eter Particles Directly Probed Using Atomic Force ... - ACS Publications

(13) Cappa, C. D.; Wilson, K. R. Atmos Chem Phys 2011, 11, 1895-1911. (14) Calvo ... (24) Murray, B. J.; Wilson, T. W.; Dobbie, S.; Cui, Z. Q.; Al-Jum...
0 downloads 0 Views 1MB Size
Article Cite This: Anal. Chem. XXXX, XXX, XXX-XXX

pubs.acs.org/ac

Solid, Semisolid, and Liquid Phase States of Individual Submicrometer Particles Directly Probed Using Atomic Force Microscopy Hansol D. Lee,† Kamal K. Ray,† and Alexei V. Tivanski*,† †

Department of Chemistry, University of Iowa, Iowa City, Iowa 52242, United States S Supporting Information *

ABSTRACT: Currently, the impact of various phase states of aerosols on the climate is not well understood, especially for submicrometer sized aerosol particles that typically have extended lifetime in the atmosphere. This is largely due to the inherent size limitations present in current experimental techniques that aim to directly assess the phase states of fine aerosol particles. Herein we present a technique that uses atomic force microscopy to probe directly for the phase states of individual, submicrometer particles by using nanoindentation and nano-Wilhelmy methodologies as a function of relative humidity (RH) and ambient temperature conditions. When using these methodologies for substrate deposited individual sucrose particles, Young’s modulus and surface tension can be quantified as a function of RH. We show that the force profiles collected to measure Young’s modulus and surface tension can also provide both qualitative and quantitative assessments of phase states that accompany solid, semisolid, and liquid particle phases. Specifically, we introduce direct measurements of relative indentation depth and viscoelastic response distance on a single particle basis at a given applied force to quantitatively probe for the phase state as a function of RH and corresponding viscosity. Thus, we show that the three phase states and phase state transitions of sucrose can be identified and ultimately propose that this technique may also be used to study other atmospherically relevant systems.

D

or ice nucleating particles, and regulate the particle reactivity with gas phase precursors.4,20,26 Therefore, it is important to accurately determine the phase state of the aerosol particle as a function of RH and viscosity. This is particularly significant for submicrometer sized systems, because both primary and secondary aerosols can exist as submicrometer sized systems with much greater lifetime in the atmosphere than supermicrometer sized aerosols.29 However, due to their small size and the inherent size limitation of experimental approaches, techniques that can directly probe for phase states have not been shown until now. Experimental approaches that indirectly measure the phase state or probe supermicrometer sized systems already exist and show promising results. The new electrical low pressure impactor (ELPI) method was used to distinguish between liquid and nonliquid particles. By generating a size distribution of aerosols and impacting them onto the substrate to measure the bounce factor, the phase state of the impacted particles can be implied.30,31 Furthermore, to measure the viscosity, or diffusion coefficient, and infer phase state, several studies on macroscopic particles have been performed. The bead-mobility and poke-flow experiments were carried out to determine the

epending on the temperature and pressure conditions, chemically identical systems can engender different physical phase states. This is evident in atmospheric aerosols, where relative humidity (RH) can strongly influence the phase state of the aerosol particle.1−5 For example, atmospheric saccharides that contribute significantly to the overall mass fraction of the aerosol can access wide ranges of viscosities resulting in metastable states.6−10 Several recent studies clearly showed that aerosols can be found at different physical states depending on surrounding RH, temperature, and source.11−13 Two types of aerosols are distinguished as primary and secondary aerosols. Primary aerosols are produced from various natural and anthropogenic sources, with examples including mineral dust, volcanic ash, soot, and sea spray aerosol.14,15 Secondary aerosols are formed from volatile precursors that underwent heterogeneous chemical reactions that determine their size, shape, and chemical composition.16 A well-known example is secondary organic aerosol (SOA).1,3,11,12,17−19 Primary and secondary aerosol particles containing highly viscous saccharides have strong implications in the overall climate and atmospheric processes by playing a dominant role in radiative forcing through hygroscopicity, ice nucleation, and heterogeneous reactivity.18,20−28 More specifically, it has been shown that particle phase state can strongly influence the overall water uptake behavior of the aerosol particle, dictate whether aerosol particles will act as cloud condensation nuclei © XXXX American Chemical Society

Received: July 14, 2017 Accepted: November 14, 2017 Published: November 14, 2017 A

