Probing photoreactions of individual suspended carbonaceous aerosols by multi-wavelength OT-SERS

Abstract

This work reports the first application of multi-wavelength optical trapping surface-enhanced Raman spectroscopy (OT-SERS) for in situ probing of photochemical reactions on individual suspended aerosol particles. A composite aerosol model (4-MBA/silver nanoparticles/activated carbon) with a size ranging from 2 to 10 μm, comparable to the particle sizes of PM_2.5 and PM₁₀, was designed to validate the ability of OT-SERS to probe interfacial photoreactions on aerosols. By integrating non-contact optical trapping with SERS detection, we directly monitored the interfacial photoreactions occurring on suspended composite particles under laser irradiation at wavelengths of 473, 589 and 671 nm. Under 473 nm irradiation, the temporal OT-SERS spectra showed dynamic intermediate formation and aromatic ring cleavage, which were attributed to the photooxidations induced by activated carbon. In the experiments involving 589 nm irradiation, relatively weak photooxidations were observed compared to those under 473 nm irradiation. Among the three wavelengths (473, 589 and 671 nm), the irradiation at 671 nm resulted in the weakest photooxidation. OT-SERS was successfully employed for in situ detection of the photooxidation process on the composite particles, which confirmed the formation of hydroxylated intermediates and the cleavage of benzene rings caused by photooxidation.

---

Introduction

The detection of photochemical reactions on an individual aerosol represents a significant challenge in analytical chemistry, particularly for the complex interfacial processes under realistic atmospheric conditions. Traditional methods for studying aerosol photochemistry rely on bulk-phase analyses, such as depositing aerosols onto a substrate or dispersing them in a solution.^1–3 These methods mask heterogeneities caused by particle morphology, surface defects, and dynamic environmental conditions, making it difficult to reflect the dynamic interactions of individual particles under realistic atmospheric conditions.⁴ Therefore, it is necessary to detect individual suspended aerosol particles without substrate interference to obtain information on their morphology, chemical composition, and photochemical reaction processes.

Optical trapping (OT) suspends particles using the interaction force between the laser and particles.^5,6 By suspending individual particles in air, they more accurately simulate their native state in the atmosphere and avoid interference from the substrate.⁷ When combined with Raman spectroscopy, OT enables characterization of particle structures and real-time monitoring of compositional changes during reactions.^8–11 However, the trace amounts of organic molecules adsorbed on aerosol surfaces fall far below the detection thresholds of conventional Raman techniques, making it difficult for conventional Raman to directly observe photochemical reactions at the air–particle interface.

Surface-enhanced Raman spectroscopy (SERS) offers a solution to this sensitivity limitation by amplifying Raman signals through localized surface plasmon resonance (LSPR) generated by noble metal nanoparticles.¹² By integrating SERS with optical trapping, optical trapping surface-enhanced Raman spectroscopy (OT-SERS) eliminates interference from the substrate and amplifies the Raman signal of aerosol particles. Recent studies have demonstrated that silver nanoparticles deposited on the surface of aerosol particles through mixing or electrospray methods, and the Raman signal of aerosol particles can be enhanced by 10 to 100 times.^13–15 However, there has been no report on the application of OT-SERS for characterizing the molecules adsorbed on the surface of individual aerosols and investigating their photochemical reactions.

Here, we have developed multi-wavelength OT-SERS to study photochemical reactions at the individual suspended carbonaceous aerosol. This marks the first use of OT-SERS in such investigations. We designed carbonaceous aerosol particles composed of 4-mercaptobenzoic acid (4-MBA), silver nanoparticles (AgNPs), and activated carbon as a soot analog to investigate photochemical processes. 4-MBA is a monocyclic aromatic compound containing both –SH and –COOH functional groups. Due to its environmental relevance and redox activity, it can be used as a photochemical probe.¹⁶ Activated carbon is used to simulate the elemental carbon in soot. Appropriately sized AgNPs exhibiting a localized surface plasmon resonance (LSPR) peak at 473 nm were synthesized by a seed growth method and deposited on the activated carbon surface as a SERS substrate.

