinen, L., Junninen, H., Liu, Y., Lu, Y., Nie, W., Rusanen, A., Vakkari, V., Wang, Y., Yang, G., Yao, L., Zheng, J., Kujansuu, J., Kangasluoma, J., Petäjä, T., Paasonen, P., Järvi, L., Worsnop, D., Ding, A., Liu, Y., Wang, L., Jiang, J., Bianchi, F., and Kerminen, V.-M.: Is reducing new particle formation a plausible solution to mitigate particulate air pollution in Beijing and other Chinese megacities?, Faraday Discuss., 226, 334–347, 10.1039/d0fd00078g, 2021.Lee, J. W. L., Carrascón, V., Gallimore, P. J., Fuller, S. J., Björkegren, A., Spring, D. R., Pope, F. D., and Kalberer, M.: The effect of humidity on the ozonolysis of unsaturated compounds in aerosol particles, Phys. Chem. Chem. Phys., 14, 8023–8031, 10.1039/c2cp24094g, 2012.Li, W., Li, H., Li, J., Cheng, X., Zhang, Z., Chai, F., Zhang, H., Yang, T., Duan, P., Lu, D., and Chen, Y.: TOF–SIMS surface analysis of chemical components of size–fractioned urban aerosols in a typical heavy air pollution event in Beijing, J. Environ. Sci., 69, 61–76, 10.1016/j.jes.2017.04.005, 2018.Lindblom, G. and Orädd, G.: NMR Studies of translational diffusion in lyotropic liquid crystals and lipid membranes, Prog. Nucl. Mag. Res. Sp., 26, 483–515, 10.1016/0079-6565(94)80014-6, 1994.Lindblom, G. and Wennerström, H.: Amphiphile diffusion in model membrane systems studied by pulsed NMR, Biophys. Chem., 6, 167–171, 10.1016/0301-4622(77)87006-3, 1977.Mauersberger, K., Barnes, J., Hanson, D., and Morton, J.: Measurement of the ozone absorption cross-section at the 253.7 nm mercury line, Geophys. Res. Lett., 13, 671–673, 10.1029/GL013i007p00671, 1986.McKay, M. D., Beckman, R. J., and Conover, W. J.: A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code, Technometrics, 21, 239, 10.2307/1268522, 1979.Mele, S., Söderman, O., Ljusberg-Wahrén, H., Thuresson, K., Monduzzi, M., and Nylander, T.: Phase behavior in the biologically important oleic acid/sodium oleate/water system, Chem. Phys. Lipids, 211, 30–36, 10.1016/j.chemphyslip.2017.11.017, 2018.Mezzenga, R., Meyer, C., Servais, C., Romoscanu, A. I., Sagalowicz, L., and Hayward, R. C.: Shear rheology of lyotropic liquid crystals: A case study, Langmuir, 21, 3322–3333, 10.1021/la046964b, 2005.Milsom, A., Squires, A. M., Boswell, J. A., Terrill, N. J., Ward, A. D., and Pfrang, C.: An organic crystalline state in ageing atmospheric aerosol proxies: spatially resolved structural changes in levitated fatty acid particles, Atmos. Chem. Phys., 21, 15003–15021, 10.5194/acp-21-15003-2021, 2021a.Milsom, A., Squires, A. M., Woden, B., Terrill, N. J., Ward, A. D., and Pfrang, C.: The persistence of a proxy for cooking emissions in megacities: a kinetic study of the ozonolysis of self-assembled films by simultaneous small and wide angle X-ray scattering (SAXS/WAXS) and Raman microscopy, Faraday Discuss., 226, 364–381, 10.1039/D0FD00088D, 2021b.Milsom, A., Lees, A., Squires, A. M., and Pfrang, C.: MultilayerPy, Zenodo [code], 10.5281/zenodo.6411189, 2022a.Milsom, A., Squires, A. M., Ward, A. D., and Pfrang, C.: Data supporting the study “The impact of molecular self-organisation on the atmospheric fate of a cooking aerosol proxy” by Milsom et al., Zenodo [data set], 10.5281/zenodo.6421940, 2022b.Molina, L. T.: Introductory lecture: air quality in megacities, Faraday Discuss., 226, 9–52, 10.1039/D0FD00123F, 2021.Morris, M. D.: Factorial Sampling Plans for Preliminary Computational Experiments, Technometrics, 33, 161–174, 10.1080/00401706.1991.10484804, 1991.Mu, Q., Shiraiwa, M., Octaviani, M., Ma, N., Ding, A., Su, H., Lammel, G., Pöschl, U., and Cheng, Y.: Temperature effect on phase state and reactivity controls atmospheric multiphase