2024 Volume 37 Issue 3
Article Contents

Piao Xu, Jianduo Lu, Dongdong Wang, Duoduo Li, Jinyou Long, Song Zhang, Bing Zhang. Ultraviolet Photoemission Spectra of Aqueous Thymidine by Liquid-Microjet Photoelectron Spectroscopy: A Combined Experimental Error Assessment†[J]. Chinese Journal of Chemical Physics, 2024, 37(3): 439-448. doi: 10.1063/1674-0068/cjcp2402024
Citation: Piao Xu, Jianduo Lu, Dongdong Wang, Duoduo Li, Jinyou Long, Song Zhang, Bing Zhang. Ultraviolet Photoemission Spectra of Aqueous Thymidine by Liquid-Microjet Photoelectron Spectroscopy: A Combined Experimental Error Assessment†[J]. Chinese Journal of Chemical Physics, 2024, 37(3): 439-448. doi: 10.1063/1674-0068/cjcp2402024

Ultraviolet Photoemission Spectra of Aqueous Thymidine by Liquid-Microjet Photoelectron Spectroscopy: A Combined Experimental Error Assessment†

通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(4)  /  Tables(3)

Article Metrics

Article views(255) PDF downloads(0) Cited by(0)

Access History

Ultraviolet Photoemission Spectra of Aqueous Thymidine by Liquid-Microjet Photoelectron Spectroscopy: A Combined Experimental Error Assessment†

Abstract: Photoemissions from aqueous solutions are prone to elastic and inelastic electron scattering, streaming potential, and space charge effect errors; however, direct assessments of their individual contributions are still a daunting challenge. Herein, to assess how these errors in a combined experimental error manner influence on the genuine photoelectron kinetic energy (PKE) spectra, the ultraviolet PKE (or electron binding energy, eBE) spectra of aqueous thymidine at different solution flow rates and laser energies were investigated by femtosecond liquid-microjet photoelectron spectroscopy. The accurate vertical eBE value of 11.33 eV for the 1b1 orbital of liquid water was used to calibrate the observed eBE spectra of aqueous thymidine, and the corresponding combined experimental error for the eBE of the 1b1 orbital of liquid water was estimated to be about 200 meV. The average combined experimental error for the vertical eBE values of the cationic ground state D0(π−1) and the cationic first excited state D1(n−1) of aqueous thymidine was estimated to be less than 3%, i.e., in the range of several tens of meV to 200 meV, depending on different flow rates and laser energies.

    • Gas-phase time-resolved photoelectron energy spectroscopy (PES) and photoelectron imaging (PEI) have been widely applied to investigate the excited-state dynamics of isolated polyatomic molecules [1−6]. These “bottom-up” studies mainly focus on the photophysical and photochemical processes of excited states of isolated biologically relevant chromophores [1−6], allowing us to understand the excited-state relaxation mechanisms underneath DNA photoprotection and damage [6, 7]. Gas-phase studies do not consider the interactions of the solute with the solvent or protein environment in a real biomolecular system. However, the solution environment, especially, the water environment, is usually vital to biology and chemistry. Sometimes, electronic structures and excited-state relaxation dynamics are very sensitive to the solvent environment, especially the polar water solvent [7]. In light of the importance of solvent environments, studies of solution-phase excited-state dynamics are of particular interest [8−12]. Currently, fluorescence up conversion (FU) and transient absorption spectroscopy (TAS) are regarded as the main solution-phase approaches for the investigation of solution-phase excited-state dynamics of biologically relevant chromophores and DNA constituents, as well reviewed by Improta et al. [7]. Despite these approaches, they sometimes suffer from severe spectral band overlap or optical dark states [7]. In contrast, the PES is the most efficient way to discern vibronic molecular structure and to advantageously detect a dark state when implemented with the photoionization probe technique. As expected, the PES is an ideal technique to be applied to solution-phase studies.

      However, the solution-phase PES has not been well developed for a long time due to the incompatibility between highly volatile aqueous solution and high vacuum [13]. Until the late 1990s, the PES of an aqueous solution was successfully reported by Faubel et al. [14]. The liquid microjet technique, pioneered by Faubel et al. [14], effectively resolved the above incompatibility and succeeded in injecting a fast flowing and very thin liquid jet filament into the vacuum system. Nowadays, by combining a traditional magnetic-bottle PES [15] with the liquid microjet technique [14], liquid-microjet magnetic-bottle photoelectron spectroscopy (LJ-MB-PES) has been one of the prevail liquid-microjet PES [16−19]. The advantage of the LJ-MB-PES is that it does not need to overcome the problem of high-voltage electrode discharge encountered in the PEI under low vacuum conditions. Furthermore, the LJ-MB-PES has high electron collection efficiency in comparison with the electron time-of-flight (TOF) PES and the hemispherical electron analyzer PES. Despite its several advantages, the LJ-MB-PES also has a drawback about its low electron transmission efficiency for the photoelectron kinetic energy (PKE) of less than 0.5 eV [11, 20−22].

