English

• Photochemical reactions find broad applications in diverse fields including solar energy conversion [1-3], environmental science [4,5], photodynamic therapy [6,7], etc. The deep understanding of photochemical reaction mechanism can require real time and in-situ analytical tools to characterize short-lived intermediates during reaction process [8,9]. Previous works have reported the utilizing of spectroscopy methods including transient absorption spectroscopy [10], surface-enhanced Raman spectroscopy [11,12] and X-ray photoelectron spectroscopy [13,14] to monitor the photoreaction process. While these spectroscopic methods offer high sensitivity and effective in-situ analysis for short-lived intermediates, characterizing the structure of compounds in complex reaction systems remains challenging. Compounds with similar structures often exhibit similar absorption peaks, making it difficult to distinguish between specific structures.

  Mass spectrometry (MS) is a powerful tool enabling reaction monitoring by analyzing the molecular weight of multiple compounds during reaction process [15,16]. With tandem MS (MS²) technologies, the structure of the reagents, intermediates and products could be characterized [17]. Recently, nanoelectrospray ionization (nESI) has been used to monitor photochemical reactions. The tapered glass capillary is widely used as the nESI emitter, which makes it convenient to trigger the photochemical reaction by irradiating the reaction solution loaded in the glass nESI emitter directly. For instance, Chen et al. used the laser pointer to irradiate the outer wall of the nano emitter to monitor the dehydrogenation of tetrahydroquinolines [18]. Ma et al. used a low-pressure mercury lamp to irradiate the nanopipette uniformly for tens of seconds to study the Paternò–Büchi (PB) reaction [19,20]. In these above-mentioned MS methods, the short-lived photochemical intermediates usually need seconds to minutes to transport to nESI emitter tip and sent to MS, which may limit the application in mechanism and kinetics study of fast photochemical reactions. Therefore, how to construct a photochemical reaction interface as near as possible to the MS inlet to ensure in-situ capability of MS ionization source for reaction monitoring is the key issue.

  Recently, Cheng et al. utilized a fiber laser inserting into the capillary to irradiate the deposited TiO₂ at the emitter tip to study the heterogeneous photochemical reaction [21]. The fiber laser could provide enough optical power density for in-situ monitoring the reaction in such limited volume. Here in this work, we developed an optical focusing inductive ESI (OF-iESI) platform using the emitter itself as lensed fiber (Fig. 1A and Fig. S1 in Supporting information) to monitor fast photochemical reactions. By introducing the laser (450 nm) from the rear part coaxially, the amplification of optical power density at the emitter tip could be achieved (Fig. 1B). Typically, the angle α is about 10°–20° and the length of the taper is 5 mm. Plane geometry analysis revealed that the incident angle of each reflection was relative to α and the number of reflection time m (Eq. 1):

  $ \text { Incident angle }=\frac{\pi}{2}-(2 m-1) \cdot \alpha $

  (1)

  Figure 1

  

  Figure 1.  (A) Schematic diagram of the optical focusing iESI MS platform. (B) Analysis of the optical path of a laser in the cone shape solution. The blue arrows indicate the direction and path of light propagation. Section HG (d₀) represents the diameter of the rear solution interface. The Region EFHG is designated as the total internal reflection region 1, where the laser undergoes its first total internal reflection. The Region CDFE is the total internal reflection region 2, where the laser experiences its second total internal reflection. The Region ABDC is the refraction region, where the laser escapes from emitter by refraction. (C) The photo of the emitter. (D) The curve of the ratio of cross section S₀ to S₂ with various α.