chemistry and transport of PAHs, Sci. Adv., 4, eaap7314, 10.1126/sciadv.aap7314, 2018.Ots, R., Vieno, M., Allan, J. D., Reis, S., Nemitz, E., Young, D. E., Coe, H., Di Marco, C., Detournay, A., Mackenzie, I. A., Green, D. C., and Heal, M. R.: Model simulations of cooking organic aerosol (COA) over the UK using estimates of emissions based on measurements at two sites in London, Atmos. Chem. Phys., 16, 13773–13789, 10.5194/acp-16-13773-2016, 2016.Ovadnevaite, J., Zuend, A., Laaksonen, A., Sanchez, K. J., Roberts, G., Ceburnis, D., Decesari, S., Rinaldi, M., Hodas, N., Facchini, M. C., Seinfeld, J. H., and O'Dowd, C.: Surface tension prevails over solute effect in organic-influenced cloud droplet activation, Nature, 546, 637–641, 10.1038/nature22806, 2017.Pfrang, C., Shiraiwa, M., and Pöschl, U.: Coupling aerosol surface and bulk chemistry with a kinetic double layer model (K2-SUB): oxidation of oleic acid by ozone, Atmos. Chem. Phys., 10, 4537–4557, 10.5194/acp-10-4537-2010, 2010.Pfrang, C., Shiraiwa, M., and Pöschl, U.: Chemical ageing and transformation of diffusivity in semi-solid multi-component organic aerosol particles, Atmos. Chem. Phys., 11, 7343–7354, 10.5194/acp-11-7343-2011, 2011.Pfrang, C., Rastogi, K., Cabrera-Martinez, E. R., Seddon, A. M., Dicko, C., Labrador, A., Plivelic, T. S., Cowieson, N., and Squires, A. M.: Complex three-dimensional self-assembly in proxies for atmospheric aerosols, Nat. Commun., 8, 1724, 10.1038/s41467-017-01918-1, 2017.Pöschl, U., Rudich, Y., and Ammann, M.: Kinetic model framework for aerosol and cloud surface chemistry and gas-particle interactions – Part 1: General equations, parameters, and terminology, Atmos. Chem. Phys., 7, 5989–6023, 10.5194/acp-7-5989-2007, 2007.Reid, J. P., Bertram, A. K., Topping, D. O., Laskin, A., Martin, S. T., Petters, M. D., Pope, F. D., and Rovelli, G.: The viscosity of atmospherically relevant organic particles, Nat. Commun., 9, 1–14, 10.1038/s41467-018-03027-z, 2018.Robinson, A. L., Donahue, N. M., and Rogge, W. F.: Photochemical oxidation and changes in molecular composition of organic aerosol in the regional context, J. Geophys. Res.-Atmos., 111, 1–15, 10.1029/2005JD006265, 2006.Rudich, Y., Donahue, N. M., and Mentel, T. F.: Aging of Organic Aerosol: Bridging the Gap Between Laboratory and Field Studies, Annu. Rev. Phys. Chem., 58, 321–352, 10.1146/annurev.physchem.58.032806.104432, 2007.Sagdeev, D., Gabitov, I., Isyanov, C., Khairutdinov, V., Farakhov, M., Zaripov, Z., and Abdulagatov, I.: Densities and Viscosities of Oleic Acid at Atmospheric Pressure, JAOCS, J. Am. Oil Chem. Soc., 96, 647–662, 10.1002/aocs.12217, 2019.Seddon, A. M., Richardson, S. J., Rastogi, K., Plivelic, T. S., Squires, A. M., and Pfrang, C.: Control of Nanomaterial Self-Assembly in Ultrasonically Levitated Droplets, J. Phys. Chem. Lett., 7, 1341–1345, 10.1021/acs.jpclett.6b00449, 2016.Seddon, J. M., Bartle, E. A., and Mingins, J.: Inverse cubic liquid-crystalline phases of phospholipids and related lyotropic systems, J. Phys.-Condens. Mat., 2, SA285–SA290, 10.1088/0953-8984/2/S/043, 1990.Shiraiwa, M., Pfrang, C., and Pöschl, U.: Kinetic multi-layer model of aerosol surface and bulk chemistry (KM-SUB): the influence of interfacial transport and bulk diffusion on the oxidation of oleic acid by ozone, Atmos. Chem. Phys., 10, 3673–3691, 10.5194/acp-10-3673-2010, 2010.Shiraiwa, M., Ammann, M., Koop, T., and Poschl, U.: Gas uptake and chemical aging of semisolid organic aerosol particles, P. Natl. Acad. Sci. USA, 108, 11003–11008, 10.1073/pnas.1103045108, 2011.Shrivastava, M., Lou, S., Zelenyuk, A., Easter, R. C., Corley, R. A., Thrall, B. D., Rasch, P. J., Fast, J. D., Simonich, S. L. M., Shen, H., and