      Photoionization is a highly efficient detection technique to probe any states with different multiplicities, especially the dark states, as long as the ionization wavelength is short enough [11]. In particular, almost all the aromatic molecules or chromophores, including DNA bases, can be ionized easily by an ultraviolet (UV) multiphoton process. Therefore, UV multiphoton photoionization is the most efficient way to detect electronic structures of aromatic molecules when implemented with PES. For instance, in the LJ-MB-PES, the PKE distribution can be directly measured, and then the corresponding eBE of the molecular orbital from which the electron is ionized can be determined by eBE= nhν–PKE, where hν represents the ionization photon energy and n is the number of photons. Usually, the ionization wavelength is tunable to perform the resonantly enhanced multiphoton ionization (REMPI) scheme (LJ-MB-REMPI-PES). When the LJ-MB-PES liquid-microjet PES is implemented with the REMPI scheme, it has the advantage to separate the resonant signals from the nonresonant ones by varying the ionization wavelength, which helps to identify the origin of the ionization channels. Moreover, if a femtosecond laser pulse is used, the femtosecond LJ-MB-REMPI-PES can provide a dynamical probe of photoionization from both optically prepared bright excited states and secondarily populated bright or dark states within the pulse duration, as long as the lifetimes of these excited states are appropriate. As shown in FIG. 1, these features are utilized in our femtosecond LJ-MB-REMPI-PES in the present work.

      Photoionization processes in solutions may differ from those in the gas phase, resulting in new photoionization products or different relaxation mechanisms. As an example, ionizing damage of DNA or its constituents, especially in aqueous solution, is of particular interest. This kind of damage can come from direct electron loss (radical cation) or attachment (radical anion), as well as indirect damage induced by water-free radicals [23]. Compared with the extensive researches on the excited-state dynamics of DNA constituents (such as nucleobases, nucleosides, and nucleotides), and other chromophores by gas-phase PES and PEI, solution-phase UP and TAS [1−12], their studies in aqueous solution by the LJ-MB-PES, are an ongoing experimental and theoretical topic [11, 12, 21, 22, 24]. The excited-state relaxation mechanism for the DNA constituents in aqueous solution still remains a topic of controversy [7, 11, 24]. It is now confirmed that pyrimidine nucleobases exhibit an additional relaxation channel to an optically dark 1nπ* state from the bright 1ππ* state. However, the quantum yield (QY) for the dark 1nπ* state of thymidine nucleobases in aqueous solution, is very puzzling, which is crucial to understanding the current controversy. For example, a recent result indicates that the 1nπ* QY for aqueous thymidine (Thd) is one order of magnitude higher than those for other pyrimidine or substituted-pyrimidine nucleobases [11]. In particular, the 1nπ* state is an optically dark state, but it determines the excited-state relaxation mechanism of aqueous pyrimidine nucleobases. Therefore, as shown in FIG. 1, aqueous thymidine is selected as the target in the present work. As discussed above, we apply the femtosecond LJ-MB-REMPI-PES to provide the most efficient way to detect the 1nπ* dark state.

      Yet, in spite of many technical advances in LJ-PES, several special aspects have emerged with the implementation of LJ-PES. One special aspect is the electron scattering with solvents in solution, including elastic and inelastic electron scatterings [20, 25, 26]. Another special aspect is the electron probing depth, from which the electrons created could escape into vacuum [26, 27]. Moreover, a serious problem encountered in LJ-PES of nonconducting solutions is the charging of the microjet surface, i.e., streaming potential [14, 22, 28−30]. In addition, with similarity to gas-phase PES, the space charge effect among the photoelectrons in aqueous solution might result in a slight PKE shift when the ionized photoelectrons are huge. The space charge effect can be minimized when the laser energy is controlled to an appropriate extent [31, 32]. Notably, herein, we define the combined experimental error, including the elastic and inelastic electron scatterings, streaming potential, and space charge effect. This work aims to assess the combined experimental error by a detailed analysis of the eBE or PKE spectra of aqueous thymidine. The vertical eBE values for the cationic D0(π−1) and D1(n−1) states are known or estimated from previous experimental work [11, 33], providing us with a referred energy standard. In the experiments, the elastic and inelastic electron scatterings, streaming potential and space charge effect can be assessed by adding electrolyte salt of NaF to the solution, together with adjusting the flow rates of the solution through the microjet and the laser energies.

