In situ analysis of the bulk and surface chemical compositions of organic aerosol particles

The study addresses how chemical structures and reactions at aerosol particle surfaces differ from those in the particle bulk and from planar air/water interfaces. While aerosol surfaces are critical gateways for heterogeneous and multiphase atmospheric processes, there has been a lack of direct, surface-specific in situ techniques to probe molecular identities, orientations, and adsorption energetics on submicron aerosol particles. The authors aim to directly identify organic species and characterize their surface behavior on airborne aerosol particles using vibrational sum frequency scattering (VSFS), and to compare surface adsorption on curved aerosol surfaces to that on planar air/water interfaces.

The paper reviews prior work highlighting the distinct chemistry at aerosol interfaces versus bulk phases, including heterogeneous oxidation, surface pH effects, and reaction acceleration at interfaces. It notes that many studies have been ex situ analyses of collected particles. Recent in situ optical methods (second harmonic scattering, SHS; electronic sum frequency scattering, ESFS) revealed molecular presence, polarity, and orientation at aerosol surfaces, but lacked chemical specificity. The need for a technique that provides surface-specific chemical identification on airborne aerosols motivates the development of VSFS, alongside bulk-sensitive hyper-Raman scattering (HRS), building on nonlinear Mie theory and prior SFG/SHG surface spectroscopy literature.

The authors developed an in situ experimental platform enabling simultaneous acquisition of VSFS (surface-specific) and HRS (bulk-sensitive) spectra from airborne, laboratory-generated aerosol particles. - Light sources: A PHAROS femtosecond amplifier (100 kHz, ~8 W) pumped an ORPHEUS-ONE OPA to generate tunable mid-IR (2500–4500 nm). An etalon produced a spectrally narrow picosecond 1025 nm beam (linewidth ~8 cm−1) from the residual pump. - Beam geometry and focusing: The 1025 nm and IR beams were incident non-collinearly at ~0° and 5° relative to the X axis (incident plane parallel to the optical table). Pulse energies were ~6.0 µJ (1025 nm) and ~2.0 µJ (IR). Focusing optics yielded focal spot diameters ~90 µm (1025 nm) and ~80 µm (IR). Polarizations were controlled via half-wave plates; thin-film polarizers selected the scattered signal polarization. Polarization conventions: H (parallel to table), V (perpendicular). - Signal collection and detection: VSFS signals were collected at ~90° relative to the X axis with a ~60° collection angle using lenses (2", f = 3.2 cm and 7.5 cm), dispersed by an Acton 300i spectrometer and detected with a Princeton Instruments CCD (LN/CCD-1340/400). Integration time: 180 s. HRS signals were collected at 0° with a similar optical train and spectrometer/CCD; integration time: 60 s. VSFS frequency: ωVSFS = ω1025nm + ωIR; HRS frequency: ωHRS = 2ω1025nm − ωIR. - Aerosol generation: A 0.5 M NaCl seed solution was atomized (TSI 3076) at 40 psi and 4 slpm. Particle number density was ~3.8 × 10^6 cm−3 with diameters centered near 40 nm (distribution 10–300 nm), characterized by TSI Optical Particle Sizer 3330 and Nanoparticle Sizer 3910. Aerosols were delivered to an enclosed chamber; an exhaust pump prevented accumulation. For VSFS intensity vs particle density, the aerosol stream was diluted with humidified N2 via a bubbler and flow control. - Planar VSFG reference: A separate 800 nm, 1 kHz amplifier (UpTek Solutions) pumped a TOPAS OPA to produce a 3250 nm IR pulse (5 µJ). The residual 800 nm (14 µJ) was combined in reflection geometry onto samples in a PTFE dish (2" diameter) with detection via an Acton SpectraPro 2300i and CCD. Samples matched VSFS solutions. - Samples and chemicals: Propionic (propanoic) acid and butanoic acid (Acros Organics) were used as received; ultrapure water (18.2 MΩ·cm). NaCl (Fisher) was baked at 600 °C for 10 h before use. Acids were added to 0.5 M NaCl to target concentrations. - Polarization-resolved VSFS: Spectra were collected for HHH, VVH, VHV, and HVV combinations to assess molecular orientation at aerosol surfaces. - Isotherms and adsorption analysis: VSFS (aerosol surface) and VSFG (air/water interface) electric fields for the CH3-as mode were measured versus bulk acid concentration. A Langmuir-type relation Es ∝ Ns·Kc/(Kc + 55.5) was fitted to extract adsorption constants K and corresponding adsorption free energies.

