Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Mar 1.
Published in final edited form as: J Am Soc Mass Spectrom. 2016 Jan 4;27(3):402–409. doi: 10.1007/s13361-015-1300-1

Molecular Characterization of Secondary Aerosol from Oxidation of Cyclic Methylsiloxanes

Yue Wu 1, Murray V Johnston 1,*
PMCID: PMC4929013  NIHMSID: NIHMS797501  PMID: 26729452

Abstract

Cyclic volatile methylsiloxanes (cVMS) have been identified as important gas-phase atmospheric contaminants, but knowledge of the molecular composition of secondary aerosol derived from cVMS oxidation is incomplete. Here, the chemical composition of secondary aerosol produced from the OH-initiated oxidation of decamethylcyclopentasiloxane (D5, C10H30O5Si5) is characterized by high performance mass spectrometry. ESI-MS reveals a large number of monomeric (300 < m/z < 470) and dimeric (700 < m/z < 870) oxidation products. With the aid of high resolution and MS/MS, it is shown that oxidation leads mainly to the substitution of a CH3 group by OH or CH2OH, and that a single molecule can undergo many CH3 group substitutions. Dimers also exhibit OH and CH2OH substitutions and can be linked by O, CH2 and CH2CH2 groups. GC-MS confirms the ESI-MS results. Oxidation of D4 (C8H24O4Si4) exhibits similar substitutions and oligomerizations to D5, though the degree of oxidation is greater under the same conditions and there is direct evidence for the formation of peroxy groups (CH2OOH) in addition to OH and CH2OH.

Keywords: secondary organic aerosol, particle phase, oxygen-containing functional groups, organo silicon aerosol, particle formation

Introduction

Ambient nanoparticles (smaller than 100 nm in diameter) can disproportionately affect climate and human health relative to their mass loading in the atmosphere. These particles influence climate by scattering the incoming solar radiation [1-3] and/or serving as cloud condensation nuclei (CCN) [4]. They influence human health by penetrating deep into the respiratory system [5-7]. To better understand these effects, knowledge of the chemical composition is needed [8]. Organic aerosol constitutes a large portion of this matter and most of this contribution is secondary, meaning that it is formed through reaction of gas phase volatile organic compounds (VOC) with oxidants (OH, O3, NO3) to give semi- or non-volatile products [8, 9].

Recently, silicon was reported as a frequent component of ambient nanoparticles [10-12]. Measurements with Nano Aerosol Mass Spectrometer (NAMS), which provides quantitative elemental composition of particles in the 10-30 nm diameter range [13-15], showed that Si was often observed in urban and suburban environments but rarely detected in a remote environment [16]. The location dependence suggests that Si in these particles is associated with human activity. One possible source is atmospheric oxidation of cyclic volatile methylsiloxanes (cVMS), which are commonly used in personal care products [17-19]. cVMS are considered to be environmentally acceptable cleaning agents [20] due to their inertness to ozone [21]. Owing to high vapor pressure, they are easily released into atmosphere where they may react with OH [21] to form semi- or non-volatile products. An atmospheric transport model for three most common cVMS in personal care products: (octamethylcyclotetrasiloxane, D4; decamethylcyclopentasiloxane D5; and dodecamethylcyclohexasiloxane, D6) is consistent with cVMS as a possible source of nanoparticulate Si [16].

In this work, the molecular composition of secondary aerosol obtained from cVMS oxidation is studied in detail. D5 was chosen as the precursor and D5-derived secondary aerosol was produced through a Photo-oxidation Chamber (PC) to simulate photo-oxidation in the atmosphere by reaction with hydroxyl radical. Several measurements were performed including high-resolution ESI-MS, ESI-MSMS and EI-MS. Previous work used only low-resolution GC-MS analysis, which could only provide partial characterization of the oxidation products [22].

Experimental Section

Aerosol Formation and Particle Collection

Aerosol was generated in a Photo-oxidation chamber (PC) consisting of a 50L box shaped chamber made of a perfluoroalkoxy copolymer [23]. Before each experiment, the chamber was flushed continuously with clean, dry, air for 2-3 days. After flushing, background particle concentrations in the PC were below 100/cm3 and 0.5 μg/m3. The air was cleaned by passing compressed air through charcoal and high efficiency particulate matter air (HEPA) filters to remove organic vapors and particulates, and through silica gel to remove moisture. A Scanning Mobility Particle Sizer (SMPS; Model 3080/3078, TSI, St. Paul, Minnesota) was used to measure particle size distributions before and during experiments.