Tao, S.: Global long-range transport and lung cancer risk from polycyclic aromatic hydrocarbons shielded by coatings of organic aerosol, P. Natl. Acad. Sci. USA, 114, 1246–1251, 10.1073/pnas.1618475114, 2017.Slade, J. H., Ault, A. P., Bui, A. T., Ditto, J. C., Lei, Z., Bondy, A. L., Olson, N. E., Cook, R. D., Desrochers, S. J., Harvey, R. M., Erickson, M. H., Wallace, H. W., Alvarez, S. L., Flynn, J. H., Boor, B. E., Petrucci, G. A., Gentner, D. R., Griffin, R. J., and Shepson, P. B.: Bouncier Particles at Night: Biogenic Secondary Organic Aerosol Chemistry and Sulfate Drive Diel Variations in the Aerosol Phase in a Mixed Forest, Environ. Sci. Technol., 53, 4977–4987, 10.1021/acs.est.8b07319, 2019.Storn, R. and Price, K.: Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces, J. Global Optim., 11, 341–359, 10.1023/A:1008202821328, 1997.Tandon, P., Raudenkolb, S., Neibert, R. H. H., Rettig, W., and Warewig, S.: X-ray diffraction and spectroscopic studies of oleic acid–sodium oleate, Chem. Phys. Lipid, 109, 37–45, 10.1016/S0009-3084(00)00207-3, 2001.Tuckermann, R.: Surface tension of aqueous solutions of water-soluble organic and inorganic compounds, Atmos. Environ., 41, 6265–6275, 10.1016/j.atmosenv.2007.03.051, 2007.Vicente, A. M. P., Rocha, S., Duarte, M., Moreira, R., Nunes, T., and Alves, C. A.: Fingerprinting and emission rates of particulate organic compounds from typical restaurants in Portugal, Sci. Total Environ., 778, 146090, 10.1016/j.scitotenv.2021.146090, 2021.Vieceli, J., Ma, O. L., and Tobias, D. J.: Uptake and collision dynamics of gas phase ozone at unsaturated organic interfaces, J. Phys. Chem. A, 108, 5806–5814, 10.1021/jp0494584, 2004. Vignes, A.: Variation in Diffusion Coefficient with Composition, Ind. Eng. Chem. Fund., 5, 189–199, 1966.Virtanen, A., Joutsensaari, J., Koop, T., Kannosto, J., Yli-Pirilä, P., Leskinen, J., Mäkelä, J. M., Holopainen, J. K., Pöschl, U., Kulmala, M., Worsnop, D. R., and Laaksonen, A.: An amorphous solid state of biogenic secondary organic aerosol particles, Nature, 467, 824–827, 10.1038/nature09455, 2010.Virtanen, P., Gommers, R., Oliphant, T. E., Haberland, M., Reddy, T., Cournapeau, D., Burovski, E., Peterson, P., Weckesser, W., Bright, J., van der Walt, S. J., Brett, M., Wilson, J., Millman, K. J., Mayorov, N., Nelson, A. R. J., Jones, E., Kern, R., Larson, E., Carey, C. J., Polat, İ., Feng, Y., Moore, E. W., VanderPlas, J., Laxalde, D., Perktold, J., Cimrman, R., Henriksen, I., Quintero, E. A., Harris, C. R., Archibald, A. M., Ribeiro, A. H., Pedregosa, F., van Mulbregt, P., Vijaykumar, A., Bardelli, A. P., Rothberg, A., Hilboll, A., Kloeckner, A., Scopatz, A., Lee, A., Rokem, A., Woods, C. N., Fulton, C., Masson, C., Häggström, C., Fitzgerald, C., Nicholson, D. A., Hagen, D. R., Pasechnik, D. V., Olivetti, E., Martin, E., Wieser, E., Silva, F., Lenders, F., Wilhelm, F., Young, G., Price, G. A., Ingold, G. L., Allen, G. E., Lee, G. R., Audren, H., Probst, I., Dietrich, J. P., Silterra, J., Webber, J. T., Slavič, J., Nothman, J., Buchner, J., Kulick, J., Schönberger, J. L., de Miranda Cardoso, J. V., Reimer, J., Harrington, J., Rodríguez, J. L. C., Nunez-Iglesias, J., Kuczynski, J., Tritz, K., Thoma, M., Newville, M., Kümmerer, M., Bolingbroke, M., Tartre, M., Pak, M., Smith, N. J., Nowaczyk, N., Shebanov, N., Pavlyk, O., Brodtkorb, P. A., Lee, P., McGibbon, R. T., Feldbauer, R., Lewis, S., Tygier, S., Sievert, S., Vigna, S., Peterson, S., More, S., Pudlik, T., et al.: SciPy 1.0: fundamental algorithms for scientific computing in Python, Nat. Methods, 17, 261–272, 10.1038/s41592-019-0686-2, 2020.Wang, Q. and