DOI: 10.1021/acs.analchem.7b02755 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

was approximately 10−15 min, to ensure that the particle reached thermodynamic equilibrium in the humidity cell. From imaging the hydration growth of sucrose particles, volumeequivalent growth factor (GF) versus RH data were generated using methodology published previously.41 GF data are reported as the mean value, and the magnitude of the error bar represents two standard deviations. AFM Force Spectroscopy: Nanoindentation. Silicon nitride AFM probes (MikroMasch, Model CSC37) that were used for imaging were also used for nanoindentation. Scan rate was 0.8 Hz for all force measurements, unless noted otherwise. The humidity in the AFM chamber was increased from 5% to 75% RH, i.e., in hydration mode. Before each experiment was started at low RH, the sample was kept at 5% RH for approximately 1 h. The equilibration period was approximately 10 min between changes in RH. For the Young’s modulus measurements, nanoindentation force measurements were performed approximately 200 times over different individual particles per RH value with the maximum applied force of 10 nN. From the force profiles collected in this study, the Johnson−Kendall−Roberts (JKR) model was used to calculate YM (eqs S1−S5), with Poisson’s ratio of 0.25.35,42−44 The YM values at each RH were compiled as a histogram, and the Gaussian fit was performed to determine the mean value and standard deviation (Figure S1). The error bars are reported as two standard deviations. For the viscoelastic response distance (VRD) and relative indentation depth (RID) measurements, nanoindentation force measurements were repeated approximately 5 times per each RH value with 10 nN as the maximum force applied to the sample, unless noted otherwise. AFM Force Spectroscopy: Nano-Wilhelmy. High aspect ratio, constant diameter Ag2Ga nanoneedles (NN-HAR-FM60, NaugaNeedles) with a nominal spring constant of 2.7−3.3 N/ m and diameter of 50−200 nm were used for surface tension measurements. Overall AFM methodology was explained in detail previously.36,45 In consideration of the high viscosity of sucrose, surface tension measurements were done in dehydration mode, where the RH was initially increased to the maximum value (∼94% RH) and then slowly decreased. At least three force plots were collected per RH value. After every RH change, each surface tension measurement spanned approximately 10−15 min of equilibration time. The effective radius of the tip was quantified by force calibration before and after single particle analysis by measuring retention force from a bulk liquid droplet with well-known surface tension, clean vacuum oil (DirecTorr, CAS No. 64742-65-0). If the force profile showed drastic changes in the tip radius before and after single particle analysis, then the data was discarded and the experiment was repeated with a new nanoneedle. AFM surface tension data are reported as the mean value, and the magnitudes of the error bars are reported as two standard deviations. Surface tension results were compared to bulk surface tension measurements using a Kibron AquaPi force tensiometer (Kibron, Finland). To predict the bulk surface tension at high concentrations, a predictive model of surface tension versus concentration obtained from bulk trend line fit was used.36

viscosity of substrate deposited particles that are supermicrometer in size and subsequently made strong implications toward the phase state.3,32 The optical tweezer method was used to coalesce two particles of the same chemical components and observe the oscillations, which allows for calculation of the viscosity as a function of RH.7,33 The same method was recently used in conjunction with theoretical simulation to determine the diffusion coefficient of supermicrometer sized particles from the whispering gallery modes.34 But none have directly probed for the phase state of submicrometer sized systems. In this work, we present a new atomic force microscopy (AFM) technique that can directly and quantitatively identify the phase state of substrate deposited particles on a submicrometer size scale. Capability to directly measure the phase state is highly advantageous since it does not require supplementary modeling, which can sometimes be prone to inherent uncertainties due to assumptions used. Specifically, we combine the nanoindentation and nano-Wilhelmy methodologies to perform Young’s modulus (YM) and surface tension measurements over an individual particle as a function of RH.35,36 For this work, we studied single particle sucrose as an ideal model system for three reasons: it has access to wide ranges of viscosity and phases in subsaturated RH ( 40% the particle response becomes dominated by capillary forces and the model can no longer be used. Conversely, as RH decreases from ca. 94%, the particle loses water, and at RH< 63% it becomes a semisolid until the particle response becomes dominated by viscosity and eq 1 for surface tension can no longer be used.