D5 vapor was generating by passing clean, dry air with a flow rate of ~700 cpm (cm3/min) over the surface of D5 fluid (CAS No.541-02-6, Gelest, Morrisville, Pennsylvania) in a gently heated gas flask bubbler. Ozone was generated by passing clean air around a mercury lamp (Model No.81-1025-01, BHK Inc., Ontario, California) and the ozone concentration was monitored with a Model 49i O3 Analyzer (Thermo Fisher Scientific Inc., Waltham, MA). The gas flow of D5 was sent into the PC along with additional 1 cpm flow of 15-20 ppmv ozone and 240cpm flow of water vapor, which was created by bubbling air through deionized water. These flows yielded a nominal residence time of 40 min in the PC. Once the gas flows, ozone concentration and relative humidity were stabilized, reaction was initiated by irradiating the chamber with UV lamps to generate OH from ozone and water. Using the method described by Hall and Johnston [23], the OH mixing ratio for these experiments was estimated to be ~1.5×108 cm−3. Particulate matter in the air flow exiting the PC was collected at a flow rate of ~2 lpm for 20-24 hrs onto a Teflon coated, glass fiber filter (GF/D, CAT No.1823-025, Whatman, GE Healthcare, Piscataway, NJ) for chemical analysis. Under the conditions used in this study, the secondary aerosol mass concentration was ~60 μg/m3, more than one hundred times larger than the background concentration. Generally, 0.3 to 0.5 mg of aerosol was collected in each experiment for analysis. In all experiments, background aerosol samples were collected prior to the injections of reactants to identify and eliminate any artifacts.

After particle collection, the filter was sonicated for one hour with ~8 ml acetonitrile (ACN) (CAS No. 75-05-8, Fisher Scientific, Pittsburgh, Pennsylvania) and the resulting solution was filtered through a polyvinylidenedifluoride (PVDF) filter (CAT No. 6747-2504, GE Healthcare, Piscataway, NJ). The filtered solution was then concentrated nearly to dryness (<0.5ml) with a Speed Vac Concentrator (Model SC110A, Thermo Scientific, Waltham, Massachusetts). Finally, the sample was reconstituted to 1 mg/ml with ACN. Reconstitution with methanol or ACN/H2O 50/50 by volume gave similar mass spectrometry results.

A few experiments were performed by bubbling air through a solution of D4 rather than D5. All other conditions for the experiment remained the same. The D4 concentration in the chamber for these experiments was not estimated, though the aerosol mass loading was similar to the D5 oxidation experiments.

Mass spectrometry

High resolution electrospray mass spectra (ESI-MS) in both positive and negative ion modes were obtained with Q-Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Scientific, Waltham, Massachusetts) equipped with a heated electrospray ionization source (HESI). Samples were injected at a flow rate of 0.5 μL/min with a spray voltage of 3.5 kV and a temperature of 250 °C. Mass spectra were obtained over the range 50-950 m/z with a nominal mass resolving power of 70000. Typically, data were acquired for 1 min with ~60 spectra averaged. For ESI-MS/MS experiments, a quadrupole mass filter isolated each chosen precursor, with subsequent analysis using collision induced dissociation (CID) with the collision energy adjusted between 10 and 30 eV depending on the precursor ion.

Samples were also analyzed by GC-MS with a Waters GCT Premier high-resolution time-of-flight mass spectrometer with a scan range of 50-950 m/z coupled to an Agilent 7890A gas chromatograph. Ionization was performed by 70eV EI with a source temperature of 180 °C. Chromatographic separation was performed with an Agilent HP-5 column using a 3μL injection volume and an injection port temperature of 280 °C.

Four separate experiments of D5 oxidation were performed under the conditions given above, and only those ions which were detected by ESI-MS in all four experiments are discussed below. These ions represented ~85% of the total number of ions detected. Ions that were detected in fewer than four experiments generally had very low signal intensities, which may have inhibited detection in some experiments and/or may have compromised accurate m/z measurement.

