Advances in the Separation and Detection of Secondary Organic Aerosol Produced by Decamethylcyclopentasiloxane (D5) in Laboratory-Generated and Ambient Aerosol

Decamethylcyclopentasiloxane (D5), a common ingredient in many personal care products (PCPs), undergoes oxidation in the atmosphere, leading to the formation of secondary organic aerosol (SOA). Yet, the specific contributions of D5-derived SOA on ambient fine particulate matter (PM2.5) have not been characterized. This study addresses this knowledge gap by introducing a new analytical method to advance the molecular characterization of oxidized D5 and its detection in ambient aerosol. The newly developed reversed phase liquid chromatography method, in conjunction with high-resolution mass spectrometry, separates and detects D5 oxidation products, enabling new insights into their molecular and isomeric composition. Application of this method to laboratory-generated SOA and urban PM2.5 in New York City expands the number of D5 oxidation products observed in ambient aerosol and informs a list of molecular candidates to track D5-derived SOA in the atmosphere. An oxidation series was observed in which one or more methyl groups in D5 (C10H30O5Si5) is replaced by a hydroxyl group, which indicates the presence of multistep oxidation products in ambient PM2.5. Because of their specificity to PCPs and demonstrated detectability in ambient PM2.5, several oxidation products are proposed as molecular tracers for D5-derived SOA and may prove useful in assessing the impact of PCPs-derived SOA in the atmosphere.


Mobile phase optimization
For mobile phase optimization, the response of surrogate standards was assessed under three different solvent ratios of acetonitrile to water (60:40, 50:50, and 40:60), three different buffer compositions (20 mM ammonium acetate, 20 mM ammonium hydroxide with ammonium bicarbonate, and 20 mM ammonium hydroxide), two different concentrations (10 mM and 20 mM), and two pH values (10 and 11).
The effect of the acetonitrile-to-water solvent ratio on the response of surrogate standards at 1000 μg L -1 each was examined at 60:40, 50:50, and 40:60.There were no significant differences in intensities (Figure S1) and the 50:50 acetonitrile-to-water ratio was used for subsequent optimization.
Three buffer systems, ammonium acetate (pH 6.67), ammonium hydroxide with ammonium carbonate (8.63), and ammonium hydroxide (9.95), were examined at concentrations of 20 mM in a surrogate standard solution at 1000 μg L -1 .Among these, 20 mM ammonium hydroxide resulted in the highest intensities for all six standards compared to the other two buffer systems (Figure S1).It is expected that the highest intensities were achieved for ammonium hydroxide because its pH is the most basic and is closest to the pKa values of the standards.To increase buffer capacity, ammonium bicarbonate (20 mM) was added.When the pH was increased from pH 10 to pH 11, there was no further increase in intensity (Figure S2), so the lower pH value of 10 was utilized.
The buffer concentration was lowered from 20 mM to 10 mM and the intensity of the signals did not significantly decrease (Figure S2), so the 10 mM buffer concentration was utilized.Having a lower concentration of buffer pH is generally preferred in ESI mass spectrometry to improve sensitivity by decreasing the number of ions created from the background compounds.

Additional source parameters
A 1000 μg L -1 solution of standards was introduced into the mass spectrometer through direct infusion while mobile phase flow was maintained via a T-junction in order to optimize the ESI source parameters that control the spraying of the LC eluent and negative ionization of the molecules.Because those parameters are affected by flow rate and mobile phase composition, a T-connection was necessary to simulate chromatographic conditions.The sheath gas is the inner nitrogen flow that nebulizes the sample flowing from the needle to form fine droplets, whereas the auxiliary gas is the heated outer nitrogen flow that aids the sheath nitrogen in sample desolvation.
The temperature of the heater was kept at 413 C to achieve optimal sample desolvation.A spray voltage of 2.5 kV in negative mode was used to achieve ionization.The ion transfer capillary is responsible for transferring the ionized species generated by the ESI source to the S-lens while also ensuring that any residual solvent is evaporated.The temperature of the capillary was kept constant at 256 C.
A 1000 μg L -1 standard solution was directly infused to acquire data-dependent MS 2 (dd-MS 2 ) data (Table 1).In the data-dependent acquisition mode, each precursor with an isolation window of 1 Da was isolated, fragmented, and its product ions are detected.The obtained data were used to assign formulas to deprotonated molecules and product ions.Product ion spectra provided by the above standards under applied (-) ESI conditions are shown in Figure S3.Tris (tert-butoxy)silanol (Figure S3a) fragmented to m/z 207, 151, and 95 fragments from the H abstraction followed by the heterolytic cleavage of each O-C bond in Si-O-C bond sequence forming a silanol functional group.The bond energy of Si-O is higher (110 kCal/mol) compared to the bond energy of O-C (85.5 kCal/mol) leading to preferential breaking of the O-C bond during the fragmentation.A similar fragmentation pattern was observed in tris (tert-pentoxy) silanol (Figure S3b).

Table S1:
The elemental composition of product ions observed by the most intense peak of laboratory OFR sample (marked with a star in Figure 1a).

Figure S1 :
Figure S1: Log of the intensity of the mixed standard solution of 1000 μg L -1 at three different buffer compositions and three different solvent ratios

Figure S4 :Figure S5 :Figure S6 :
Figure S4: Extracted chromatograms from a mixed solution containing 100 μg L -1 of each of four standards that was detected and was in the accepted range of linearity (R 2 ≥ 0.995) in the C-18 column

Figure S7 :
Figure S7: Product ion spectra provided by the a) C9H28O6Si5 (m/z 371.0655) under applied (-) ESI conditions for the peak observed in the OFR sample at tR 10.9 min

Figure S8 :
Figure S8: Comprehensive flow diagram outlining the qualitative data analysis process for data acquired through LC-MS analysis, covering both laboratory and New York City (NYC) samples.