    II.   METHOD
    • Thymidine (≥99% purity) used in the experiments was purchased from the commercial company (Aladdin) without further purification. The sample solutions of thymidine (100 mmol/L) were prepared using ultrapure water (18.2 MΩ·cm) at room temperature. Notably, 30 mmol/L sodium fluoride (NaF, ≥99%, Aladdin) was added to minimize the streaming potential of the aqueous microjet and does not result in any photoionization signals.

      The experimental apparatus used in this work consists of a commercial femtosecond laser system and a liquid-microjet magnetic-bottle photoelectron spectrometer. Briefly, a Ti:sapphire regenerative amplifier system (Coherent Mira-Legend) outputs the fundamental wavelength of 800 nm with a repetition rate of 1 kHz and a pulse duration of 100 fs. The total output energy of 4 mJ/pulse was split into three parts, one of which was used to pump an optical parametric amplifiers (OPA) (Light Conversion, TOPAS-C) to directly output the experimental UV wavelength of 246.33 nm. The experimental energies for the UV pulse were attenuated to be on the order of several hundreds of nJ/pulse to avoid gas-phase signals. The UV pulse was softly focused on the liquid microjet with a spherical plano-convex lens (f = 300 mm). The liquid-microjet injection system was similar to that of our liquid-microjet PEI spectrometer [34]. Briefly, a home-made liquid microjet nozzle consisted of a fused silica capillary with an inner diameter (I.D.) of 25 μm. Note that the conductive graphite was carefully coated to the outer surface of the nozzle to improve the silica’s electrical conductivity. The magnetic-bottle photoelectron spectrometer was also previously described [35]. Briefly, the magnetic-bottle system was constructed by a combination of a strong permanent magnetic field (SmCo magnet, ~0.98 T at its tip) located inside the source chamber and a weak union magnetic field (solenoid coils, 10G) located inside the TOF chamber. At the end of the TOF chamber, an electron detector consisted of a gate, two V-shaped microchannel plates (MCP), and a conical anode. The flight time of the photoelectrons was output from the anode, and then preamplified to a high-speed acquisition card (MCS6, FAST ComTec). The flight time of the photoelectrons was recorded. The permanent magnet was electrically grounded. In addition, a polyetheretherketone (PEEK) tube with an inner diameter of 100 μm without capillary was used at the gas nozzle exit to introduce the gasous nitric oxide (NO) to perform a dialy calibration of the spectrometer. As a result, a relationship between the photoelectron flight time and the PKE was calibrated by NO [36]. The energy resolution was determined to be ~50 meV within 10 eV. Moreover, an estimate of the UV pulse duration was approximately 180 fs.

      In this work, based on the previous work [34], a capillary of internal diameter of 25 μm was chosen as the microjet orifice and the flow rates were controlled to be 0.4, 0.5 and 0.6 mL/min by a HPLC pump (PU-4180, JASCO). Under the working conditions described above, the vacuum in the source chamber was maintained at around 1×10−2–3×10−2 Pa, whilst the vacuum in the TOF detection chamber was maintained at around 2×10−4–3×10−4 Pa.

    • Two kinds of quantum chemical calculations were used to calculate the structures and energies of the ground, excited and cationic states of aqueous thymidine. Firstly, the polarizable continuum model (PCM) was used and was found to give a big deviation to the experimental results when we calculated the aqueous Thd. This is because the PCM does not take into account specific solute-solvent interactions like hydrogen bonds [37−39]. Therefore, the cluster model of Thd+nH2O was then chosen because this method has an appropriate computational cost and has the advantage of the inclusion of specific solute-solvent interactions, which has also been widely used in aqueous systems [7].

      The structures and energies of the ground, excited and cationic states were calculated based on the cluster model of Thd+nH2O by using the Gaussian software [40] and the Molclus program [41]. Firstly, we examined the effect of microhydration of water (i.e., the number of waters) on the energies of the ground-state Thd+nH2O by adding a conformational search and found that six water molecules can nearly represent the cluster model at an appropriate computational cost. Therefore, we chose the simple cluster model of Thd+6H2O to perform all the calculations, as shown in FIG. 2. The optimized structures and energies of the ground, excited and cationic states were calculated at different basis sets, and the responding vertical excitation energies, vertical ionization energies were summarized in the following section.