  DownLoad: Full-Size Img PowerPoint

  This result indicated that not more than two total internal re-flection occurred in a single optical path (incident angle <42°). The solution and glass are optically denser media compared to air, acting as a total internal reflection prism. When the laser is perpendicularly incident into this region, two total internal reflections can occur: the first in Region EFGH (total internal reflection region 1) and the second in Region CDFE (total internal reflection region 2). When the laser enters Region ABDC (refraction region), it no longer satisfies the conditions for total internal reflection and escapes from tube by refraction. Therefore, the amplification factor of optical intensity could be considered as the ratio of Section GH (S₀) to Section CD (S₂). The value S₀/S₂ is a function of α and the detailed verification process was shown in Fig. S2 (Supporting information). Obviously, angle α is the critical element to improve the optical focusing performance. Fig. 1D displays the change of value of S₀/S₂ vs. angle α. For our emitters, angle α was typically 13° (Fig. 1C), indicating that the optical power was focused inside the tip with a confined space of 45 nL (Region ABFE in Fig. 1B) and the optical power density could be enhanced ca. 16 times (Fig. 1D) theoretically. We also tested the actual improvement of optical power density in the taper. Distilled water (1.5 µL) was added to the emitter as well as another untapered capillary with coaxial irradiation applied. As the result, it took averagely 16.7 s to evaporate 0.36 µL water in the tapered emitter and 120.0 s to evaporate 0.24 µL water in untapered capillary. Considering the thermal capacity and latent heat of water, the actual measured optical density enhancement factor was 11. The reason for the lower measured optical density enhancement factor may be due to the divergence of laser beam, causing a small portion of the laser to transmit and refract through emitter capillary wall directly, ultimately resulting in a loss of optical power density. In this work, the power of laser source was 1.5 W with a rectangular laser spot of 5.0 mm × 1.1 mm. Therefore, the optical power density of laser was estimated to be 2.7 × 10⁵ W/m² and this value could be amplified to ca. 3.0 × 10⁶ W/m² at emitter tip.

  The Paternò–Büchi (PB) reaction has been widely used in the field of organic synthesis and lipids analysis [22-26] however, the order of reaction of triplet energy transfer based PB reaction [27,28] remains unknown. We therefore used OF-iESI platform to study the kinetics of this reaction. The Dip&Go method was applied to load small volume samples (<100 nL) to emitters [29] and perform the MS analysis. Furthermore, we tried to capture the short-lived donor-acceptor collision complex intermediates by this platform.

  We first evaluated the in-situ capability of OF-iESI platform for photochemical reaction monitoring. To avoid the thermal effect caused by continuous laser irradiation, a transistor-transistor logic (TTL) controller was applied to control the pulse width, frequency and duty cycle of the laser. The coaxial irradiation setting of OF-iESI was shown in Fig. 2A. The Dip&Go method was utilized to load nanoliter scale samples to the emitter for iESI analysis. As demonstrated in Fig. 2A, we utilized a TTL controller to modulate the laser irradiation pulse width to control the reaction time of photochemical reactions. The pules frequency of the laser was 1 Hz and the pules width was set from 1 ms to 500 ms according to the needs of different experiments. The photochemical reaction was triggered by a single pulse. The dehydrogenation of tetrahydroquinoline (THQ) was examined as an example to evaluate performance of OF-iESI for photochemical reaction monitor [18,21]. The lifetime of the radical species of THQ dehydrogenation is reported at sub 10 microseconds level (Fig. 2B) [30]. In our experiment, about 50 nL reaction solution containing 1 mmol/L THQ and 10 µmol/L Ir[dFppy]₂(dtbbpy)PF₆ was added to the emitter. A voltage of 1.5 kV was applied to trigger the electrospray, and the iESI was always on during the reaction monitoring. The coaxial irradiation with a pulse width of 50 ms (reaction time) was applied to trigger the dehydrogenation of THQ. Fig. 2C displayed the mass spectrum of the THQ reaction before irradiation. The [M+H]⁺ signal of THQ (detected: m/z 134.0960, calculated: m/z 134.0965) was clearly observed and no other radical species were detected. After irradiation, the radical species THQ^+• (detected: m/z 133.0885, calculated: m/z 133.0886) and dihydroquinoline^+• ([DHQ]^+•, detected: m/z 131.0731, calculated: m/z 131.0730) were successfully observed by OF-iESI. In addition, the DHQ (detected: m/z 132.0808, calculated: m/z 132.0808) and quinoline (detected: m/z 130.0651, calculated: m/z 130.0652) were also detected (Fig. 2D). These results showed that the radical species intermediates in the reaction process of the dehydrogenation of THQ (Fig. 2B) can be captured and analyzed by OF-iESI, suggesting that OF-iESI is capable for monitoring microseconds level photochemical reaction intermediates.