qualitatively indicates that the particle is too soft and nanoindentation response can no longer be described with the JKR method. Thus, the nanoindentation methodology is not suitable for determining YM at the RH values above ∼40% for sucrose, which corresponds to below ∼106.3 Pa s in viscosity, and the particle phase state that is now closer to that of a liquid rather than a solid. When the water uptake dictates the particle phase state to be closer to that of a liquid rather than a solid, the nano-Wilhelmy method can now be used instead to measure surface tension at higher RH. The methodology relies on forming a stable capillary between the constant-radius nanoneedle and the particle, which is subsequently stretched and pulled away from the surface until the meniscus breaks off. The relation is the following36,45

Figure 2. Force versus tip−sample separation plots for selected RH ranging from 7% to 82% of a sucrose particle. Tip sample separation of 0 indicates the initial position of AFM tip contact with the surface of the particle. The evolution of the force profiles from low to high RH is shown from left to right, using x-offset. Red lines indicate approach to the sucrose particle, and blue lines retract away from the particle. The purple lines indicate the JKR model fit in the contact region. (A) Force profiles at 7%, 19%, and 40% RH collected at the maximum applied force of 10 nN. The corresponding viscoelastic response distance (VRD) is specified below. (B) Force profiles at 48%, 54%, and 61% RH collected at the maximum applied force of 10 nN. The corresponding indentation depth (I) is specified below. (C) Force profiles at 58%, 64%, and 82% RH collected at the maximum applied force of 1 nN. The retention force (Fret) used to calculate surface tension is specified below.

Fret = 2πrσ

(1)

where Fret is retention force, r is radius of the nanoneedle, and σ is surface tension of a droplet at a given RH. Previous publications have developed and systematically studied the application of this methodology to a wide range of systems, including atmospherically relevant electrolyte salts, dicarboxylic acids, and saccharides.36,45 Similarly, force profiles in Figure 2C were used to calculate the surface tension from the measured retention force. Noteworthy, although RH decreases from 82% to 64% RH yields nearly identical force profiles, the decrease from 64% to 58% RH shows significant changes to the force profiles. Specifically, the force profile shown at 58% RH is qualitatively similar to the force profiles collected on a solid surface with thin liquid layer, where the capillary force from the formation of a meniscus on the nanoneedle contributes to the retention force. This likely indicates that the particle at 58% RH is too viscous. This result is shown in Figure 3, where single particle surface tension measurements are shown in orange circles with error bars, indicating two standard deviations. The

viscosity for sucrose was formulated from a previous publication that quantified viscosity with optical tweezers, in the viscosity range from 1016 to 10−3 Pa s.37 As RH increases, the sucrose concentration decreases with corresponding increase in the total mass of water absorbed, which in turn decreases the YM as the particle becomes progressively more hydrated and thus softer. We note that measurements of YM beyond 40% RH were not incorporated into the YM calculations. This is due to highly viscous nature of the response observed for the approach curve in the contact region and the lack of good overlap to the JKR fit (Figure 2B, unsuccessful fits using JKR model are not shown). This D

DOI: 10.1021/acs.analchem.7b02755 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

curve shows that the probe indents completely through the particle reaching the underlying substrate with the slope approaching the perpendicular, thus indicative of a phase transition from semisolid to liquid. Therefore, we can tentatively predict the transition to be at around 60% RH, which would equate to viscosity of 102.5 Pa s. In fact, this agrees reasonably well with 102 Pa s that is currently accepted.7,11,50 Beyond 60% RH, force profiles do not show any significant changes, even beyond the saturation point of sucrose, which is approximately 88% RH.36 In order to quantitatively measure the semisolid to liquid phase transition, we propose to define and measure relative indentation depth (RID), a unitless value, as a function of RH. RID at a particular RH is defined as a ratio of indentation depth and height of the particle