Using this OT-SERS, we investigated the wavelength-dependent photoreaction behaviours of the composite particles under multi-wavelength laser irradiation (473, 589 and 671 nm) and monitored the changes in 4-MBA during the photoreaction process at the single-aerosol level. By combining non-contact optical trapping with multi-wavelength SERS, this work establishes a platform for studying interfacial photochemistry in aerosol systems.

Experimental

Materials

Silver nitrate (AgNO₃, AR) was purchased from Sinopharm Chemical Reagent Ltd. 4-Mercaptobenzoic acid (4-MBA, 99%) was purchased from Sigma-Aldrich. Tannic acid (TA, 98%) and sodium citrate (SC, 99%) were purchased from Macklin Biochemical Co., Ltd. All the chemicals were used without further purification. Ultrapure water (18.2 MΩ cm) was used in this study.

Synthesis and characterization of AgNPs prepared by stepwise seeded growth

The preparation of Ag nanoparticles was conducted following a previously reported method with slight modifications.¹⁷ First, 100 mL of an aqueous solution containing 5 mM SC and 0.1 mM TA was taken in a round-bottom flask, and the mixture was vigorously stirred at 100 °C for 15 min using a reflux condenser to prevent solvent evaporation. Then, 1 mL of 25 mM AgNO₃ aqueous solution was added, and the reaction was continued for 1 h, resulting in a first-generation AgNP solution.

Twenty millilitres of the sample were taken from the round-bottom flask and stored in a refrigerator, and 17 mL of water was added to the flask, while maintaining the temperature at 90 °C. Next, 0.5 mL of 25 mM SC, 1.5 mL of 2.5 mM TA, and 1 mL of 25 mM AgNO₃ solution were added sequentially (with 1 min intervals between each addition). After 20 min, another 20 mL of the product was collected and stored in a refrigerator, resulting in a second-generation AgNP solution. Then, 17 mL of water was added to the flask again, followed by sequential addition of 0.5 mL of 25 mM SC, 1.5 mL of 2.5 mM TA, and 1 mL of 25 mM AgNO₃ solution (with 1 min intervals between each addition). By repeating this process, the size of the AgNPs gradually increased, resulting in different generations of AgNP solutions. Fig. S1a (ESI†) shows the images of the solutions from different generations of AgNP solutions after 20 times dilution with water. As the AgNPs grew, the colour of the solution gradually changed from yellow to red and finally became milky white.

The absorption spectra of AgNP solutions were measured using a UV-Vis spectrophotometer. After diluting the AgNP solutions 15 times with water, they were placed in cuvettes for absorption measurements at room temperature in the wavelength range of 350–750 nm. Fig. S1b (ESI†) shows the UV-Vis absorption spectra of AgNP solutions from different generations. As the generation of AgNPs increased, the absorption peak gradually shifted to longer wavelengths, which is consistent with the red-shift of the surface plasmon resonance peak caused by the increase in the size of AgNPs.^17,18 The absorption peak positions of the AgNPs are shown in Table S1 (ESI†). Notably, the absorption peak of the 15th-generation AgNP solution matches most closely with the laser wavelength (473 nm), making it capable of generating a strong surface-enhanced Raman scattering effect.

Preparation of 4-MBA/AgNPs/C and 4-MBA/C samples

Activated carbon (10 mg), 1 mL of the 15th-generation AgNP solution, and 200 μL of 10 mM 4-MBA ethanol solution were mixed in a 5 mL centrifuge tube and ultrasonicated for 30 min. The mixtures were allowed to stand at room temperature for 24 h and then dried in an oven at 60 °C for 12 h. The dried samples were placed in different cuvettes and shaken to suspend the particles, which were then trapped in an optical trap. The preparation conditions of 4-MBA/C particles were identical to those of 4-MBA/AgNPs/C particles, except that 4-MBA/C particles do not contain AgNPs.