  Figure 2

  

  Figure 2.  (A) Schematic of the coaxial irradiation method used in optical focusing iESI MS platform and the photochemical reaction analysis process. (B) The proposed reaction mechanisms of Ir[dFppy]₂(dtbbpy)PF₆ catalytic dehydrogenation of tetrahydroquinoline. (C) Mass spectra of reaction solution before irradiation. (D) Mass spectra of reaction solution after irradiation of 50 ms.

  DownLoad: Full-Size Img PowerPoint

  The above results have proved the in-situ capability of the OF-iESI platform. To further demonstrate the advantages of the OF-iESI platform, we used PB reaction as a model reaction to compare the reaction yields triggered by different irradiation ways (Fig. S3 in Supporting information). Considering that the laser pulse could be controlled precisely, the OF-iESI platform was a potential tool for reaction kinetics study. PB derivation is a widely used strategy for lipid C═C double bond location. Recently, triplet energy transfer based PB reaction was introduced to distinguish the cis and trans isomers [22]. However, the kinetics of this reaction remains unknown. Therefore, we tried to use OF-iESI platform to investigate the kinetics of triplet energy transfer based PB reaction. The methyl benzoylformate (MBF) - oleic acid - Ir[dFppy]₂(dtbbpy)PF₆ PB reaction was chosen to study (Fig. 3A). To determine the concentration of oleic acid after PB reaction, stearic acid was chosen as the internal standard. We first evaluated the linear range of the concentration of oleic acid. The result demonstrated that the concentrations from 1 µmol/L to 50 µmol/L had good linear relationship (Fig. S4 in Supporting information). We also found that the signal intensity of oleic acid (I_o) was close to the intensity of stearic acid (I_s) when the ratio between the concentration of oleic acid and the concentration of stearic acid was 1:5 (Fig. S5 in Supporting information). Therefore, we loaded 100 nL reaction solution containing 50 µmol/L MBF, 50 µmol/L (C₀) oleic acid, 250 µmol/L stearic acid (C_s) and 1 µmol/L Ir[dFppy]₂(dtbbpy)PF₆ to the emitter by Dip&Go strategy to study the kinetics of PB reaction. The internal standard calibration factor was established as f = 0.201. And C₀–C can be determined as follows (Eq. 2):

  C0−C=f·Cs·IoIs

  (2)

  where C₀–C and C were the concentrations of oleic acid and PB reaction products after PB reaction, respectively. Before the kinetics study, we first investigated the influence of temperature of the mass spectrometer capillary on the PB reaction. The capillary was set at 275 ℃ and 120 ℃ respectively. The irradiation time of 0.1 s was applied, and the results demonstrated that the capillary temperature has no significant influence on the PB reaction (Fig. S6 in Supporting information). Subsequently, the MS² analysis was applied to confirm the oleic acid and stearic acid, and the irradiation time of 0.1, 0.2, 0.3, 0.4 and 0.5 s were applied and the I_o (at m/z 281, [M−H]⁻) and I_s (at m/z 283, [M−H]⁻) at negative mode were recorded respectively (Fig. S7 in Supporting information). The positive mode mass spectra were shown in Fig. S8 (Supporting information). We tried to fit the kinetics curve according to 0^th-order, 1^st-order and 2^nd-order reaction mechanisms and found the curve of 2nd-order showed the best R² (Fig. 3B). According to the kinetics experiments, we proposed a model to illustrate the reaction process (Eqs. 3–6):