    III.   RESULTS AND DISCUSSION
    • FIG. 3 shows the eBE spectra of 100 mmol/L aqueous Thd (with an addition of 30 mmol/L NaF) at different flow rates and laser energies obtained from resonantly enhanced 1+2 PKE spectra at 246.33 nm. Notably, the eBE spectra shown in FIG. 3 have been calibrated and corrected by a combined experimental error. Thus, the eBE=3hν–PKE, where hν, representing the photon energy, corresponds to the eBE of cationic states. These spectra featured a broad eBE band within 6.5–10.5 eV, and a weak band peaked at about 11.33 eV. For the broad eBE band within 6.5–10.5 eV, all the line shapes at different flow rates and laser energies can be well fitted as a superposition of two Gaussian functions with a more intense eBE component centered at ~8.5 eV and a smaller eBE component centered at ~9.5 eV, as indicated by the two color-filled Gaussian components in FIG. 3. For the weak band, which peaks at about 11.33 eV, all the line shapes at different flow rates and laser energies can be satisfactorily fitted by one Gaussian function.

      In order to assign the observed eBE spectra of aqueous Thd shown in FIG. 3, we calculate the vertical excitation energies (VEEs) of the two lowest electronically-excited 1ππ* and 1nπ* states and the vertical ionization energies (VIEs) of the cationic ground D0(π−1) and first electronically-excited D1(n−1) states by using the PCM and the cluster model of Thd+6H2O at different theoretical levels. The corresponding values are summarized in Table I. We found that the calculated VEEs and VIEs for the cluster model of Thd+6H2O, especially the VIE for the D0(π−1), are much closer to the observed values than those for the PCM. On the basis of a significant body of previous work on aqueous Thd [11, 33] concerning the known or estimated vertical eBE(D0)=8.3 eV [33], eBE(D0)≈8.5 eV [11], eBE(D1)≈9.7 eV [33], eBE(D1)≈9.8 eV [11], we assign the eBE component centered at ~8.5 eV to the cationic ground state D0(π−1) and the eBE component centered at ~9.5 eV to the cationic first excited state D1(n−1), which correspond to the two resonant ionization channels from 1ππ*→D0(π−1) and 1nπ*→D1(n−1), respectively. Importantly, the weak band that peaks at about 11.33 eV is easily assigned to the vertical eBE value of the 1b1 orbital of liquid water.

      For a better understanding of the photoionization picture, as shown in FIG. 1, the first photon at 246.33 nm prepares the optically bright 1ππ* state, and then ultrafast internal conversion (IC) from 1ππ* to 1nπ* within the pulse duration of ~180 fs populates the low-lying dark 1nπ* state. The ensuing two-photon absorption preferentially ionizes the 1ππ* and 1nπ* states into the D0(π−1) and D1(n−1), respectively. The IC from 1ππ* to 1nπ* was recently measured to be as fast as 87 fs [11], together with earlier values of 120 fs [21] and 100 fs [42] for the 1ππ* state to relax on its surface, supporting the possibility of populating the dark 1nπ* state within the pulse duration. Therefore, the first photon at 246.33 nm is resonant with the 1ππ* and 1nπ* states during the 1+2 ionization process.

      We now focus on a detailed analysis of the combined experimental error. From our LJ-MB-PES, the eBE spectrum can be obtained indirectly by the eBE=nhν–PKE–ΔE, where PKE is the measured photoelectron kinetic energy, n is the photon number of photons (n=3 in this work), and ΔE is the combined energy error. The ΔE is a combined energy offset that is mainly caused by elastic and inelastic electron scatterings, streaming potential and space charge effect. When the ΔE has been corrected, the eBE spectrum can be represented by the normal form of eBE=nhν–PKE', where PKE' is the corrected photoelectron kinetic energy, as shown in FIG. 3.

      Firstly, the daily spectral system error is directly calibrated by using a well-known PKE of gasous NO. For our LJ-MB-PES, the PKE spectra measured in the experiment were calibrated based on the NO ionization spectra measured by Jarvis [36] and Buchner [16] and the relationship between flight time of photoelectrons and their PKE was obtained by the following equation:

      Among them, S0 represents the electron flight distance, t0 represents the time difference between the time of the data collection and electron ionization, and E0 represents the energy offset value of the PKE that results from the spectrometer. In the experiment, the PKE is calibrated by nonresonant multiphoton ionization signals of NO at 200 nm and 266.7 nm. The fitted parameters are S0 = 1.134 m, t0 = 35.99 ns, E0 = −0.16534 eV. The S0 is in good agreement with the flight tube length of 1.16 m designed for the experimental setup, allowing for a photoelectron kinetic energy resolution approximately equal to 50 meV below 10 eV when a femtosecond pulse is used. As a result, the PKE (or eBE) can be obtained, and the daily spectral system error of −0.16534 eV is corrected in the first step.