Yu, J. Z.: Ambient Measurements of Heterogeneous Ozone Oxidation Rates of Oleic, Elaidic, and Linoleic Acid Using a Relative Rate Constant Approach in an Urban Environment, Geophys. Res. Lett., 48, e2021GL095130, 10.1029/2021GL095130, 2021.Wang, Q., He, X., Zhou, M., Huang, D. D., Qiao, L., Zhu, S., Ma, Y. G., Wang, H. L., Li, L., Huang, C., Huang, X. H. H., Xu, W., Worsnop, D., Goldstein, A. H., Guo, H., Yu, J. Z., Huang, C., and Yu, J. Z.: Hourly Measurements of Organic Molecular Markers in Urban Shanghai, China: Primary Organic Aerosol Source Identification and Observation of Cooking Aerosol Aging, ACS Earth Sp. Chem., 4, 1670–1685, 10.1021/acsearthspacechem.0c00205, 2020.Wang, T., Huang, R. J., Li, Y., Chen, Q., Chen, Y., Yang, L., Guo, J., Ni, H., Hoffmann, T., Wang, X., and Mai, B.: One-year characterization of organic aerosol markers in urban Beijing: Seasonal variation and spatiotemporal comparison, Sci. Total Environ., 743, 140689, 10.1016/j.scitotenv.2020.140689, 2020.Weschler, C. J.: Ozone in indoor environments: Concentration and chemistry, Indoor Air, 10, 269–288, 10.1034/j.1600-0668.2000.010004269.x, 2000.Woden, B., Skoda, M. W. A., Milsom, A., Gubb, C., Maestro, A., Tellam, J., and Pfrang, C.: Ozonolysis of fatty acid monolayers at the air–water interface: organic films may persist at the surface of atmospheric aerosols, Atmos. Chem. Phys., 21, 1325–1340, 10.5194/acp-21-1325-2021, 2021.Worsnop, D. R., Morris, J. W., Shi, Q., Davidovits, P., and Kolb, C. E.: A chemical kinetic model for reactive transformations of aerosol particles, Geophys. Res. Lett., 29, 19–22, 10.1029/2002GL015542, 2002.Wu, Z. J., Zheng, J., Shang, D. J., Du, Z. F., Wu, Y. S., Zeng, L. M., Wiedensohler, A., and Hu, M.: Particle hygroscopicity and its link to chemical composition in the urban atmosphere of Beijing, China, during summertime, Atmos. Chem. Phys., 16, 1123–1138, 10.5194/acp-16-1123-2016, 2016.
Zabara, A. and Mezzenga, R.: Controlling molecular transport and sustained drug release in lipid-based liquid crystalline mesophases, J. Control. Release, 188, 31–43, 10.1016/j.jconrel.2014.05.052, 2014.Zahardis, J. and Petrucci, G. A.: The oleic acid-ozone heterogeneous reaction system: products, kinetics, secondary chemistry, and atmospheric implications of a model system – a review, Atmos. Chem. Phys., 7, 1237–1274, 10.5194/acp-7-1237-2007, 2007.Zahardis, J., LaFranchi, B. W., and Petrucci, G. A.: Direct observation of polymerization in the oleic acid-ozone heterogeneous reaction system by photoelectron resonance capture ionization aerosol mass spectrometry, Atmos. Environ., 40, 1661–1670, 10.1016/j.atmosenv.2005.10.065, 2006.Zhao, X., Hu, Q., Wang, X., Ding, X., He, Q., Zhang, Z., Shen, R., Lü, S., Liu, T., Fu, X., and Chen, L.: Composition profiles of organic aerosols from Chinese residential cooking: Case study in urban Guangzhou, south China, J. Atmos. Chem., 72, 1–18, 10.1007/s10874-015-9298-0, 2015.Zhou, S., Shiraiwa, M., McWhinney, R. D., Pöschl, U., and Abbatt, J. P. D.: Kinetic limitations in gas-particle reactions arising from slow diffusion in secondary organic aerosol, Faraday Discuss., 165, 391–406, 10.1039/c3fd00030c, 2013.Zhou, S., Hwang, B. C. H., Lakey, P. S. J., Zuend, A., Abbatt, J. P. D., and Shiraiwa, M.: Multiphase reactivity of polycyclic aromatic hydrocarbons is driven by phase separation and diffusion limitations, P. Natl. Acad. Sci. USA, 116, 11658–11663, 10.1073/pnas.1902517116, 2019.Zhou, Z., Zhou, S., and Abbatt, J. P. D.: Kinetics and Condensed-Phase Products in Multiphase Ozonolysis of an Unsaturated Triglyceride, Environ. Sci. Technol., 53, 12467–12475, 10.1021/acs.est.9b04460, 2019.