Configuration of the optical trapping system

The OT Raman system integrates optical trapping, Raman detection, and optical image measurement as shown in Fig. 1. Three continuous-wave lasers with 473, 589 and 671 nm wavelengths were split into two beams of equal intensity using a beam splitter. These beams then passed through axicons to form two counter-propagating hollow beams that were focused onto the trapped particles using two objectives. A 20× objective was used for signal acquisition, and the scattering light was split using a beam splitter. One part of the light was directed to a CMOS camera for optical imaging, and the other part was directed into the spectrometer for spectral measurement. The grating of the spectrometer was 1200 grooves per mm and the slit width was set to 200 μm.

Fig. 1 Schematic diagram of the optical trapping Raman spectroscopy system (HWP: half-wave plate; PBS: polarizing beam splitter; M: mirror; AL: axicon; 10 : 90 BS: 10 : 90 beam splitter; FC: fiber coupler).

Photochemical reaction process monitored by OT-SERS

In this experiment, the laser power of each wavelength used was 100 mW. Once the particles were trapped, they were continuously monitored by Raman spectroscopy under sustained exposure to the trapping laser to study the photoreaction induced by the laser. The integration time of the Raman spectra was 100 s and all Raman spectra were baseline-corrected.

Results

OT-SERS of individual 4-MBA/AgNPs/C particles under 473, 589 and 671 nm laser excitation

Fig. 2 demonstrates the capability of OT-SERS to resolve the molecular signatures of 4-MBA on individual suspended 4-MBA/AgNPs/C particles under 473, 589 and 671 nm laser excitation. Remarkably, all characteristic Raman peaks of 4-MBA were detected with high resolution without substrate interference. When using a 473 nm laser, the Raman peak at 1584 cm⁻¹ becomes stronger compared to the other peaks. This phenomenon is caused by the chemical enhancement effect resulting from the charge transfer between the SERS substrate and 4-MBA molecules under the action of a short-wavelength laser.^19,20 The specific assignments of the Raman peaks of 4-MBA are listed in Table S2 (ESI†).^21–24 Fig. S2 (ESI†) shows the optical trapping Raman spectra of 4-MBA/C particles under the three different lasers, with no Raman signals of 4-MBA detected under these wavelengths. The results from Fig. 2 and Fig. S2 (ESI†) indicate that the AgNPs can effectively enhance the Raman signals of 4-MBA adsorbed on the composite particles.

Fig. 2 OT-SERS and optical images of 4-MBA/AgNPs/C particles under trapping and excitation by lasers with wavelengths of 473 nm, 589 nm, and 671 nm. Scale bar: 10 μm.

Photoreaction of individual 4-MBA/AgNPs/C particles under 473 nm laser irradiation

Using OT-SERS, we investigated the photochemical reactions of individual 4-MBA/AgNPs/C particles under multi-wavelength laser irradiation and captured real-time molecular transformations.

Fig. 3 shows the temporal OT-SERS spectra of 4-MBA/AgNPs/C particles under 473 nm laser irradiation. At the beginning of the spectral collection, the SERS spectrum of the 4-MBA/AgNPs/C particle only showed peaks at 1078, 1182, 1418 and 1584 cm⁻¹. As 473 nm laser irradiation continued for up to 2000 s, the temporal SERS spectra of the 4-MBA/AgNPs/C particle exhibited changes. The intensity of the peak at 1418 cm⁻¹ increased and a new peak appeared at 1628 cm⁻¹. However, as the 473 nm laser irradiation time reached 3100 s, the intensity of the peak at 1418 cm⁻¹ decreased and the peak at 1628 cm⁻¹ completely disappeared. Meanwhile, new peaks appeared at 1262 and 1340 cm⁻¹, and both of these peaks vanished at 3700 s. As the 473 nm laser irradiation continued, two peaks of 4-MBA at 1182 and 1418 cm⁻¹ disappeared. The peak at 1418 cm⁻¹ vanished at 4000 s, followed by the peak at 1182 cm⁻¹ at 6500 s. With further 473 nm laser irradiation, a series of additional peaks were observed at 1126, 1135, 1239, 1245, 1271, 1309, 1326, 1335, 1377, 1387, 1407, 1454 and 1503 cm⁻¹. These peaks likely correspond to various intermediates or degradation products formed during the photochemical reaction of 4-MBA. The 100-second spectral integration time led to the superposition of signals from multi-stage reaction products being recorded in a single spectrum. Additionally, laser-induced thermal effects caused molecules to frequently enter and exit the SERS hotspots, causing significant peak intensity fluctuations.²⁵ These two factors together generated the complex Raman peak patterns observed.