      In aqueous solution, the elastic and inelastic electron scatterings, streaming potential and space charge effect can simultaneously affect the PKE. It is still a daunting challenge to directly assess their individual contributions from experimental measurements and theoretical simulations [24, 25, 43]. Usually, the streaming potential of a liquid-microjet can be controlled by adding electrolyte salts of NaX (X=F, Cl, Br, and I) to the solution [28−30], or adjusting the flow rate of the solution through the microjet [28−30]. Salt solutions of halogens do not cause photoionization signals [11, 12, 18, 28]. Experimentally, 30 mmol/L NaF was added to minimize the streaming potential of the aqueous microjet. This addition of electrolyte salt does not cause any extra photoionization signals at UV wavelengths. Moreover, the streaming potential is inversely proportional to the flow rate of the microjet [14]. Thus, the flow rate of the microjet can be adjusted to test its influence on the streaming potential [28−30]. Experimentally, the best flow rate range for a diameter of 25 μm microjet was tested to be 0.4, 0.5, and 0.6 mL/min. When the flow rate is less than 0.4 mL/min, the laminar flow region is reduced and sometimes turbulent. When the flow rate is much larger than 0.6 mL/min, the laminar flow region is no longer apparently lengthened, however, there is a higher performance and cost requirement for the background pressure output of the HPLC pump. What is more, when the flow rate is higher, the density of the aqueous microjet is higher, resulting in a bigger ionizing signal. Sometimes, the space charge effect is thus accompanied [31, 32]. Further, the space charge effect can be more obvious when the laser energy is tuned to a much higher level. However, the laser energy is usually controlled to be on the order of several hundreds of nJ to avoid the ionization of gaseous molecules around the surface of the microjet. This laser energy results in a neglectable space charge effect. Furthermore, the elastic electron scattering does not change the photoelectron kinetic energy, whereas the inelastic electron scattering may change the photoelectron kinetic energy [20, 25, 26]. If the inelastic electron collisions are huge, the line shape of the PKE spectrum could be distorted, resulting in an asymmetrical line shape and a slight PKE shift. In the experiments, we adjusted the flow rates and the laser energies, and added the electrolyte salt NaF to solution to make a combined assessment of the elastic and inelastic electron scatterings, streaming potential and space charge effect.

      As discussed above, the PKE (or eBE) can be corrected by the daily spectral system error of −0.16534 eV in the first step. Importantly, the weak band peaked at about 11.33 eV is easily assigned to the vertical eBE value of the 1b1 orbital of liquid water, providing us with an internal energy standard [11, 25]. The accurate eBE value of 11.33 eV for the 1b1 orbital of liquid water is used to calibrate the observed eBE spectrum [11, 25]. As a result, an energy difference between the observed vertical eBE value of the 1b1 orbital of liquid water (herein, i.e., the eBE (or PKE) that was corrected by the daily spectral system error of –0.16534 eV in the first step) and the accurate eBE value of 11.33 eV can be obtained. As shown in FIG. 4, the enlarged eBE spectra for the 1b1 orbital of liquid water at different flow rates and laser energies can be satisfactorily fitted by one Gaussian function, although their signal-to-noise ratio is not good enough at the low signal level. The peak centers for the 1b1 orbital of liquid water (i.e., vertical eBE) at different flow rates and laser energies are thus extracted from the fitting values. Then the energy differences ($\Delta E_{{\rm{H}}_2{\rm{O}}}$) between these fitting values with the accurate eBE value of 11.33 eV can be obtained and summarized in Table II. As a result, the $\Delta E_{{\rm{H}}_2{\rm{O}}} $ was found to be around 0.20 eV at different flow rates and laser energies. Under each laser energy, the deviation from the accurate eBE value of 11.33 eV becomes larger and larger as the flow velocity increases from 0.4 mL/min to 0.6 mL/min. At a given flow rate, the deviation from the accurate eBE value of 11.33 eV is in the range of about 30–70 meV at different energies of 750 and 500 nJ. Therefore, the PKE (or eBE) can be further corrected by the $\Delta E_{{\rm{H}}_2{\rm{O}}} $ value at its given flow rate and given laser energy. As a result, each eBE spectrum shows accurate eBE value of 11.33 eV for the 1b1 orbital of liquid water.