orange dashed line indicates surface tension prediction from bulk solution surface tension measurements. Above 63% RH, there is a good overlap between the AFM and bulk surface tension data. However, in agreement with significant changes in the force profiles, strong deviation in single particle surface tension compared to extrapolation from bulk surface tension is seen below 63% RH. This result implies that when the particle’s phase state is not liquid, capillary force is no longer the sole contributor to the measured retention force; thus, significant contribution to the measured retention force stems from the particle viscosity, which further supports our assessment of the particle phase transition that occurs within the two RH regions. Therefore, the nano-Wilhelmy methodology is not suitable for use below 63% RH for sucrose, which corresponds to above ∼102 Pa s in viscosity, and the particle phase state that is closer to that of a semisolid rather than a liquid. Qualitative and Quantitative Assessment of the Particle Phase States. The identification of experimental limitations shown above can be extended to make both qualitative and quantitative assessments of the particle phase states at varying RH, from solid to semisolid to liquid. We note that the phase transitions for amorphous solids such as sucrose should not be as discrete and sharp as for crystalline solids; instead, continuous and smooth transitions are expected. Here, we elaborate on the factors considered to identify the phase state transition points. The transition between solid and semisolid phase state was determined by observing the presence of viscoelastic response in the force profile (Figure 2A) from 7% to 19% RH. This viscoelastic phenomenon indicates that the sucrose particle has undergone transition from purely elastic to viscoelastic material. For quantitative assessment of the viscoelastic response distance (VRD) at varying RH, we propose to measure the hysteresis in the distance from the approach to retract curves. By normalizing to the 0 nN force position on the profile, the difference in distance between the approach and retract can be determined and then correlated to the extent of viscoelastic response. In order to ensure that the level of noise in our experiments does not affect the measured VRD at low RH, we incorporated a smooth function for all force profiles (see the Supporting Information for details). Results indicate that at 7% RH, the VRD is 0.2 ± 0.1 nm and shows relatively small increase to 0.4 ± 0.3 nm at 15% RH. At 19% RH, the VRD is 0.6 ± 0.2 nm, which is statistically significant in comparison to the level of noise in our distance measurements (ca. 0.2 nm). Subsequent increase in RH shows apparent exponential increase in VRD (Figure S2). Moreover, systematically varying the scan rate (0.4, 0.8, and 1.6 Hz) in our VRD measurements does not significantly change the results for RH below 25% (Figure S3A). Thus, we designate the phase transition between solid and semisolid to be at ∼18% RH, which corresponds to 1011.2 Pa s. Our assessment agrees well with the currently accepted value of 1012 Pa s from the literature,7,11,50 confirming applicability of the method to directly determine the change in the phase state. The transition from semisolid to liquid can be qualitatively determined by observing change in the indentation depth as a function of RH. Force profiles show evolution from 54% to 61% RH to be unique in this regard (Figure 2B); at 54% RH, the approach curve in the contact region does not indent through the particle completely as seen by the negative shallow slope with continued decrease in the separation distance as force increases. However, at elevated RH of 61%, the approach

RID =

I H

(2)

where I is the indentation depth which represents total amount of travel at the apex of the AFM probe in the z-direction from the surface of the particle to the maximum depth indented into the particle and H is the height of the particle relative to the substrate. At low RH ranges, the particle is dehydrated and stiff (high Young’s modulus) due to high viscosity, and the indentation depth is low. However, as the particle uptakes more water, it becomes softer and the same applied maximum force allows the AFM probe to indent further into the particle, shown in Figure 2B where the indentation was ∼0.36 μm at 54% RH and ∼0.40 μm at 61% RH. In Figure 4A, we show the combination of both VRD and RID versus RH and corresponding viscosity. Blue circles and red triangles indicate average VRD and RID measurements with error bars, respectively, collected with force plots using 10 nN maximum applied force. The dotted line for RID data is fit using a fourparameter logistic sigmoidal function and is shown for clarity only. Due to high viscosity and subsequently high Young’s modulus, from 7% to 34% RH, the corresponding RID is small at ∼0.04 and does not significantly increase until 44% RH. This shows that the particle is relatively stiff, consistent with the YM measurements shown in Figure 3, and the AFM probe cannot make appreciable amount of indentation compared to the height of the particle. However, beyond 44% RH, the viscosity becomes low enough that the AFM probe starts to indent considerably relative to the height of the particle. The RID at 47% RH is approximately 0.18, indicating that the material is becoming soft leading up to the transition point at 60% RH, at which point the RID is 0.98 and the particle behaves more like liquid. This transition is not discrete but rather continuous, as expected for an amorphous sucrose particle. For RH > 60%, RID reaches the maximum value of 1, which means that the probe indents completely through the particle and further increase in RH will decrease the viscosity, but the RID will remain at 1. To test whether experimental parameters affect the measured RID, both applied force and scan rate were systematically varied from 2 to 20 nN (Figure 4B) and 0.4 to 1.6 Hz (Figure S3B), respectively. Dotted lines were fitted using a four-parameter logistic sigmoidal function when increasing the maximum applied force and are shown for clarity. Initially, applying 2 nN of maximum force onto the sucrose particle to indent is shown to underestimate the RID even up to 73% RH. However, further increasing the applied maximum force to 5 nN and as high as 20 nN on the same particle shows systematically E