Results and Discussion

Figure 1 shows mass spectra of D5-derived secondary aerosol from one of the four replicate experiments. Both monomer and dimer products are observed. As described in supplementary information, a few additional experiments were performed with a lower mass loading of aerosol in the PC. The monomer signal intensity appeared to increase relative to the dimer signal intensity when the mass loading was decreased (Figure S1), though more work is needed to explore this dependence thoroughly.

Figure. 1.

Figure. 1

Open in a new tab

Representative ESI mass spectra of secondary aerosol from D5 oxidation.

Ions observed in all four experiments were further characterized in the following way. First, background peaks from the PC and filter were removed from consideration, as were peaks with <0.5% (positive spectrum) and 0.1% (negative spectrum) relative intensity. This left a total of 190 (positive spectrum) and 233 (negative spectrum) peaks between 50 and 950 m/z. Chemical formulas with theoretical m/z values within ± 5 ppm of the measured m/z values were determined with the mass spectrometer software (X calibur). Reasonable formulas were selected based on the following general criteria: a) C0-10H1-30O1-20Si1-5 for monomers and C10-20H30-60O10-30Si5-10 for dimers, b) up to 1 Na atom for peaks in the positive ion spectrum, and c) H/C atomic ratio between 2.7 and 3.8.

Of the original 190 peaks in the positive spectrum, 154 peaks (81%) were assigned a reasonable single formula based on the above criteria. Of the original 233 peaks in the negative spectrum, 203 peaks (87%) were assigned a single, reasonable formula. The average mass difference between measured and theoretical m/z for peak assignments was 1.28 ppm. The majority of the assigned peaks in the positive ion spectrum (72%) contained one Na atom. Additionally, there was no evidence for multiply charged peaks. Only 30 peaks from the positive spectra and 26 peaks from the negative spectra could not be assigned a reasonable formula, and their relative intensities were generally < 1%. Assigned peaks are given in Table S1 of the supplementary information. The mass weighted signal intensity fractions (MIF) [24] of these ions showed that the average O/Si atomic ratio increased from 1 in D5 to ~1.3 in the oxidized aerosol while the C/Si atomic ratio decreased from 2 to ~1.7 under the experimental conditions used in this study.

Peak assignments from the positive and negative ion spectra were combined, and redundancies due to charge (assuming MH+, MNa+, M-H−) and isotopic substitution were removed, giving 135 unique molecular formulas for the (neutral) products. The products were divided into three types: fragmented products, unsaturated products and saturated substituted products. Fragmented products were defined as those where the siloxane ring had to be broken, i.e. silicon numbers of 1-4 (monomers) or 6-9 (dimers); representative examples are C8H24O4Si4 and C10H32O8Si6. Unsaturated products were defined as those having silicon numbers of 5 (monomers) or 10 (dimers) but with an H/C atomic ratio less than 3, indicating the presence of unsaturated functional group(s); representative examples are C9H26O8Si5 and C18H52O12Si10. Saturated, substituted products were defined as those having silicon numbers of 5 (monomers) or 10 (dimers), but H/C ≥3 and O/Si>1; representative examples are C9H28O7Si5 and C18H54O11Si10. Under the conditions of these experiments, more that 95% of the total mass weighted signal intensity was encompassed by fragmented and saturated products (Table S2). Therefore, only these two product types are discussed below.

Fragmented Oxidation Products of D5

A total of 58 neutral fragmented products were assigned. The molecular formulas suggest the presence of both cyclic and linear siloxane structures with different numbers of subunits, e.g. D2, D3, D4, L2, L3, L4 (D=cyclic; L=linear) with various fragmentations and substitutions e.g. loss of a CH3 group, substitution of an OH group for a CH3 group, etc. (See Table S3.) Figure S2 gives the positive ESI mass spectrum of pure D5 as a control. Several fragmentation peaks are observed, some examples being 355.069 m/z assigned as (CH3)9(OSi)5+ and 297.082 m/z assigned as (CH3)8(OSi)4H+. It is likely that oxidation products of D5 will also fragment during ESI analysis, making it difficult to distinguish whether or not the fragmentation products observed in this work are the result of photooxidation or an ESI artifact.