Fig. 3 Temporal OT-SERS spectra and optical image of the 4-MBA/AgNPs/C particle under 473 nm laser irradiation. Scale bar: 10 μm.

Fig. 4 shows the Raman peak of the benzene ring of the 4-MBA/AgNPs/C particle at 1078 cm⁻¹ before and after being irradiated by a 473 nm laser for 2 hours. It can be seen that the intensity of this peak decreased significantly after 473 nm laser irradiation. This observation indicates that, under 473 nm laser irradiation, the functional groups of 4-MBA undergo significant modifications, and part of the aromatic ring has also changed.

Fig. 4 Raman peak of the 4-MBA/AgNPs/C particle at 1078 cm⁻¹ (from Fig. 3) before and after 473 nm laser irradiation for 2 h.

Fig. S3 (ESI†) shows the temporal OT-SERS spectra of three additional 4-MBA/AgNPs/C particles under 473 nm laser irradiation. It can be seen that, similar to Fig. 3, the changes in Raman peaks during the reactions are mainly observed in the range of 1000–1600 cm⁻¹, such as those at 1140, 1236, 1264, 1278, 1323, 1337, 1363, 1380, 1396, 1484 and 1666 cm⁻¹. While the overall reaction patterns were consistent, particle-to-particle variations in peak intensities and intermediate lifetimes were observed. Fig. S4 (ESI†) shows the Raman peaks of the three 4-MBA/AgNPs/C particles at 1078 cm⁻¹ before and after 2-hour irradiation with a 473 nm laser. It can be seen that the 1078 cm⁻¹ peak of all the 4-MBA/AgNPs/C particles experiences a significant decrease after 473 nm laser irradiation.

Photoreaction of individual 4-MBA/AgNPs/C particles under 589 nm laser irradiation

Fig. 5 shows the temporal OT-SERS spectra of the 4-MBA/AgNPs/C particle under 589 nm laser irradiation. Initially, the SERS spectrum exhibited characteristic peaks of 4-MBA at 1078, 1141, 1182, 1428, 1483 and 1585 cm⁻¹. As 589 nm laser irradiation progressed, the temporal OT-SERS revealed a complex and intermittent dynamic evolution.

Fig. 5 Temporal OT-SERS spectra and optical image of the 4-MBA/AgNPs/C particle under 589 nm laser irradiation. Scale bar: 10 μm.

Notably, the peak at 1428 cm⁻¹ gradually decreased, while the peaks at 999 and 1021 cm⁻¹ gradually increased. Throughout the 589 nm laser irradiation process, the peak at 1078 cm⁻¹ gradually decreased, and a series of transient peaks emerged within the 1000–1600 cm⁻¹ region, such as those at 1210, 1219, 1231, 1236, 1283, 1287, 1338, 1362, 1374, 1527, 1531, 1550, and 1630 cm⁻¹.

Fig. 6 shows the Raman peak of the 4-MBA/AgNPs/C particle at 1078 cm⁻¹ before and after 2-hour irradiation with a 589 nm laser. It can be seen that, after laser irradiation, the intensity of the peak decrease.

Fig. 6 Raman peak of the 4-MBA/AgNPs/C particle at 1078 cm⁻¹ (from Fig. 5) before and after 589 nm laser irradiation for 2 h.

Fig. S5 (ESI†) shows the temporal OT-SERS spectra of three additional 4-MBA/AgNPs/C particles under 589 nm laser irradiation. Consistent with the results in Fig. 5, all spectra exhibited similar dynamic evolution patterns, characterized by the formation and persistence of peaks at 999 and 1021 cm⁻¹, along with intermittent emergence of peaks within the 1000–1600 cm⁻¹ region. The spectral variation trends in Fig. S5b and c (ESI†) are quite similar to those in Fig. 5, while there are relatively obvious differences between Fig. S5a (ESI†) and Fig. 5. Specifically, the transient peaks in Fig. S5a (ESI†) are more pronounced, and the intensity at 1078 cm⁻¹ shows a relatively obvious decrease under 589 nm laser irradiation.

