pH-Dependent Water Clusters in Photoacid Solution: Real-Time Observation by ToF-SIMS at a Submicropore Confined Liquid-Vacuum Interface

Water clusters are ubiquitously formed in aqueous solutions by hydrogen bonding, which is quite sensitive to various environment factors such as temperature, pressure, electrolytes, and pH. Investigation of how the environment has impact on water structure is important for further understanding of the nature of water and the interactions between water and solutes. In this work, pH-dependent water structure changes were studied by monitoring the changes for the size distribution of protonated water clusters by in-situ liquid ToF-SIMS. In combination with a light illumination system, in-situ liquid ToF-SIMS was used to real-time measure the changes of a light-activated organic photoacid under different light illumination conditions. Thus, the proton transfer and pH-mediated water cluster changes were analyzed in real-time. It was found that higher concentration of free protons could lead to a strengthened local hydrogen bonding network as well as relatively larger protonated water clusters in both organic acid and inorganic acid. Besides, the accumulation of protons at the liquid-vacuum interface under light illumination was observed owing to the affinity of organic molecules to the low-pressure gas phase. The application of in-situ liquid ToF-SIMS analysis in combination with in-situ light illumination system opened up an avenue to real-time investigate light-activated reactions. Besides, the results regarding water structure changes in acidic solutions showed important insights in related atmospheric and physiochemical processes.