Saturated Oxidation Products of D5

Under the experimental conditions of this study, saturated products represent the majority of the mass-weighted signal intensity (~70%). Relative to D5 and its dimer, these products contain higher amounts of oxygen and lower amounts of carbon and hydrogen. Figure 2 provides a graphical summary of these formulas as a function of carbon to silicon (C/Si) and oxygen to silicon (O/Si) atomic ratios. The change in O/Si and C/Si ratios between the D5 reactant and its oxidation products indicate the types of reactions that take place [11]. For example, substituting a CH2OH group for a CH3 group increases the O/Si ratio while the C/Si ratio remains constant, i.e. oxidizing (CH3)10(SiO)5 to produce (CH3)10-x(CH2OH)x(SiO)5. Substituting an OH group for a CH3 group decreases C/Si ratio and increases O/Si ratio, i.e. oxidizing (CH3)10(SiO)5 to produce (CH3)10-y(OH)y(SiO)5. For monomers, the assigned product formulas indicate various combinations of OH and CH2OH substitution with up to 6 CH3 groups replaced (Fig. 2a). The same types of substitutions are observed for dimers (Fig. 2b), but with the additional complication that the two monomers in a dimer can be linked with several species i.e. O, CH2 and CH2CH2.

Figure. 2.

Figure. 2

Open in a new tab

Plots of C/Si versus O/Si for saturated products of D5 oxidation in the monomer (a) and dimer (b) regions. Each dot represents an assigned molecular formula of a saturated oxidation product. The arrows in the monomer plot extend from unoxidized D5 and represent the two types of substitution that can occur. The arrows in the dimer plot take into account different possible linkages between monomers.

OH substitution for a CH3 group

Figure 3a and 3b show ESI-MSMS spectra of the C9H28O6Si5 product in positive and negative ion modes, which is considered to have one OH substitution for a CH3 group. The positive spectrum in Figure 3a shows the dissociation of the precursor at 373.080 m/z(+) which is assigned as C9H29O6Si5+, with −2.43 ppm mass accuracy, while the negative spectrum in Figure 3b shows the dissociation of the precursor at 371.066 m/z(−) which is assigned as C9H27O6Si5− with mass accuracy −1.06 ppm. The first neutral loss in positive spectrum is H2O (C9H29O6Si5+ to C9H27O5Si5+), while the corresponding loss in the negative spectrum is CH4 (C9H27O6Si5− to C8H23O6Si5−). These products are consistent with protonation of the OH group in the positive spectrum and deprotonation of the OH group in the negative spectrum. Other diagnostic ions are protonated and deprotonated trimethyl silanol, [Si(CH3)3OH+H+] and [Si(CH3)3O−] respectively, neutral loss of Si(CH4)4, and a series of protonated/deprotonated products corresponding to successive loss of siloxane subunits e.g. loss of C4H12O2Si2 or C6H18O3Si3.

Figure 3.

Figure 3

Open in a new tab

Product ion spectra for isolation of a) 373 m/z(+), b) 371 m/z(−) and c) 387 m/z(+). These precursors correspond to the oxidation product of D5 having one OH substitution for a CH3 group (Fig.3a and Fig. 3b), and D5 having one CH2OH substitution for a CH3 group (Fig.3c).

It should be noted that a nearby precursor ion is observed in the positive mass spectrum at 373.043 m/z(+), which is assigned as the unsaturated product C8H25O7Si5+ (−2.17 ppm). While this ion could lead to some of the products in Figure 3a, several product ions including the first neutral loss cannot be explained by its molecular formula (precursor has fewer carbon atoms than product). Furthermore, when the collision energy is increased from 10eV to 30eV, the product ion signal intensities increase substantially. At the same time, the normalized intensity of the precursor of interest 373.080 m/z(+) decreases by almost a factor of 6 (1.8E6 to 3.4E5), while the normalized intensity for the “impurity” precursor at 373.043 m/z(+) hardly changes at all (8.0E5 to 6.5E5). Taken together, the inconsistency of several product ions with the “impurity” precursor and the disparities in signal intensity suggest that the OH substituted product at 373.080 m/z(+) is the main contributor to the spectrum in Figure 3a and 3b.