DOI: 10.1021/acs.analchem.7b02755 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

states of sucrose were directly probed: solid, semisolid, and liquid. The quantitative measurements of viscoelastic response distance (VRD) show that solid to semisolid transition occurs at ca. 18% RH (viscosity 1011.2 Pa s). The relative indentation depth (RID) measurements show that semisolid to liquid transition occurs at ca. 60% RH (viscosity 102.5 Pa s). Corresponding phase state transitions and RH are in good agreement with the literature accepted values, thus validating the technique. This approach can be readily used to directly determine the particle phase states of other atmospherically or environmentally relevant chemical systems, by utilizing an imaging AFM probe and measuring the VRD and RID at an appropriate maximum applied force as a function of RH. Both laboratory generated model systems as well as nascent aerosol particles collected on a substrate may be characterized using this technique. Our established work outlined above can contribute to better understanding the role of various phase states available to viscous aerosols and their impact on the climate and the atmospheric processing.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b02755. Young’s modulus histograms based on Johnson− Kendall−Roberts fit; quantification of viscoelastic response distance with respect to relative humidity; role of scan rate in viscoelastic response distance and relative indentation depth measurements (PDF)

Figure 4. (A) AFM viscoelastic response distance (VRD, left, blue circles) and relative indentation depth (RID, right, red triangles) versus RH and corresponding viscosity collected with 10 nN of maximum applied force. The error bars represent two standard deviations for the VRD and RID, although sometimes smaller than the symbol. The RH−viscosity relationship is taken from Song et al. The red dotted line is shown for clarity and represents the fit using a fourparameter logistic sigmoidal function. Color bars indicate the various phase states of the sucrose particle at given RH from solid, semisolid, and liquid. The dashed black lines represent phase transition points determined by the force profiles of sucrose. The phase transition from solid to semisolid states occurred at ∼18% RH and 1011.2 Pa s, while the semisolid to liquid state transition occurred at ∼60% RH and 102.5 Pa s. (B) RID versus RH and corresponding viscosity from 25% to 70% RH with varying maximum applied forces from 2 to 20 nN. The colored dotted lines are shown for clarity and represent the fit using a four-parameter logistic sigmoidal function.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Alexei V. Tivanski: 0000-0002-1528-2421 Author Contributions

VRD and RID experiments were performed by H.D.L., and YM and water uptake experiments were performed by K.K.R. The data analysis, interpretation, and writing were performed through contribution of all authors. All authors have given approval to the final version of the manuscript. Notes

increasing RID values over similar RH below ca. 60% but which still converge to the value of 1 after a phase transition at 60% RH, in very good agreement with our proposed phase transition point from semisolid to liquid for sucrose. Systematically varying the scan rate for RID measurements does not show any significant dependence, except at 45−55% RH range where the slowest scan rate of 0.4 Hz produced the highest RID values (Figure S3B).

The authors declare no competing financial interest.

CONCLUSIONS In this work, new AFM technique was developed to directly probe for the phase states of substrate deposited submicrometer sucrose particles. The technique combines complementary nanoindentation and nano-Wilhelmy methodologies to collect force profiles over an individual particle at different RH that revealed distinct changes in phase states within viscosity ranges from 1016 to 10−3 Pa s. By varying RH, three different phase





ACKNOWLEDGMENTS This work was funded by the National Science Foundation through the Center for Aerosol Impacts on Chemistry of the Environment under grant no. CHE 1305427. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.



REFERENCES

(1) 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.; Worsnop, D. R.; Virtanen, A. Atmos. Chem. Phys. 2012, 12, 7517−7529. (2) Wang, B. B.; O’Brien, R. E.; Kelly, S. T.; Shilling, J. E.; Moffet, R. C.; Gilles, M. K.; Laskin, A. J. Phys. Chem. A 2015, 119, 4498−4508.