      As discussed above, several special aspects have emerged with the implementation of the LJ-PES. One special aspect is the electron scattering with solvents in solution, including elastic and inelastic electron scatterings. The elastic electron scattering does not change the photoelectron kinetic energy, whereas the inelastic electron scattering may change the photoelectron kinetic energy when the inelastic electron collisions are huge. The electron inelastic mean free path (IMFP) is defined as the average length of the path traveled by an electron with a given energy between inelastic collisions in a material [25, 26]. For most materials (including aqueous solution), IMFPs exhibit a qualitatively very similar dependence (usually called as V-shaped curve) on PKE, characterized by a minimum of about 5–10 Å in the 50–100 eV PKE range with a steep increase towards the low-energy side, and a steady but slower increase for higher PKEs; ~20 Å at 1000 eV and ~30 Å at 2000 eV [25, 26]. However, it is extremely difficult to directly measure the IMFP as this would require an exact understanding of elastic and inelastic scattering processes along the electron trajectory in aqueous solution. Another special aspect is concerning on the electron probing depth, from which the created electrons could escape into vacuum. The electron probing depth, or the effective electron attenuation length (EAL), is estimated to be an approximately exponential attenuation of the electron intensity as a function of the depth from the starting point to the vacuum surface, at which the initial electron signal is reduced to 1/e [26]. The electron probing depth in aqueous solution is not accurately known, however, the depth at PKE of less than 5 eV is speculated to be 10–100 nm [27]. Additionally, the streaming potential can be largely reduced to an energy level of ~100–300 meV by adding electrolyte salts of NaX (X=F, Cl, Br, and I) to the solution [18, 28−30]. Taken these aspects together, it seems reasonable that the $\Delta E_{{\rm{H}}_2{\rm{O}}} $ is on the order of around 200 meV at different flow rates and laser energies.

      As discussed above, when each eBE spectrum is further corrected by the accurate eBE value of 11.33 eV for the 1b1 orbital of liquid water, the calibration of each eBE spectrum is expected to be finished. However, the VIEs of the D0(π−1) and D1(n−1) states also exhibit some flow rate and laser energy dependences. As summarized in Table III , the VIEs for both D0(π−1) and D1(n−1) states decrease as the flow rate increases. Likewise, the VIEs for both D0(π−1) and D1(n−1) states decrease as the laser energy increases. The decrease of the VIEs means an energy shift towards a higher PKE. The ionizing signal increases when the flow rate or laser energy increases, indicating more photoelectrons are generated when the flow rate or laser energy is bigger. The energy shift towards higher PKE is consistent with the repulsion among the electrons. Thus, the flow rate and laser energy dependences might result from the space charge effect [31, 32]. The energy shift is in the range of several tens of meV to 200 meV at different flow rates and laser energies, resulting in an average energy shift of less than 3%. Although we cannot assess the individual contribution from the elastic and inelastic electron scatterings, streaming potential, and space charge effect, we can make a combined assessment of the elastic and inelastic electron scatterings, streaming potential, and space charge effect. After all the corrections, the combined experimental error still exists and is less than 3%, i.e., in the range of several tens of meV to 200 meV.

    IV.   CONCLUSION
    • In aqueous solution, the genuine photoelectron kinetic energy spectra are easily affected by elastic and inelastic electron scattering, streaming potential and space charge effect, which can be assessed in a combined experimental error manner by adding electrolyte salt to the solution, together with adjusting the solution flow rates through the microjet and the laser energies. We used femtosecond liquid-microjet photoelectron spectroscopy to measure the ultraviolet PKE (or eBE) spectra of aqueous thymidine at different solution flow rates and laser energies. As an energy standard, the accurate vertical eBE value of 11.33 eV for the 1b1 orbital of liquid water was used to calibrate the observed eBE spectrum, and the corresponding combined experimental error for the eBE of the 1b1 orbital of liquid water was estimated to be about 200 meV. The vertical eBE values of the cationic ground state D0(π−1) and the cationic first excited state D1(n−1) of aqueous thymidine were found to exhibit an energy shift in the range of several tens of meV to 200 meV at different flow rates and laser energies, resulting in an average energy shift of less than 3%.

    Figure (4)  Table (3) Reference (43)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return