CH2OH substitution for a CH3 group

Figure 3c shows the product ion mass spectrum for the 387.098 m/z(+) precursor C10H31O6Si5+ (+0.22 ppm) which corresponds to one CH2OH substitution for a CH3 group. The first neutral loss is H2O, consistent with protonation of the OH functionality. The second and more intense neutral loss is 355.069 m/z(+) (C9H27O5Si5+, 2.92 ppm), which corresponds to the loss of CH3OH from the precursor. Another diagnostic ion is C4H13OSi+, which corresponds to protonated trimethylsilyl-methanol.

Dimer Linkages by O and CH2

Figure 4a shows the product ion spectrum of the 727.140 m/z(+) precursor assigned as C18H55O11Si10+ (−3.47 ppm), while Figure 4b shows an expansion of the product ion spectrum between 350 and 380 m/z(+). Two dimer linkages are possible for the precursor molecular formula. One is linkage by an O atom with all side groups as CH3 (Scheme 1a) and the other by a CH2 group with one side group as OH and the remaining as CH3 (Scheme 1b). The two C9 product ions labeled in black in Figure 4a cannot distinguish between the two linkages since they are expected products of both structures as shown in Scheme 1a and b. However, insight can be gained from other product ions in the spectrum. The first fragmentation from this precursor (C17H51O11Si10+, −1.90 ppm) corresponds to a neutral loss of CH4 while no H2O loss is detected (Figure 4a). When compared with the neutral losses in Figures 3a and 3c, this observation suggests that very few precursors have an OH group and, therefore, the two siloxane rings are linked by an O atom in most dimers. However, the expansion in Figure 4b shows the presence of very low intensity fragment ions that are consistent with a CH2 linkage. In particular, the C10H29O5Si5+ product ion cannot be obtained from a dimer linked by an O atom (see structures in Scheme 1a and 1b). Taken together, there is strong evidence that both O and CH2 linkages exist, though products unique to the O linkage have much greater intensity than those unique to the CH2 linkage.

Figure 4.

Figure 4

Open in a new tab

Product ion spectrum of 727 m/z(+). a) Complete product ion spectrum. b) An expansion of 350 to 380 m/z(+). Peaks corresponding to an O linkage are marked with red. Peaks corresponding with a CH2 linkage are marked with blue. Peaks consistent with both are marked in black. c) Expansion of the product ion spectrum of 795 m/z(+) between 380 and 470 m/z(+).Peaks corresponding with a CH2CH2 linkage are marked with blue.

Scheme 1.

Scheme 1

Open in a new tab

Dimer structures with a) −O−, b) −CH2−, and c) −CH2CH2- linkages. The OH group in b) could be located anywhere around the siloxane ring. The locations of the two CH2OH groups in c)can be anywhere on the left hand ring, while the location of the OH group can be anywhere on the right hand ring. The C11 product confirms the existence of a CH2CH2- linkage. Other labeled ions are consistent with, but not unique to, this linkage.

Dimer linkage by CH2CH2 (m/z 795.128)

A CH2CH2 linked dimer has been discussed previously as a product of D4 oxidation [22]. Here, there are many precursors having formulas consistent with a CH2CH2 linkage, but it is very difficult in to distinguish by MSMS a structure having a CH2 linkage with a CH2OH substitution from one having a CH2CH2 linkage with an OH substitution. Nonetheless, one precursor at 795.128 m/z(+) assigned as C19H56O13Si10Na+ with (−3.18 ppm) provides clear evidence for a CH2CH2 linkage. This precursor must have either a CH2 linkage with 3 CH2OH substitutions or a CH2CH2 linkage with 2 CH2OH substitutions and 1 OH substitution. The key product ion shown in Figure 4c is C11H30O7Si5Na+ (−1.12 ppm), which can only be formed from a precursor having a CH2CH2 linkage as illustrated by the structure in Scheme 1c.

Comparison to previous work

GC-MS data are included as supplementary information. These data provide the opportunity to check consistency with ESI MSMS analysis and compare with previous work. Similar to the study of Sommerlade [22], GC-MS confirms that major components of D5-derived secondary aerosol are D5 (presumably partitioned to the particle phase and/or bound within dimers), and an oxidation product corresponding to one OH substitution for CH3 (see Figure S3). Our GC-MS results and those of Sommerlade provide evidence for a CH2OH substitution product and the presence of dimers (O and CH2 linked dimers), which is consistent with ESI-MS results, though the full range of products observed with ESI cannot be characterized by GC-MS.