- Simultaneous surface and bulk vibrational fingerprints: • VSFS (surface) of aerosolized 4.0 M propionic acid (0.5 M NaCl) showed peaks at 2887.3 cm−1 (CH3 symmetric stretch), 2950.9 cm−1 (Fermi resonance), and 2991.8 cm−1 (CH3 asymmetric stretch). • HRS (bulk) from the same aerosols showed peaks at 2873.0, 2933.6, and 2986.1 cm−1 (assigned to CH3-ss, Fermi, CH3-as), with differing relative intensities from VSFS, consistent with bulk orientational randomness. - Surface vs bulk scaling: • HRS intensity scaled linearly with bulk propionic acid concentration. • VSFS intensity exhibited a nonlinear dependence on concentration, consistent with surface adsorption behavior and differing from bulk response. • VSFS intensity scaled linearly with aerosol particle number density upon dilution with humidified N2, confirming a surface scattering origin. - Polarization selection and orientation: • Detectable VSFS signals appeared for HHH and VVH polarizations; VHV and HVV were silent, indicating orientational ordering of surface propionic acid CH3 groups. Qualitative agreement with selection rules from nonlinear Mie theory was observed, but quantitative discrepancies remained. - Adsorption energetics on curved aerosol vs planar interfaces (CH3-as mode isotherms in 0.5 M NaCl): • Propionic acid adsorption constant K: (1.67 ± 0.48) × 10^2 (aerosol surface) vs (4.95 ± 0.53) × 10^2 (air/water planar interface). • Corresponding adsorption free energies: −12.69 ± 0.28 kJ mol−1 (aerosol surface) vs −15.38 ± 0.11 kJ mol−1 (planar), showing weaker adsorption on curved aerosol surfaces. • Butanoic acid at aerosol surfaces exhibited adsorption free energy −15.96 ± 0.24 kJ mol−1; as expected, longer chain length increased adsorption strength (more negative ΔG).

The work directly addresses the lack of surface-specific chemical identification for airborne aerosol particles by implementing VSFS, while simultaneously probing bulk composition via HRS. The spectral assignments confirm that VSFS selectively reports on interfacial molecules, whereas HRS reflects in-particle bulk structure. Polarization-resolved VSFS indicates nonrandom orientation of propionic acid at aerosol surfaces, a key factor influencing heterogeneous reaction kinetics and mechanisms. Crucially, concentration-dependent isotherms reveal that adsorption of propionic acid is weaker on curved aerosol surfaces than at the planar air/water interface, challenging the common assumption that planar interfacial data can be directly applied to aerosol particles. This has implications for modeling heterogeneous atmospheric chemistry, secondary organic aerosol formation, and potentially the behavior of biological/viral aerosols, since surface composition and orientation govern reactivity, uptake, and aging processes.

The study demonstrates an in situ, surface-specific VSFS method to chemically identify and characterize organic molecules at aerosol particle surfaces, with simultaneous HRS enabling bulk-phase analysis. It establishes orientational ordering of surface propionic acid and quantifies adsorption energetics, showing weaker adsorption on curved aerosol surfaces than on planar air/water interfaces. These findings suggest that models relying on planar interface data may overestimate adsorption on aerosols and should be re-evaluated. Future work should focus on quantitative orientation analysis incorporating refractive index mismatch and small-particle scattering effects, expanding to diverse organic and inorganic species, probing chemical aging and oxidation processes, and leveraging concurrent VSFS/HRS to disentangle surface vs bulk reaction pathways in real time.

- Quantitative discrepancies between measured polarization-resolved VSFS intensities and predictions from existing nonlinear Mie/small-particle theories indicate incomplete modeling; refractive index mismatch and small-particle scattering effects likely contribute and require further treatment. - Experiments were conducted on laboratory-generated NaCl-seeded aerosols with selected organic acids (propionic and butanoic acids), which may not fully represent complex atmospheric aerosol compositions. - Adsorption analyses used a Langmuir-type model and focused on CH3-as mode isotherms; broader validation across modes/species and environmental conditions is needed.