Fig. S6 (ESI†) shows the Raman peak of the three 4-MBA/AgNPs/C particles at 1078 cm⁻¹ before and after 2-hour irradiation with a 589 nm laser. In Fig. S6b and c (ESI†), the intensity of the peak decreased slightly.

Photoreaction of individual 4-MBA/AgNPs/C particles under 671 nm laser irradiation

Fig. 7 shows the temporal OT-SERS spectra of the 4-MBA/AgNPs/C particle under 671 nm laser irradiation. Initially, the SERS spectrum exhibited characteristic peaks of 4-MBA at 1079, 1143, 1185, 1430, 1485 and 1590 cm⁻¹. As 671 nm laser irradiation progressed, new peaks emerged at 999, 1021, 1285, 1333 and 1390 cm⁻¹. Once formed, these peaks persisted throughout. In addition, transient peaks were observed in only a few spectral lines, which is significantly fewer compared to the cases in Fig. 3 and 5. Throughout the detection process, the overall intensity of the Raman spectra displayed fluctuating behaviour. This may be due to the dynamic changes in the adsorption sites of molecules on the SERS substrate.

Fig. 7 Temporal OT-SERS spectra and optical image of the 4-MBA/AgNPs/C particle under 671 nm laser irradiation. Scale bar: 10 μm.

Fig. S7 (ESI†) shows the temporal OT-SERS spectra of three additional 4-MBA/AgNPs/C particles under 671 nm laser irradiation. Consistent with the observations in Fig. 7, all spectra showed peaks at 999, 1021, 1285, 1333 or 1390 cm⁻¹. Moreover, transient peaks were only observed in a small number of spectral lines.

Discussion

The results from Fig. 3 and Fig. S3 (ESI†) demonstrate that 473 nm laser irradiation induces a series of photochemical transformations in 4-MBA, characterized by the formation and degradation of various intermediates. The intensities of the Raman peak representing the benzene ring at 1078 cm⁻¹ showed a significant decrease after 2 hours of 473 nm laser irradiation, indicating that a ring-opening reaction has occurred in part of the aromatic framework of the benzene ring.

The Raman peaks at 1418 cm⁻¹ and 1628 cm⁻¹ are attributed to ν(COO–) and ν(C=O) in 4-MBA, respectively.²⁴ As the 473 nm laser irradiation progresses, these two Raman peaks completely disappear, which means that the carboxyl group on 4-MBA is completely consumed. Subsequently, the signal from the δ(CH) on the benzene ring at 1182 cm⁻¹ also gradually weakens and then disappears, indicating that an oxidation reaction occurs on the benzene ring and some hydrogen atoms are replaced by other functional groups.

The transient nature of intermediate peaks observed in the 1000–1600 cm⁻¹ region suggests that the 4-MBA functional groups further undergo significant modifications. This indicates a dynamic, multi-step reaction pathway. This phenomenon was also observed under 589 nm laser irradiation. Specifically, the peaks at 1126–1135 cm⁻¹ could be related to δ(CH) and ν(CO); the peaks at 1210–1309 cm⁻¹ to δ(CH), ν(ring) and ν(CO); the peaks at 1323–1407 cm⁻¹ to δ(CH), δ(OH), δ(C–OH) and ν(ring); the peaks at 1454–1503 cm⁻¹ to ν(ring) and δ(OH); and the peaks at 1527–1550 cm⁻¹ to ν(ring), δ(C–OH), and δ(CH).^26–28 Additionally, the peaks at 1628–1630 and 1666 cm⁻¹ might be attributed to ν(C=O).^24,29 We used DFT to calculate several possible reaction products, and the optimized geometric structures and coordinates of the products are presented in Table S3 (ESI†). A scaling factor of 0.97 was used for the calculations. As shown in Fig. S8 (ESI†), the results presented strong Raman bands with frequencies similar to the experimental results. The specific vibrational frequencies from the DFT calculations are presented in Table S4 (ESI†). It is postulated that, during the photooxidation reaction under 473 nm laser irradiation, 4-MBA molecules undergo hydroxylation, generating intermediates and final products containing C–OH and C=O functional groups.