The major products characterized in this work can be explained by OH abstraction of a hydrogen atom from D5 as the initial step. Scheme 2 shows possible formation pathways for several of these products. Formation of the OH substitution product is similar to the reaction sequence proposed by Atkinson [25]. Note that formation of a multiply substituted product does not necessarily require multiple OH abstractions, since an auto-oxidation process based on internal hydrogen rearrangements [26] could lead to several substitutions.

Scheme 2.

Scheme 2

Open in a new tab

Possible pathways for product formation from OH oxidation of D5.

D4-derived secondary aerosol

Secondary aerosol derived from OH oxidation of D4 (C8H24O4Si4) was briefly studied for comparison to D5-derived aerosol. The ESI-MS spectra for D4-derived aerosol showed many more peaks than D5-derived aerosol, especially in the dimer and trimer regions. By applying the analysis procedure as discussed previously for D5, of the original 736 peaks in the positive spectrum, 671 peaks (91%) were assigned a reasonable, single formula. Of the original 455 peaks in the negative spectrum, 409 peaks (90%) were assigned a single, reasonable formula. The average mass difference between measured and theoretical m/z for peak assignments was 2.56 ppm. After removing redundancies due to charge (assuming MH+, MNa+, M-H−) and isotopic substitution, 529 unique molecular formulas for the (neutral) products were obtained, which could be subdivided into fragmented, unsaturated and saturated products. About 96 % of the total products were fragmented or saturated, but more fragmented products were observed for D4-derived aerosol than D5-derived aerosol. MIF analysis showed the O/Si ratio increased from 1 in D4 to ~1.48 in the oxidized aerosol while the C/Si ratio decreased from 2 to ~ 1.56 for the conditions used in this study.

Saturated products are shown graphically in Figure 5 where C/Si vs. O/Si is plotted for each observed molecular formula. Both OH and CH2OH substitutions for CH3 were identified, and relative to D5, a greater fraction of the CH3 groups were found to be oxidized in D4. Furthermore, monomer products having more than 8 oxygen atoms added to the D4 molecule were detected, indicating that peroxy groups (CH2OOH) must also be formed.

Figure 5.

Figure 5

Open in a new tab

Plots of C/Si versus O/Si for saturated products of D4 oxidation in the a) monomer and b) dimer regions. Each dot represents an assigned molecular formula of a saturated oxidation product. The arrows in the monomer plot extend from unoxidized D4 and represent the two main types of substitution that occur. The arrows in the dimer plot take into account different possible linkages between monomers. Circles represent monomer products where all eight CH3 groups have been substituted. Formulas to the right of the circles (same C/Si but higher O/Si) must contain peroxy groups.

Conclusions

The molecular composition of cVMS-derived secondary aerosol was characterized by high performance mass spectrometry. The products can be divided into three types: fragmented products, unsaturated substituted products and saturated substituted products. Fragmented products arise at least in part as artifacts of electrospray ionization. Saturated substituted products exhibit the highest signal intensities. Based on a mass-weighted intensity fraction analysis, gas phase D5 (O/Si=1, C/Si=2) was oxidized to produce aerosol whose detected products have an average O/Si ≈ 1.3 and an average C/Si ≈ 1.7 under the conditions studied. Gas phase D4 was oxidized to produce trimers in addition to monomers and dimers, with the detected products having average O/Si ≈ 1.5 and C/Si ≈ 1.6 under the conditions studied. ESI-MS/MS analysis of the D5-derived aerosol provided evidence for major substitution types along the siloxane ring and the linkages of two siloxane rings to produce dimers. The results showed that OH and CH2OH substitutions are prevalent, and dimers can be linked by O, CH2 and CH2CH2 groups. D4-derived aerosol gave evidence for the presence of peroxy groups (CH2OOH). GC-MS with EI generally confirmed the ESI analysis, though products were incompletely characterized owing to the greater extent of ion fragmentation.