F

DOI: 10.1021/acs.analchem.7b02755 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

(32) Renbaum-Wolff, L.; Grayson, J. W.; Bertram, A. K. Atmos. Chem. Phys. 2013, 13, 791−802. (33) Power, R.; Bones, D. L.; Reid, J. P. Proc. SPIE 2012, 8458, 845829-10. (34) Preston, T. C.; Davies, J. F.; Wilson, K. R. Phys. Chem. Chem. Phys. 2017, 19, 3922−3931. (35) Mohapatra, H.; Kruger, T. M.; Lansakara, T. I.; Tivanski, A. V.; Stevens, L. L. Soft Matter 2017, 13, 5684−5695. (36) Lee, H. D.; Estillore, A. D.; Morris, H. S.; Ray, K. K.; Alejandro, A.; Grassian, V. H.; Tivanski, A. V. J. Phys. Chem. A 2017, 121, 8296− 8305. (37) Song, Y. C.; Haddrell, A. E.; Bzdek, B. R.; Reid, J. P.; Bannan, T.; Topping, D. O.; Percival, C.; Cai, C. J. Phys. Chem. A 2016, 120, 8123−8137. (38) Bones, D. L.; Reid, J. P.; Lienhard, D. M.; Krieger, U. K. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 11613−11618. (39) Baustian, K. J.; Wise, M. E.; Tolbert, M. A. Atmos. Chem. Phys. 2010, 10, 2307−2317. (40) Baltrusaitis, J.; Grassian, V. H. J. Phys. Chem. A 2012, 116, 9001−9009. (41) Morris, H. S.; Estillore, A. D.; Laskina, O.; Grassian, V. H.; Tivanski, A. V. Anal. Chem. 2016, 88, 3647−3654. (42) Dokukin, M. E.; Sokolov, I. Langmuir 2012, 28, 16060−16071. (43) Dokukin, M. E.; Sokolov, I. Macromolecules 2012, 45, 4277− 4288. (44) Prokopovich, P.; Perni, S. Colloids Surf., A 2011, 383, 95−101. (45) Morris, H. S.; Grassian, V. H.; Tivanski, A. V. Chemical Science 2015, 6, 6021−6021. (46) Karunatilaka, C.; Bucar, D. K.; Ditzler, L. R.; Friscic, T.; Swenson, D. C.; MacGillivray, L. R.; Tivanski, A. V. Angew. Chem., Int. Ed. 2011, 50, 8642−8646. (47) Rupasinghe, T. P.; Hutchins, K. M.; Bandaranayake, B. S.; Ghorai, S.; Karunatilake, C.; Bucar, D. K.; Swenson, D. C.; Arnold, M. A.; MacGillivray, L. R.; Tivanski, A. V. J. Am. Chem. Soc. 2015, 137, 12768−12771. (48) Morris, H.; Grassian, V.; Tivanski, A. Chemical Science 2015, 6, 3242−3247. (49) Johnson, K. L.; Kendall, K.; Roberts, A. D. Proc. R. Soc. London, Ser. A 1971, 324, 301−313. (50) Power, R. M.; Simpson, S. H.; Reid, J. P.; Hudson, A. J. Chemical Science 2013, 4, 2597−2604.

