Substrate-Coated Illumination Droplet Spray Ionization: Real-Time Monitoring of Photocatalytic Reactions
- 261 Downloads
Real-time monitoring of photocatalytic reactions facilitates the elucidation of the mechanisms of the reactions. However, suitable tools for real-time monitoring are lacking. Herein, a novel method based on droplet spray ionization named substrate-coated illumination droplet spray ionization (SCI-DSI) for direct analysis of photocatalytic reaction solution is reported. SCI-DSI addresses many of the analytical limitations of electrospray ionization (ESI) for analysis of photocatalytic-reaction intermediates, and has potential for both in situ analysis and real-time monitoring of photocatalytic reactions. In SCI-DSI-mass spectrometry (MS), a photocatalytic reaction occurs by loading sample solutions onto the substrate-coated cover slip and by applying UV light above the modified slip; one corner of this slip adjacent to the inlet of a mass spectrometer is the high-electric-field location for launching a charged-droplet spray. After both testing and optimizing the performance of SCI-DSI, the value of this method for in situ analysis and real-time monitoring of photocatalytic reactions was demonstrated by the removal of cyclophosphamide (CP) in TiO2/UV. Reaction times ranged from seconds to minutes, and the proposed reaction intermediates were captured and identified by tandem mass spectrometry. Moreover, the free hydroxyl radical (·OH) was identified as the main radicals for CP removal. These results show that SCI-DSI is suitable for in situ analysis and real-time monitoring of CP removal under TiO2-based photocatalytic reactions. SCI-DSI is also a potential tool for in situ analysis and real-time assessment of the roles of radicals during CP removal under TiO2-based photocatalytic reactions.
KeywordsMass spectrometry Reactive intermediate Droplet spray ionization Real-time monitoring Photocatalytic reaction
Recently, pollutants, including organometallics, pharmaceuticals, and pesticides are manufacturing targets that are continuously being developed and manufactured [1, 2]. They are excreted as metabolites in urine and feces, and are commonly detected in water sources, sediments, and soils at ng·L−1 to μg·L−1 levels . Some pollutants are highly stable, are poorly biodegradable, and can cause potentially adverse health effects. TiO2-based photocatalytic techniques have been considered as an effective approach for the removal of these organic contaminants , and as an effective recovery method of clean energy from wastewaters . Notably, studies on elucidating oxidation pathways of single organic pollutant in TiO2-based photocatalytic processes are growing constantly. However, few techniques for monitoring photocatalytic reaction in real time exist.
Traditionally, detection of photocatalytic-reaction intermediates and products is studied by electron-spin resonance (ESR) spectroscopy , ultraviolet/visible (UV/Vis) spectroscopy , capillary electrophoresis (CE) , nuclear magnetic resonance (NMR) , chromatography , and mass spectrometry (MS) . However, these techniques have two drawbacks. First, these techniques are time-consuming and require complex sample pretreatment. With inappropriate reagents used, samples may suffer contamination and loss from sample pretreatment, such as liquid–liquid extraction, filtration, and dilution. Second, some reaction intermediates are short-lived, are low in concentration within an aqueous solution, and are difficult to isolate and detect. These two drawbacks limit accurate elucidation of a reaction mechanism.
Ambient ionization (AI) has emerged as a popular technique in mass spectrometry since the introduction of desorption electrospray ionization (DESI) . AI is defined as the ionization of samples in their natural state with little to no pretreatment. AI can also be considered as an interface for online reaction monitoring. In this sense, AI can convey the sample existing in solution with different concentrations into the gas phase for observation with disparate relative abundance. In the past two decades, quite a number of AI sources were developed and were applied to the analysis of complex samples in both condensed phase and gas phase [13, 14, 15, 16, 17, 18]. Up to now, a number of them for real-time monitoring of chemical reactions have been reported, and have been well documented [19, 20, 21]. For example, paper spray ionization (PSI) has been used for the monitoring of both clandestine chemical synthesis  and the redox process of methylene blue . With these techniques, chemists can directly track reaction progress, study short-lived reactive intermediates, and discover new reaction mechanisms.