INTRODUCTION
Proton transfer (PT) along hydrogen-bond (HB) network in water plays a key role in multiple physical, chemical, and biological processes, such as acid-base reactions (Mohammed et al., 2005), organic synthesis (Guo et al., 2016), electro-catalysis (Badalyan and Stahl, 2016), and biological redox processes (Di Luca et al., 2017). According to the Grotthuss mechanism, protons transfer in the form of hydronium (H 3 O + ) or hydroxide (OH − ) by forming HBs with adjacent water molecules in aqueous solutions, resulting in the formation of protonated or hydroxide water clusters with different size and structure (Agmon, 1995). This mechanism explained the anomalously high mobility of hydronium than other ions (Miyazaki et al., 2004;Headrick et al., 2005;Natarajan et al., 2015;Chen et al., 2018;Daldrop et al., 2018). The ultrafast proton transfer via HB network is the basic of biological processes (Garczarek and Gerwert, 2006;Park et al., 2012). Water cluster is the dominant species in most biological systems and its size, structure, and structural transformation determine the dynamics of biological protons (Mohammed et al., 2007). Additionally, the protonated water cluster as a specific acid with unique proton transfer properties have a significant impact on catalysis. Thus, for better understanding of proton transfer processes in water, it is necessary to investigate the properties of water cluster in different environments (Ellmer et al., 2014;Gould et al., 2020). It was known that the structure of water clusters are sensitive to the environment, such as confined space, interface, temperature, pressure, electrolytes, and pH (Burnham et al., 2006;Li and Lazaridis, 2006;Wang et al., 2008). Thus, identifying the environmental effects on the dynamic and structural properties of water clusters at molecular level is of significant importance for better understanding the nature of water and the interactions between water and solutes.
Much attention has been paid to study how the temperature and electrolytes have effects on the dynamics of HB network and water cluster structure (Stirnemann et al., 2013;Zhao et al., 2015). However, the study of acid effect on HB network and water cluster structures, which is fundamental to some important chemical and physicochemical processes in the atmosphere and biological systems, are rare (Gutberlet et al., 2009). Investigation of PT process in acid solutions seemed more complicated due to the strong ability of acid to transfer protons to water and this process is largely solvation-dependent. So far, most researches regarding acid-water cluster interactions focused on the water cluster-assisted acid dissociation (Lee et al., 1996;Ding et al., 2003;Gutberlet et al., 2009;Li et al., 2019). But how the dissociated protons from acid affect the protonated water cluster structures is hardly known. Besides, most previous studies investigated the interactions between water clusters and inorganic acids, such as HCl, H 2 SO 4 , HF, and H 2 S. Nevertheless, the interaction between water clusters and organic acids, which is of significant importance for some key physiological processes, is little reported owing to the complexity of organic molecular structures.
Photoacids are a class of proton-containing organic compounds that undergo photo-induced proton dissociation and thermal reassociation (Johns et al., 2014). Upon irradiation, the acidity of the solution could become stronger, even to the degree of strong acids (Shi et al., 2011). Once the light is turned off, the conjugate base would be protonated to regenerate photoacids, showing excellent reversibility of the proton transfer process. Thus, photoacids could be used to noninvasively control the acidity of various reaction systems, such as dynamic selfassembly of nanoparticles, and photo polymerization (Fu et al., 2016;Liao, 2017;Yucknovsky et al., 2019). More importantly, the accurate control of solution acidity makes photoacids a good candidate for real-time and in-situ experimental study of proton transfer and the interaction between water and protons.
For decades, extensive experimental and computational studies have been used to investigate the water structure dynamics and the PT mechanism in aqueous solutions, including NMR spectroscopy (Meiboom, 1961), dielectric spectroscopy (Marcus and Hefter, 2006), ultrafast vibrational dynamic spectroscopies (Park and Fayer, 2007;Shalit et al., 2017), Femtosecond elastic second harmonic scattering (fs-ESHS) (Chen et al., 2016), X-ray diffraction (XRD) (Bouazizi et al., 2006), X-ray absorption spectroscopy (XAS) (Waluyo et al., 2014), surface-sensitive Sum Frequency Generation (SFG) (Ye et al., 2001), density functional theory (DFT) (Shi et al., 2017), and molecular dynamics (MD) simulations (Chen et al., 2018). However, the results from different methods sometimes resulted in diverse interpretations. This might be attributed to the local sensitivity of these techniques to specific properties of HB networks, such as the relaxation or reorientation time of water, the intramolecular O-H stretch vibration and HB dynamics, which greatly limited the understanding of protonwater interactions from molecular level. Mass spectrometry is a powerful method for the protonated water cluster studies which can be used to monitor the water cluster size distribution under various conditions. Time-of-flight secondary ion mass spectrometry (ToF-SIMS) is a highly surface sensitive technique with high spatial and time resolutions. In combination with a microfluidic reactor, in-situ liquid ToF-SIMS analysis could be conducted to overcome the limitation of liquid sample analysis in high vacuum environment of ToF-SIMS and thus provide important spatial and temporal chemical information at the poreconfined liquid-vacuum interface (Yang et al., 2011;Liu et al., 2018Liu et al., , 2019.
In this work, in-situ liquid ToF-SIMS in combination with a light illumination system was employed to real-time monitor the proton transfer between a light-triggered photoacid and water. The dynamics of the photoacid dissociation was measured. Besides, the effect of dissociated protons from photoacid on the dynamic of HB network and water cluster structure was detected in real-time. By comparing the size distribution of organic photoacid and inorganic acid solutions, the accumulation of organic photoacid molecules at the pore-confined liquid-vacuum interface was observed and interpreted.

Synthesis of the Photoacid
A photoacid with protonated merocyanine (MEH) structure was synthesized. Scheme of the synthesis process was shown in Supplementary Figure 1. pr-MEH was synthesized following a literature method (Mason et al., 2005). 2,3,3-trimethylindolenine (1.59 g, 10 mmol) was added to propane sultone (1.22 g, 10 mmol). The mixture was stirred at 90 • C overnight under nitrogen atmosphere. The crude product was filtered, washed with cold diethyl ether, and dried in vacuo to get pr-MEH (2.50 g, 85%). For the synthesis of MEH, pr-MEH (1 g, 3.6 mmol) and 2-hydroxybenzaldehyde (0.48 g, 3.9 mmol) were added into anhydrous ethanol (15 mL). The mixture was allowed to reflux overnight. The orange solid of MEH was obtained by filtration (1 g

Fabrication of the Microfluidic Reactor
Fabrication of the microfluidic reactor was described in our previous paper . Briefly, a micro-chamber was made by pouring PDMS on a silicon mode fabricated by soft lithography. The as-prepared PDMS block was sealed with a silicon-framed silicon nitride window via air plasma. An inlet and outlet were drilled through the PDMS block. After injection of sample solution, the two ends were sealed by a PEEK union. After that, the microfluidic reactor was introduced into the high vacuum chamber for subsequent in-situ liquid ToF-SIMS measurements.