(3) Song, M. J.; Liu, P. F. F.; Hanna, S. J.; Zaveri, R. A.; Potter, K.; You, Y.; Martin, S. T.; Bertram, A. K. Atmos. Chem. Phys. 2016, 16, 8817−8830. (4) Hodas, N.; Zuend, A.; Mui, W.; Flagan, R. C.; Seinfeld, J. H. Atmos. Chem. Phys. 2015, 15, 5027−5045. (5) Tong, H. J.; Reid, J. P.; Bones, D. L.; Luo, B. P.; Krieger, U. K. Atmos. Chem. Phys. 2011, 11, 4739−4754. (6) Jayarathne, T.; Sultana, C. M.; Lee, C.; Malfatti, F.; Cox, J. L.; Pendergraft, M. A.; Moore, K. A.; Azam, F.; Tivanski, A. V.; Cappa, C. D.; Bertram, T. H.; Grassian, V. H.; Prather, K. A.; Stone, E. A. Environ. Sci. Technol. 2016, 50, 11511−11520. (7) Angell, C. A. Science 1995, 267, 1924−1935. (8) Rickards, A. M. J.; Song, Y. C.; Miles, R. E. H.; Preston, T. C.; Reid, J. P. Phys. Chem. Chem. Phys. 2015, 17, 10059−10073. (9) Reinhardt, A.; Emmenegger, C.; Gerrits, B.; Panse, C.; Dommen, J.; Baltensperger, U.; Zenobi, R.; Kalberer, M. Anal. Chem. 2007, 79, 4074−4082. (10) Lu, J. W.; Rickards, A. M. J.; Walker, J. S.; Knox, K. J.; Miles, R. E. H.; Reid, J. P.; Signorell, R. Phys. Chem. Chem. Phys. 2014, 16, 9819−9830. (11) Koop, T.; Bookhold, J.; Shiraiwa, M.; Poschl, U. Phys. Chem. Chem. Phys. 2011, 13, 19238−19255. (12) Virtanen, A.; Joutsensaari, J.; Koop, T.; Kannosto, J.; Yli-Pirila, P.; Leskinen, J.; Makela, J. M.; Holopainen, J. K.; Poschl, U.; Kulmala, M.; Worsnop, D. R.; Laaksonen, A. Nature 2010, 467, 824−827. (13) Cappa, C. D.; Wilson, K. R. Atmos. Chem. Phys. 2011, 11, 1895− 1911. (14) Calvo, A. I.; Alves, C.; Castro, A.; Pont, V.; Vicente, A. M.; Fraile, R. Atmos. Res. 2013, 120−121, 1−28. (15) Fuzzi, S.; Baltensperger, U.; Carslaw, K.; Decesari, S.; Denier van der Gon, H.; Facchini, M. C.; Fowler, D.; Koren, I.; Langford, B.; Lohmann, U.; Nemitz, E.; Pandis, S.; Riipinen, I.; Rudich, Y.; Schaap, M.; Slowik, J. G.; Spracklen, D. V.; Vignati, E.; Wild, M.; Williams, M.; et al. Atmos. Chem. Phys. 2015, 15, 8217−8299. (16) Kroll, J. H.; Seinfeld, J. H. Atmos. Environ. 2008, 42, 3593−3624. (17) Baustian, K. J.; Wise, M. E.; Jensen, E. J.; Schill, G. P.; Freedman, M. A.; Tolbert, M. A. Atmos. Chem. Phys. 2013, 13, 5615− 5628. (18) 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.; Pankow, J. F.; Finlayson-Pitts, B. J. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 2836−2841. (19) Schill, G. P.; De Haan, D. O.; Tolbert, M. A. Environ. Sci. Technol. 2014, 48, 1675−1682. (20) Kuwata, M.; Martin, S. T. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 17354−17359. (21) Renbaum-Wolff, L.; Grayson, J. W.; Bateman, A. P.; Kuwata, M.; Sellier, M.; Murray, B. J.; Shilling, J. E.; Martin, S. T.; Bertram, A. K. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 8014−8019. (22) Zobrist, B.; Marcolli, C.; Pedernera, D. A.; Koop, T. Atmos. Chem. Phys. 2008, 8, 5221−5244. (23) Murray, B. J. Atmos. Chem. Phys. 2008, 8, 5423−5433. (24) Murray, B. J.; Wilson, T. W.; Dobbie, S.; Cui, Z. Q.; Al-Jumur, S. M. R. K.; Mohler, O.; Schnaiter, M.; Wagner, R.; Benz, S.; Niemand, M.; Saathoff, H.; Ebert, V.; Wagner, S.; Karcher, B. Nat. Geosci. 2010, 3, 233−237. (25) Mikhailov, E.; Vlasenko, S.; Martin, S. T.; Koop, T.; Poschl, U. Atmos. Chem. Phys. 2009, 9, 9491−9522. (26) Shiraiwa, M.; Ammann, M.; Koop, T.; Poschl, U. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 11003−11008. (27) Shiraiwa, M.; Berkemeier, T.; Schilling-Fahnestock, K. A.; Seinfeld, J. H.; Poschl, U. Atmos. Chem. Phys. 2014, 14, 8323−8341. (28) Berkemeier, T.; Huisman, A. J.; Ammann, M.; Shiraiwa, M.; Koop, T.; Poschl, U. Atmos. Chem. Phys. 2013, 13, 6663−6686. (29) Gong, S. L.; Barrie, L. A.; Blanchet, J. P. J. Geophys Res-Atmos 1997, 102, 3805−3818. (30) Jain, S.; Petrucci, G. A. Aerosol Sci. Technol. 2015, 49, 390−399. (31) Bateman, A. P.; Belassein, H.; Martin, S. T. Aerosol Sci. Technol. 2014, 48, 42−52. G

DOI: 10.1021/acs.analchem.7b02755 Anal. Chem. XXXX, XXX, XXX−XXX