  $ [\mathrm{MBF}]^*+\text { Lipid } \xrightarrow[k_1]{\text { Rate-determining step }} \mathrm{PB} \text { reacton products } $

  (3)

  [Ir(III)]*+[MBF]→k2[Ir(III)]*+Ir(III)

  (4)

  Ir(III)→hvk3[Ir(III)]*

  (5)

  dCpbdt=k1[MBF]*[Lipid]=k1k2k3[MBF][Lipid][Ir(III)]

  (6)

  Figure 3

  

  Figure 3.  (A) Schematic of the PB reaction system. (B) Kinetics curves fitting of the triplet energy transfer based PB reaction according to 0^th-order, 1^st-order and 2^nd-order reaction mechanisms. C₀ refers to the origin concentration of oleic acid (50 µmol/L), and C refers to the concentration of PB reaction product.

  DownLoad: Full-Size Img PowerPoint

  Ir(Ⅲ) is the catalyst, therefore, the concentration of Ir(Ⅲ) is low. And (Eq. 6) could be rewritten as (Eq. 7):

  dCpbdt=k1[MBF]*[Lipid]=k1k2k3[MBF]

  (7)

  where C_pb represents the concentration of the PB reaction products. The triplet energy transfer based PB reaction followed the pseudo-second-order kinetics. The rate-determining step is the combination of MBF^* and lipids. The optical power density might influence the initial generation of photocatalytic species Ir[dFppy]₂(dtbbpy)PF₆^* and finally increase the concentration of the downstream MBF^*. The kinetics study revealed that amplified optical power density was highly possible to enhance the intrinsic rate constant of PB reaction by promoting the generation of initial photocatalytic active iridium species Ir[dFppy]₂(dtbbpy)PF₆^* to accelerate the PB reaction process.

  Previous works have demonstrated that the key step of triplet energy transfer is the formation of donor-acceptor collision complex. We used OF-iESI MS platform to monitor the [Ir[dFppy]₂(dtbbpy)PF₆-MBF]^* collision complex (Fig. 4A). In this section, 100 nL reaction solution containing 50 mmol/L methyl benzoylformate, 1 mmol/L oleic acid and 0.5 mmol/L Ir[dFppy]₂(dtbbpy)PF₆ was used and the irradiation time of 100 ms was applied. The target collision complex was successfully captured by high resolution mass spectrometer. The isotope distribution of the experimental MS is consistent with that of the calculated one (Figs. 4A and B). Time dependent profile was carried out to further monitor the change of this signal and prove the reliability. The signal of [Ir[dFppy]₂(dtbbpy)PF₆-MBF]^* collision complex was observed when the laser was on and disappeared rapidly after the laser off (Fig. 4C) and the MS² of m/z 1141.27 displayed a fragment ion of Ir[dFppy]₂(dtbbpy)⁺ (Fig. S9 in Supporting information), indicating the success of detecting of donor-acceptor collision complex by MS. The above results suggested OF-iESI a powerful tool for in-situ reaction monitoring and short lifetime intermediates analysis.

  Figure 4

  

  Figure 4.  (A) The schematic of doner-acceptor collision complex and its in-silico isotope distribution mode. (B) The experimental mass spectra of the donor-acceptor collision complex. (C) Extracted ion chromatogram of the donor-acceptor collision complex during the irradiation experiments.

  DownLoad: Full-Size Img PowerPoint

  In summary, we developed an easy-to-construct OF-iESI MS platform to investigate the photochemical reaction. In this platform, the emitter acted as a lensed fiber to transmit and focus the light to improve the optical power density in the emitter tip, which improves the in-situ performance of nESI or iESI devices to monitor photochemical reaction. The in-situ capability of this device was illustrated by revisiting the dehydrogenation of tetrahydroquinoline. Furthermore, we used this platform to investigate the kinetics of the triplet energy transfer based PB reaction. The results revealed that the reaction followed a pseudo-second-order reaction. Using this tool, the short-lived [Ir[dFppy]₂(dtbbpy)PF₆-MBF]^* collision complex was successfully detected. We believe that this platform is a useful tool for fast photochemical reaction kinetics study and in-situ reaction monitoring.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Yongyi Li: Writing – review & editing, Writing – original draft, Conceptualization. Jin Han: Validation. Xiangyu Wang: Validation. Zhenwei Wei: Writing – review & editing, Writing – original draft, Funding acquisition, Conceptualization.