Hybrid Light/ToF-SIMS Instrumentation
A ToF-SIMS V spectrometer (IONTOF GmbH, Germany) equipped with a 30 keV Bi + 3 primary ion beam was used for in-situ liquid ToF-SIMS analysis. For in-situ liquid ToF-SIMS analysis, the target current, and lateral resolution were adjusted to 0.35 pA and 200 nm, respectively. As shown in Supplementary Figure 5A, a 100 nm thick SiN membrane supported on a silicon frame (window size 0.5 × 0.5 mm 2 ) was irreversibly bonded with the PDMS block containing a 200 × 300 µm (width × depth) channel to form the detection area. For sampling, a liquid-vacuum interface was formed by drilling a ∼2 µm pore through the SiN membrane by primary ion beam before continuous signal recording of the liquid-vacuum interface. The dynamic depth profiling by ToF-SIMS for 0.5 mM MEH solution in the positive mode was used to monitoring the submicropore condition at the SiN membrane surface. As presented in Supplementary Figure 5B, SiN membrane was punched through at around 295 s with the dramatically increased intensity of [MEH+H] + and decreased signal of Si 2 N + . The ultrahigh surface tension of liquid water in a sub micrometer pore can confine the liquid water in the microfluidic chip. Thus, a liquid-vacuum interface used for the in-situ liquid detection is formed under high vacuum condition. When the liquidvacuum interface at the confined pore was sable, the pulse width was adjusted from 160 to 80 ns immediately for better mass resolution. Then the mass spectrum signal of the water clusters from 0.5 mM MEH solution surface was acquired in the microfluidic reactor (see Supplementary Figure 5C). During analysis, a flashlight (Skyfire, white light with an optical filter at 420 nm) with a power of 12 W was used to irradiate the microfluidic chip. By adjusting the flashlight, real-time ToF-SIMS analysis of light-activated reaction was realized. The obtained ToF-SIMS spectra were calibrated using C + , CH + 2 , C 2 H + 5 , C 3 H + 3 for positive mass spectra and C − , CH − 2 , C − 2 , C − 3 for negative mass spectra, respectively. The two-dimensional (2D) images of Si + , [MEH+H] + and (H 2 O) 3 H + can be used to estimate the pore size. Besides, in order to improve the signal to noise (S/N) of mass spectrum, the mass spectra data could be reconstructed at the selected ROI area of the pore center in 2D images (Supplementary Figure 5D).

DFT Simulation
The calculations in this work were performed by the Gaussian 09 suite of programs (Frisch, 2009). The geometries of the protonated water clusters under study were fully optimized by the hybrid M06-2x functional, in which developed by Zhao and Truhlar, has proved to be reliable in the description of various types of non-covalent interactions (Zhao and Truhlar, 2008). The Dunning's basis set aug-cc-pVTZ, was utilized for all the atoms (Kendall et al., 1992). No symmetry or geometry constraint was applied during the optimizations. All the optimized-geometries were corroborated to be factual minima on the potential energy surface by means of frequency calculation at the same theoretical level.
The interaction energy ( E) between hydronium ion and waters for per water molecule was evaluated as Where E system is the energy of the optimized whole system; E H3O+ and E nH2O are the total electronic energies of the hydronium ion and waters, respectively, which kept in the same structure as that in the optimized whole system. The n denotes the number of waters.