These results indicate that 4-MBA in composite particles undergoes processes such as photooxidation under laser irradiation, leading to changes in their spectral signals. This process may be related to heterogeneous photoreactions between activated carbon and oxygen. Previous studies have shown that carbon plays a key role in the photooxidation of polycyclic aromatic hydrocarbons.^1,30 Under solar radiation, reactive species such as O₂˙⁻ and ˙OH are generated on the elemental carbon and activated carbon.^31–33 These radicals promote the oxidation and decomposition of organic molecules, resulting in the formation of products such as phenols, ketones, aldehydes, and inorganic small molecules.^3,34

Under 589 nm laser irradiation, transient peaks were also observed in the 1000–1600 cm⁻¹ region, but the intensity of their changes was weaker than that under 473 nm laser irradiation. This means that the photooxidation reaction induced by the activated carbon under 589 nm laser irradiation is weaker than that under 473 nm laser irradiation. From the results in Fig. S5 (ESI†), it can be seen that the photooxidation reaction induced by activated carbon in Fig. S5a (ESI†) is significantly stronger than those in Fig. 5 and Fig. S5b, c (ESI†). Under 589 nm laser irradiation, the peak intensity of the benzene ring decreased, particularly in the particle exhibiting the most intense photooxidation, as shown in Fig. S5a and S6a (ESI†). This decrease is attributed to the consumption of the benzene ring during the photooxidation reaction. This result also reflects the heterogeneity among the particles.

Under 671 nm irradiation, the transient peaks only appear in a small number of spectral lines, and the aromatic ring vibrations remain relatively stable. This phenomenon implies that the photooxidation reaction induced by the activated carbon becomes very weak.

In addition, under 589 nm and 671 nm laser irradiation, the peaks at 999, 1021, 1285, 1333 and 1390 cm⁻¹ are derived from benzenethiol formed by the plasmon-induced decarboxylation of 4-MBA.^22,35 Since the C–S bond and C–C stretching band in both 4-MBA and benzenethiol are similar, we used the peak around 1078 cm⁻¹ as an internal standard to quantify Raman intensity. As shown in Fig. S9 and S10 (ESI†), except for the particle with intense photooxidation in Fig. S9b (ESI†), the normalized intensity of the peaks at 999 and 1021 cm⁻¹ increases. In contrast, the variation trend of the normalized peak intensity around 1428 cm⁻¹ (ν_s(COO–)) is exactly opposite to that of the normalized peak intensities at 999 and 1021 cm⁻¹. This further confirms that the peaks at 999 and 1021 cm⁻¹ are related to the benzenethiol formed after the decarboxylation of 4-MBA.

The above results show that the photooxidation reaction of 4-MBA/AgNPs/C under these laser irradiations is significantly different, and the reactions are more intense under shorter-wavelength laser irradiation, which may be related to photon energy. The photon energy of a 473 nm laser (2.62 eV) is close to or slightly higher than the bandgap of activated carbon (1.98–2.83 eV),³⁶ which can effectively excite electron–hole pairs, thereby generating more reactive oxygen species to promote the oxidative decomposition of 4-MBA. The photon energy of a 589 nm laser (2.11 eV) may only partially meet the band gap requirement of activated carbon, resulting in a decrease in carrier yield. The photon energy of a 671 nm laser (1.85 eV) is much lower than the band gap of activated carbon and cannot effectively excite electron transitions, resulting in insufficient generation of carriers. Fig. 8 shows the proposed decomposition mechanism of 4-MBA under 473 nm laser irradiation.

Fig. 8 Schematic of the decomposition mechanism of 4-MBA under 473 nm laser irradiation.