Supplementary Material

supporting information

Acknowledgements

This research was supported by the National Science Foundation under grant number CHE-1408455. The Orbitrap mass spectrometer used in this study was purchased under grant number S10 OD016267-01 and supported by grant number 1 P30 GM110758-01, both from the National Institutes of Health. The GCT mass spectrometer used in this study was purchased under grant number CHE-1229234 from the National Science Foundation. The authors thank Andrew Horan for assistance with acquisition and processing of MS data.

References

  • (1).Smith JN, Barsanti KC, Friedli HR, Ehn M, Kulmala M, Collins DR, Scheckman JH, Williams BJ, McMurry PH. Observations of aminium salts in atmospheric nanoparticles and possible climatic implications. Proceedings of the National Academy of Sciences of the United States of America. 2010;107(15):6634–6639. doi: 10.1073/pnas.0912127107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (2).Zhang RY, Wang L, Khalizov AF, Zhao J, Zheng J, McGraw RL, Molina LT. Formation of nanoparticles of blue haze enhanced by anthropogenic pollution. Proceedings of the National Academy of Sciences of the United States of America. 2010;106(42):17650–17654. doi: 10.1073/pnas.0910125106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (3).See SW, Balasubramanian R, Wang W. A study of the physical, chemical, and optical properties of ambient aerosol particles in Southeast Asia during hazy and non-hazy days. Journal of Geophysical Research-Atmospheres. 2006;111(D10) [Google Scholar]
  • (4).Bzdek BR, Johnston MV. New Particle Formation and Growth in the Troposphere. Analytical Chemistry. 2010;82(19):7871–7878. doi: 10.1021/ac100856j. [DOI] [PubMed] [Google Scholar]
  • (5).Li N, Xia T, Nel AE. The role of oxidative stress in ambient particulate matter-induced lung diseases and its implications in the toxicity of engineered nanoparticles. Free Radical Biology and Medicine. 44(9):1689–1699. doi: 10.1016/j.freeradbiomed.2008.01.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (6).Inoue KI, Takano H, Yanagisawa R, Hirano S, Kobayashi T, Fujitani Y, Shimada A, Yoshikawa T. Effects of inhaled nanoparticles on acute lung injury induced by lipopolysaccharide in mice. Toxicology. 2007;238(2-3):99–110. doi: 10.1016/j.tox.2007.05.022. [DOI] [PubMed] [Google Scholar]
  • (7).Quadros ME, Marr LC. Environmental and Human Health Risks of Aerosolized Silver Nanoparticles. Journal of the Air & Waste Management Association. 2010;60(7):770–781. doi: 10.3155/1047-3289.60.7.770. [DOI] [PubMed] [Google Scholar]
  • (8).Riipinen I, Yli-Juuti T, Pierce JR, Petaja T, Worsnop DR, Kulmala M, Donahue NM. The contribution of organics to atmospheric nanoparticle growth. Nature Geoscience. 2012;5(7):453–458. [Google Scholar]
  • (9).Kroll JH, Seinfeld JH. Chemistry of secondary organic aerosol: Formation and evolution of low-volatility organics in the atmosphere. Atmospheric Environment. 2008;42(16):3593–3624. [Google Scholar]
  • (10).Pennington MR, Klems JP, Bzdek BR, Johnston MV. Nanoparticle chemical composition and diurnal dependence at the CalNex Los Angeles ground site. Journal of Geophysical Research-Atmospheres. 2012;117 [Google Scholar]
  • (11).Bein KJ, Zhao YJ, Wexler AS, Johnston MV. Speciation of size-resolved individual ultrafine particles in Pittsburgh, Pennsylvania. Journal of Geophysical Research-Atmospheres. 2005;110(D7) [Google Scholar]
  • (12).Phares DJ, Rhoads KP, Johnston MV, Wexler AS. Size-resolved ultrafine particle composition analysis - 2. Houston. Journal of Geophysical Research-Atmospheres. 2003;108(D7) [Google Scholar]