  Acknowledgments

  This research was financially supported by the National Natural Science Foundation of China (Nos. 22104112 and 22374110) and the Fundamental Research Funds for the Central Universities.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110708.

  
  
• 1. [1]

     K. Yang, M. Lv, Y. Chang, et al., Chin. Chem. Lett. 35 (2024) 109018. doi: 10.1016/j.cclet.2023.109018

  2. [2]

     F.A. Chowdhury, M.L. Trudeau, H. Guo, et al., Nat. Commun. 9 (2018) 1707. doi: 10.1038/s41467-018-04067-1

  3. [3]

     S. Shen, Y. Mi, Y. Ouyang, et al., Angew. Chem. Int. Ed. 62 (2023) e202316495. doi: 10.1002/anie.202316495

  4. [4]

     J. Zhou, H. Liu, H. Wang, Chin. Chem. Lett. 34 (2023) 107420. doi: 10.1016/j.cclet.2022.04.018

  5. [5]

     Q. Yang, H. Qian, Y. Guo, et al., Environ. Sci. Technol. 57 (2023) 20781–20791. doi: 10.1021/acs.est.3c05552

  6. [6]

     J. Feng, X. Tian, R. Cao, et al., Chin. Chem. Lett. 35 (2024) 109657. doi: 10.1016/j.cclet.2024.109657

  7. [7]

     Y. Li, Z. Gao, X. Yue, et al., Chin. Chem. Lett. 35 (2024) 109133. doi: 10.1016/j.cclet.2023.109133

  8. [8]

     F. Strieth-Kalthoff, M.J. James, M. Teders, et al., Chem. Soc. Rev. 47 (2018) 7190–7202. doi: 10.1039/c8cs00054a

  9. [9]

     A. Cubillas, S. Unterkofler, T. Euser, et al., Chem. Soc. Rev. 42 (2013) 8629–8648. doi: 10.1039/c3cs60128e

  10. [10]

      M. Teders, C. Henkel, L. Anhäuser, et al., Nat. Chem. 10 (2018) 981–988. doi: 10.1038/s41557-018-0102-z

  11. [11]

      T. Westrich, K. Dahlberg, M. Kaviany, et al., J. Phys. Chem. C 115 (2011) 16537–16543. doi: 10.1021/jp204405h

  12. [12]

      N. Brandt, E. Keller, R. Frontiera, J. Phys. Chem. Lett. 7 (2016) 3179–3185. doi: 10.1021/acs.jpclett.6b01453

  13. [13]

      N. Feng, Q. Wang, A. Zheng, et al., J. Am. Chem. Soc. 135 (2013) 1607–1616. doi: 10.1021/ja312205c

  14. [14]

      N. Zhang, Y. Luo, Y. Chen, et al., ACS Sustain. Chem. Eng. 11 (2023) 4813–4824. doi: 10.1021/acssuschemeng.2c07515

  15. [15]

      X. Yuan, D. Zhang, C. Liang, et al., J. Am. Chem. Soc. 145 (2023) 2800–2805. doi: 10.1021/jacs.3c00037

  16. [16]

      T. Brown, H. Chen, R. Zare, et al., Angew. Chem. Int. Ed. 54 (2015) 11183–11185. doi: 10.1002/anie.201506316

  17. [17]

      J. Chen, X. Wang, X. Cui, et al., Angew. Chem. Int. Ed. 62 (2023) e202219302. doi: 10.1002/anie.202219302

  18. [18]