Conclusions

This study demonstrates the capability of OT-SERS for investigating interfacial photochemical processes on individual suspended aerosol particles. By designing a composite aerosol model (4-MBA/AgNPs/C), we successfully monitored wavelength-dependent photochemical reactions. Under 473, 589 or 671 nm irradiation, photooxidation processes induced by activated carbon were observed, leading to aromatic ring cleavage and intermediate formation. These processes are evidenced by the attenuation of the 1078 cm⁻¹ peak and the formation of transient intermediates. The photooxidation processes on 4-MBA/AgNPs/C exhibited strong wavelength dependence, with significantly stronger responses observed under 473 nm laser irradiation than under 589 nm and 671 nm laser irradiation. This study establishes OT-SERS as a powerful platform for probing the interfacial photochemical processes in individual aerosol particles.

Author contributions

Conceptualization: H. Z., J. C., J. L.; methodology: H. Z.; validation: J. C., J. L.; formal analysis: H. Z., investigation: H. Z., Y. H., J. C., T. S.; resources: J. C., Y. H., J. C., T. S. J. L.; data curation: H. Z.; writing – original draft: H. Z.; writing – review & editing: J. C.; supervision: J. C., J. L.; project administration: J. C., J. L.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 22102162).

Notes and references

1. Y. Zhu, J. Li, Y. Zhang, X. Ji, J. Chen, D. Huang, J. Li, M. Li, C. Chen and J. Zhao, Environ. Sci. Technol., 2024, 58, 11578–11586.
2. C. Han, Y. Liu and H. He, Atmos. Environ., 2013, 64, 270–276.
3. M. Li, J. Li, Y. Zhu, J. Chen, M. O. Andreae, U. Pöschl, H. Su, M. Kulmala, C. Chen, Y. Cheng and J. Zhao, Chem, 2022, 8, 2688–2699.
4. J. M. Conny and D. L. Ortiz-Montalvo, J. Geophys. Res. Atmos., 2017, 122, 9816–9842.
5. A. Ashkin, Biophys. J., 1992, 61, 569–582.
6. J. Huisken and E. H. K. Stelzer, Opt. Lett., 2002, 27, 1223–1225.
7. Z. Liang, Y. Chu, M. Gen and C. K. Chan, Atmos. Chem. Phys., 2022, 22, 3017–3044.
8. Z. Y. Gong, Y. L. Pan, G. Videen and C. J. Wang, Anal. Chim. Acta, 2018, 1020, 86–94.
9. P. P. Chang, X. Y. Gao, C. Cai, J. Ma and Y. Zhang, Spectrochim. Acta, Part A, 2020, 228, 117727.
10. J. A. Gómez Castaño, L. Boussekey, J. P. Verwaerde, M. Moreau and Y. A. Tobón, Molecules, 2019, 24, 3325.
11. E. A. Parmentier, G. David, P. C. Arroyo, S. Bibawi, K. Esat and R. Signorell, J. Aerosol Sci., 2021, 151, 105660.
12. J. Langer, D. Jimenez de Aberasturi, J. Aizpurua, R. A. Alvarez-Puebla, B. Auguie, J. J. Baumberg, G. C. Bazan, S. E. J. Bell, A. Boisen, A. G. Brolo, J. Choo, D. Cialla-May, V. Deckert, L. Fabris, K. Faulds, F. J. García de Abajo, R. Goodacre, D. Graham, A. J. Haes, C. L. Haynes, C. Huck, T. Itoh, M. Kall, J. Kneipp, N. A. Kotov, H. Kuang, E. C. Le Ru, H. K. Lee, J. F. Li, X. Y. Ling, S. A. Maier, T. Mayerhofer, M. Moskovits, K. Murakoshi, J. M. Nam, S. Nie, Y. Ozaki, I. Pastoriza-Santos, J. Perez-Juste, J. Popp, A. Pucci, S. Reich, B. Ren, G. C. Schatz, T. Shegai, S. Schlucker, L. L. Tay, K. G. Thomas, Z. Q. Tian, R. P. Van Duyne, T. Vo-Dinh, Y. Wang, K. A. Willets, C. Xu, H. Xu, Y. Xu, Y. S. Yamamoto, B. Zhao and L. M. Liz-Marzan, ACS nano, 2020, 14, 28–117.