  • (13).Wang SY, Zordan CA, Johnston MV. Chemical characterization of individual, airborne sub-10-nm particles and molecules. Analytical Chemistry. 2006;78(6):1750–1754. doi: 10.1021/ac052243l. [DOI] [PubMed] [Google Scholar]
  • (14).Wang SY, Johnston MV. Airborne nanoparticle characterization with a digital ion trap-reflectron time of flight mass spectrometer. International Journal of Mass Spectrometry. 2006;258(1-3):50–57. [Google Scholar]
  • (15).Pennington MR, Johnston MV. Trapping charged nanoparticles in the nano aerosol mass spectrometer (NAMS) International Journal of Mass Spectrometry. 2012;311:64–71. [Google Scholar]
  • (16).Bzdek BR, Horan AJ, Pennington MR, Janechek NJ, Baek J, Stanier CO, Johnston MV. Silicon is a Frequent Component of Atmospheric Nanoparticles. Environmental Science & Technology. 2014;48(19):11137–11145. doi: 10.1021/es5026933. [DOI] [PubMed] [Google Scholar]
  • (17).Genualdi S, Harner T, Cheng Y, MacLeod M, Hansen KM, van Egmond R, Shoeib M, Lee SC. Global Distribution of Linear and Cyclic Volatile Methyl Siloxanes in Air. Environmental Science & Technology. 2011;45(8):3349–3354. doi: 10.1021/es200301j. [DOI] [PubMed] [Google Scholar]
  • (18).Krogseth IS, Kierkegaard A, McLachlan MS, Breivik K, Hansen KM, Schlabach M. Occurrence and Seasonality of Cyclic Volatile Methyl Siloxanes in Arctic Air. Environmental Science & Technology. 2013;47(1):502–509. doi: 10.1021/es3040208. [DOI] [PubMed] [Google Scholar]
  • (19).Warner NA, Evenset A, Christensen G, Gabrielsen GW, Borga K, Leknes H. Volatile Siloxanes in the European Arctic: Assessment of Sources and Spatial Distribution. Environmental Science & Technology. 2010;44(19):7705–7710. doi: 10.1021/es101617k. [DOI] [PubMed] [Google Scholar]
  • (20).Wang DG, Norwood W, Alaee M, Byer JD, Brimble S. Review of recent advances in research on the toxicity, detection, occurrence and fate of cyclic volatile methyl siloxanes in the environment. Chemosphere. 2013;93(5):711–725. doi: 10.1016/j.chemosphere.2012.10.041. [DOI] [PubMed] [Google Scholar]
  • (21).Atkinson R. Kinetics of the Gas-Phase Reactions of a Series of Organosilicon Compounds with OH and NO3 Radicals and O3 at 297 +/− 2-K. Environmental Science & Technology. 1991;25(5):863–866. [Google Scholar]
  • (22).Sommerlade R, Parlar H, Wrobel D, Kochs P. Product Analysis and Kinetics of the Gas-Phase Reactions of Selected Organosilicon Compounds with OH Radicals Using a Smog Chamber - Mass-Spectrometer System. Environmental Science & Technology. 1993;27(12):2435–2440. [Google Scholar]
  • (23).Hall WA, Johnston MV. Oligomer Content of alpha-Pinene Secondary Organic Aerosol. Aerosol Science and Technology. 2011;45(1):37–45. [Google Scholar]
  • (24).Hall WA, Pennington MR, Johnston MV. Molecular Transformations Accompanying the Aging of Laboratory Secondary Organic Aerosol. Environmental Science & Technology. 2013;47(5):2230–2237. doi: 10.1021/es303891q. [DOI] [PubMed] [Google Scholar]
  • (25).Atkinson R, Tuazon EC, Kwok ESC, Arey J, Aschmann SM, Bridier I. Kinetics and Products of the Gas-Phase Reactions of (CH3)4Si, (CH3)4SiCH2OH, (CH3)3SiOSi(CH3)3 and (CD3)3SiOSi(CD3)3with Cl Atoms and OH Radicals. J Chem Soc Faraday T. 1995;91(18):3033–3039. [Google Scholar]
  • (26).Crounse JD, Nielsen LB, Jorgensen S, Kjaergaard HG, Wennberg PO. Autoxidation of Organic Compounds in the Atmosphere. J Phys Chem Lett. 2013;4(20):3513–3520. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supporting information
NIHMS797501-supplement-supporting_information.pdf (296KB, pdf)

RESOURCES