      S. Chen, Q. Wan, A. Badu-Tawiah, et al., Angew. Chem. Int. Ed. 55 (2016) 9345–9349. doi: 10.1002/anie.201603530

  19. [19]

      X. Ma, Y. Xia, Angew. Chem. Int. Ed. 53 (2014) 2592–2596. doi: 10.1002/anie.201310699

  20. [20]

      X. Ma, L. Chong, R. Tian, et al., Proc. Natl. Acad. Sci. U. S. A. 113 (2016) 2573–2578. doi: 10.1073/pnas.1523356113

  21. [21]

      Q. Cheng, T. Wang, J. Hu, et al., Anal. Chem. 95 (2023) 14150–14157. doi: 10.1021/acs.analchem.3c03494

  22. [22]

      G. Feng, M. Gao, L. Wang, et al., Nat. Commun. 13 (2022) 2652. doi: 10.1038/s41467-022-30249-z

  23. [23]

      H. Shi, Z. Tan, X. Guo, et al., Anal. Chem. 95 (2023) 5117–5125. doi: 10.1021/acs.analchem.3c00085

  24. [24]

      C. Sun, C. Ma, L. Li, et al., Chin. Chem. Lett. 33 (2022) 2073–2076. doi: 10.1016/j.cclet.2021.08.034

  25. [25]

      X. Ma, W. Zhang, Z. Li, et al., Acc. Chem. Res. 54 (2021) 3873–3882. doi: 10.1021/acs.accounts.1c00419

  26. [26]

      Z. Qiu, L. Long, Z. Zhu, et al., ACS Catal. 12 (2022) 13282–13291. doi: 10.1021/acscatal.2c03879

  27. [27]

      K. Rykaczewski, C. Schindler, Org. Lett. 22 (2020) 6516–6519. doi: 10.1021/acs.orglett.0c02316

  28. [28]

      J. Zheng, X. Dong, T. Yoon, Org. Lett. 22 (2020) 6520–6525. doi: 10.1021/acs.orglett.0c02314

  29. [29]

      Z. Wei, Z. Xie, R. Kuvelkar, et al., Angew. Chem. Int. Ed. 58 (2019) 17594–17598. doi: 10.1002/anie.201909047

  30. [30]

      J. Hu, T. Wang, W. Zhang, et al., Angew. Chem. Int. Ed. 60 (2021) 18494–18498. doi: 10.1002/anie.202106945

• 

  Figure 1  (A) Schematic diagram of the optical focusing iESI MS platform. (B) Analysis of the optical path of a laser in the cone shape solution. The blue arrows indicate the direction and path of light propagation. Section HG (d₀) represents the diameter of the rear solution interface. The Region EFHG is designated as the total internal reflection region 1, where the laser undergoes its first total internal reflection. The Region CDFE is the total internal reflection region 2, where the laser experiences its second total internal reflection. The Region ABDC is the refraction region, where the laser escapes from emitter by refraction. (C) The photo of the emitter. (D) The curve of the ratio of cross section S₀ to S₂ with various α.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) Schematic of the coaxial irradiation method used in optical focusing iESI MS platform and the photochemical reaction analysis process. (B) The proposed reaction mechanisms of Ir[dFppy]₂(dtbbpy)PF₆ catalytic dehydrogenation of tetrahydroquinoline. (C) Mass spectra of reaction solution before irradiation. (D) Mass spectra of reaction solution after irradiation of 50 ms.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) Schematic of the PB reaction system. (B) Kinetics curves fitting of the triplet energy transfer based PB reaction according to 0^th-order, 1^st-order and 2^nd-order reaction mechanisms. C₀ refers to the origin concentration of oleic acid (50 µmol/L), and C refers to the concentration of PB reaction product.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (A) The schematic of doner-acceptor collision complex and its in-silico isotope distribution mode. (B) The experimental mass spectra of the donor-acceptor collision complex. (C) Extracted ion chromatogram of the donor-acceptor collision complex during the irradiation experiments.

  下载: 全尺寸图片 幻灯片
•