13. M. Gen, R. Kunihisa, A. Matsuki and C. K. Chan, Aerosol. Sci. Technol., 2019, 53, 760–770.
14. V. Sivaprakasam and M. B. Hart, ACS Omega, 2021, 6, 10150–10159.
15. Y. Ai, C. Wang, Y. Pan and G. Videen, Aerosol. Sci. Technol., 2024, 59, 635–648.
16. L. Yang, Q. Sang, J. Du, M. Yang, X. Li, Y. Shen, X. Han, X. Jiang and B. Zhao, Phys. Chem. Chem. Phys., 2018, 20, 15149–15157.
17. N. G. Bastús, F. Merkoçi, J. Piella and V. Puntes, Chem. Mater., 2014, 26, 2836–2846.
18. N. G. Bastús, J. Piella and V. Puntes, Langmuir, 2016, 32, 290–300.
19. X. Zhang, D. Han, Z. Pang, Y. Sun, Y. Wang, Y. Zhang, J. Yang and L. Chen, J. Phys. Chem. C, 2018, 122, 5599–5605.
20. X. Xue, L. Chen, C. Wang, C. Zhao, H. Wang, N. Ma, J. Li, Y. Qiao, L. Chang and B. Zhao, Spectrochim. Acta, Part A, 2021, 247, 119126.
21. C. J. Orendorff, A. Gole, T. K. Sau and C. J. Murphy, Anal. Chem., 2005, 77, 3261–3266.
22. H. Yeasmin, M. Akteruzzaman, M. M. R. Mazumder, S. Y. Alfaifi, R. H. Althomali and M. M. Rahman, J. Photoch. Photobio. A, 2024, 448, 115307.
23. T. Shao, J. Xu, H. Zhong, Y. Hu and J. Chen, Talanta, 2024, 274, 126008.
24. S. Zhu, C. Fan, J. Wang, J. He and E. Liang, Appl. Phys. A: Mater. Sci. Process., 2014, 117, 1075–1083.
25. C. Zong, C. Chen, X. Wang, P. Hu, G. Liu and B. Ren, Anal. Chem., 2020, 92, 15806–15810.
26. A. Espina, S. Sanchez-Cortes and Z. Jurašeková, Molecules, 2022, 27, 279.
27. S. Sánchez-Cortés and J. V. García-Ramos, J. Colloid Interface Sci., 2000, 231, 98–106.
28. M. Tammer, Colloid Polym. Sci., 2004, 283, 235.
29. M. Srivastava, N. P. Singh and R. A. Yadav, Spectrochim. Acta, Part A, 2014, 129, 131–142.
30. M. Li, F. Bao, Y. Zhang, W. Song, C. Chen and J. Zhao, Proc. Natl. Acad. Sci. U. S. A., 2018, 115, 7717–7722.
31. J. Zhu, J. Shang and T. Zhu, Environ. Pollut., 2021, 270, 116079.
32. J. Zhu, M. Sheng, J. Shang, Y. Kuang, X. Shi and X. Qiu, Environ. Sci. Technol., 2022, 56, 7668–7678.
33. I. Velo-Gala, J. J. López-Peñalver, M. Sánchez-Polo and J. Rivera-Utrilla, Appl. Catal., B, 2017, 207, 412–423.
34. S. He, Y. Liu, M. Song, X. Li, S. Lu, T. Chen, Y. Mu, S. Lou, X. Shi, X. Qiu, T. Zhu and Y. Zhang, Environ. Sci. Technol., 2023, 57, 20657–20668.
35. M. Futamata, K. Tabei, D. Wu, P. Verma and Y. D. Suh, Proc. SPIE, 2020, 1146806.
36. S. K. Shahcheragh, M. M. B. Mohagheghi and A. Shirpay, Opt. Quant. Electron., 2023, 55, 1256.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5cp01319d

---