ToF-SIMS Real-Time Analysis of the Light-Activated Proton Dissociation
A photoacid with protonated merocyanine (MEH) structure was used as a model to in-situ generate free protons in aqueous solutions. Based on previous studies (Shi et al., 2011;Liao, 2017), MEH was the first reported metastable-state photoacid.
It was known that upon illumination of blue light, one proton of MEH could be released and the MEH will turn into SP. When the light was turned off, the less stable SP will convert back into MEH ( Figure 1A). The mechanism of light-activated proton release is described in Supplementary Figure 6. Upon irradiation with visible light (420 nm), MEH firstly convert into the cis-MEH and then cis-ME after releasing a proton. Once the proton was transferred, cis-ME wound turn into SP via a nucleophilic ring closing reaction. SP tends to return to its original state (MEH) immediately in the dark, which means the whole process is reversible. To directly monitor the light-activated proton dissociation, in-situ liquid ToF-SIMS in combination with a 420 nm light source was used. A ∼2 µm pore on the SiN membrane of the microfluidic cell was drilled in-situ by the primary ion beam of ToF-SIMS. Thus, a pore-confined liquid-vacuum interface was formed for light illumination as well as direct ToF-SIMS measurement ( Figure 1A). As shown in Figure 1B, the intensity ratio between proton (m/z 1) and MEH (m/z 384, [M-H] − ) started to increase linearly upon light illumination. When the light was turned off at ∼55 s, the ratio started to decrease. It takes about 35 s for the ratio to reach the baseline, which is quite similar to the light illumination process (from 20 to 55 s). The results indicated that the kinetic constant of proton dissociation and recombination process were similar. In addition, the result demonstrated the capability of in-situ liquid ToF-SIMS for real-time monitoring of proton transfer process mediated by light.

Dynamic Change of Water Clusters Mediated by Photoacid
The dissociation of protons in aqueous acid solutions changed the micro-environment of the water structure, which could lead to the change of hydrogen bonding networks as well as the size distribution of protonated water clusters. The effect of proton released from organic photoacid on HBs and water structures was real-time measured by in-situ liquid ToF-SIMS. Before analysis of the light-mediated water cluster distribution of MEH, we firstly measured the water cluster size distribution of pure water as a control. As shown in Figures 2A-C Figure 2D). According to previous report, MEH was a weak acid with a pK a of ∼7.8. The pH of 0.5 mM MEH was ∼5.5 (under dark environment) (Gutberlet et al., 2009). The increase of free proton concentration could lead to a strengthened hydrogen bonding networks and thus increase the ratio of relatively larger water clusters. Another important reason for this change is the disturbing of organic molecules on the hydrogen bonding networks of water. It was reported that small purely hydrophobic solutes tend to strengthen nearby water hydrogen bonding networks (Grdadolnik et al., 2017). As shown in Figure 1A, the structure of MEH was relatively complicated with various hydrophobic functional groups, which could strengthen the local hydrogen bonding networks. Therefore, the ratio between larger water clusters and smaller clusters tend to increase. Here hydrochloride acid (HCl) was chosen since it was known that Cl − has negligible effects on HB networks and water clusters (Näslund et al., 2005;Liu et al., 2019). In-situ liquid ToF-SIMS measurement of the size distributions of protonated water clusters in HCl solutions under different pH were shown in Figure 3. It was clear that with the decrease of pH, the intensity ratio of (H 2 O) 4 H + gradually decreased while that for (H 2 O) 5 H + increased. When pH was decreased to one, the signal intensity of (H 2 O) 5 H + even exceeded (H 2 O) 3 H + , which was the dominant protonated water cluster under neutral pH. That demonstrated the formation of relatively larger protonated water clusters in concentrated acid solution. This is in good agreement with previous observations that an increase of the acid concentration lead to the increase of the average number of donating HBs per water molecule by X-ray Raman scattering (XRS) and XAS (Cavalleri et al., 2006;Chen et al., 2013). However, in those studies, it was not until the concentration of hydrochloric acid increased to several M that the spectra started to change. Our results provided direct molecular evidence that the structure of water was changed even when the concentration of hydrochloric acid was lower than 0.1 M.
For real-time monitoring the dynamic process of protonated water cluster size changes with time, the mass spectrum data was reconstructed every 20 s to obtain the mass spectrum peak area of water cluster signals. The normalized intensity changes of each water cluster ((H 2 O) n H + , n = 1-6) with time was presented in Figure 4. The dominant protonated water cluster in MEH solution under dark environment was (H 2 O) 4 H + (Region I in Figure 4). Once the light was turned on (Region II), the ratio of smaller water clusters of (H 2 O) n H + (n = 1-4) decreased immediately with the increased concentration of proton, while the larger ones of (H 2 O) 5 H + and (H 2 O) 6 H + increased, especially for (H 2 O) 5 H + , whose ratio is about 1.7 times that in dark environment. Interestingly, when the light was turned off, the concentration of proton decreased, the large water clusters of (H 2 O) 5   into smaller water clusters of (H 2 O) n H + (n = 1-3) and then the size distribution of water clusters gradually return to the initial level (Region IV). For better understanding of the above Proton Accumulation at the Pore-Confined Liquid-Vacuum Interface An interesting phenomenon is that the size distribution of protonated water clusters in MEH under light illumination ( Figure 2E) was quite similar to that in HCl solution with pH = 1 (Figure 3D). Previous report demonstrated that even under complete proton dissociation, the pH of 0.5 mM MEH solution was 3.2 (Shi et al., 2011). However, the results in light illuminated MEH was closer to the results of HCl with pH = 1. According to the abnormal results, we speculate that in the light-mediated proton dissociation, proton tends to accumulate at the pore-confined liquid-vacuum interface. Thus, the local concentration of protons at the sampling area significantly increased, leading to a further strengthened hydrogen bonding networks as well as larger water clusters at the liquid-vacuum interface. It was well-known that the hydrophobic parts of a complicated molecule tend to approach the gas phase at a liquidgas interface (Björneholm et al., 2016). Thus, in this case, it is reasonable to hypothesis the accumulation of MEH at the liquid-vacuum interface ( Figure 5A). ToF-SIMS 3D mapping of MEH at the center of liquid-vacuum interface at the very beginning of pore formation was conducted to directly visualize the concentration change of MEH at the interface as a function of time ( Figure 5B). It was clearly observed that the concentration of MEH continuously increased along with time, which verified the accumulation of MEH at the liquid-vacuum interface. As a result, the concentration of dissociated protons significantly increased at the interface, indicating a lower pH at the interface than in the bulk.

CONCLUSION
In this work, the effects of inorganic and organic acids on water structure and hydrogen bonding networks in aqueous solutions were investigated by in-situ liquid ToF-SIMS. The results revealed that higher concentration of dissociated protons lead to an enhanced local hydrogen bonding networks as well as relatively larger protonated water clusters. In combination with a light illumination system, in-situ liquid ToF-SIMS holds the ability to real-time analyze light-activated proton dissociation and the resulting changes of protonated water clusters. In addition, the accumulation of organic photoacid and dissociated protons at the pore-confined liquid-vacuum interface caused by the affinity of organic molecules to the liquid-vacuum interface was observed. These results provided direct molecular evidence of the interaction between protons and water, which was of significance for the fundamental investigations of related atmospheric and physiochemical processes. Besides, the lightillumination system combined with in-situ liquid ToF-SIMS method shows great potential in the study of light-sensitized processes in aqueous solutions.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
XH and Y-TL designed this work. Y-YL conducted the experiments. XH and Y-YL wrote this paper. ZZ and JZ synthesized the photoacid. SZ performed the DFT simulation. PH gave constructive suggestions in addressing the reviewers' comments. All authors contributed to the article and approved the submitted version.