Recently, droplet spray ionization (DSI) , by replacing the paper substrate with a microscope cover slip to launch a charged-droplet spray, was developed in our laboratory. DSI features contaminate-free sampling of the target samples, low cost, simplicity, and durability of the spray platform. Moreover, DSI eliminates the analysis delay introduced by the tubing leading to an ESI source. Although the delay of online reaction monitoring with the tubing between the reaction vessel and electrospray ion source is only a few seconds , the lifetimes of some short-lived intermediates are less than one second . In DSI, the corner of the slip cover was used as a microreactor. By sequentially pipetting reagents to an already spraying system, detection of reaction intermediates on the order of less than a second is possible. For example, DSI was successfully used for in situ analysis and real-time monitoring of a Ziegler-Natta catalyst reaction , where the catalytically active intermediates and polymeric products were both captured and identified. Also, DSI has been applied to monitoring macromolecule reactions in real-time . As a result, DSI is particularly suitable for capturing both short-lived and low concentration reactive intermediates involved in chemical reactions.
Herein, a novel method based on DSI named substrate-coated illumination droplet spray ionization (SCI-DSI) was developed, for in situ analysis and real-time monitoring of photocatalytic reactions. The raw glass cover slip used in DSI was replaced with a TiO2-coated slip cover and UV illumination was positioned above the slip. This combination enables a TiO2-based photocatalytic reaction. First, the performance of SCI-DSI was investigated. Second, in situ analysis and real-time monitoring of the removal of cyclophosphamide (CP) by TiO2/UV was attempted. Third, the roles of reactive species, such as free hydroxyl radical (·OH), adsorbed ·OH, and valence hole (h+), for CP removal were monitored in real time. The data present in this work demonstrated that SCI-DSI is a useful tool to directly track TiO2-based photocatalytic reaction progresses, to detect short-lived intermediates and to study the roles of reactive radicals in real time.
Acetonitrile (HPLC grade) was obtained from Sigma-Aldrich (Darmstadt, Germany). Both 4-aminophenol and cyclophosphamide (CP) (purity ≥ 99.5%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Sodium fluoride and 2-propanol were purchased from Aike Co. (Chengdu, China). Ultrapure water was obtained from a water Milli-Q purification system (Milford, MA, USA). All the individual standards were freshly prepared before use.
Substrate-Coated Illumination Droplet Spray Mass Spectrometry (SCI-DSI-MS)
There was an uncovered TiO2 area at the edge of the TiO2-coated slip with a width of 1.5 mm. This coating-free area was easily maintained and impeded the sample solution loss by wicking to the custom holder. If the slip surface is completely coated with TiO2, which transforms the hydrophobic glass surface to a hydrophilic surface, the sample solution loaded onto the TiO2-coated surface cannot form a droplet; thus the sample solution will disperse over the entire surface and flow by wicking action direct to the holder.
In Situ Analysis and Real-Time Monitoring of the Removal of CP by TiO2/UV
In Situ Analysis and Real-Time Assessment of the Roles of Radicals During CP Removal
Both 2-propanol and the fluorine anion (F-) were used as radical scavengers to determine the roles of reactive species (free ·OH, adsorbed ·OH and h+) for CP removal. This work was investigated by monitoring the signal intensity of protonated CP (m/z 261). A high voltage of 4.0 kV was applied to the corner and subsequently loaded with 10 μL of acetonitrile/water (v/v: 7/3) (0 s), 70 μL of acetonitrile/water (v/v: 7/3) containing CP (5 × 10−5 mol/L) (10 s), and 10 μL of acetonitrile/water (v/v: 7/3) containing 2-propanol or F- (5 × 10−4 mol/L) (120 s). The UV lamps were turned on at 14 s. Data were recorded continuously during these additions of reagents.
Mass Spectrometry Experiments
All the experiments were implemented on a LTQ mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Full-scan positive ion spectra were acquired and processed by the default Xcalibur package (ver. 2.2.7; Thermo Fisher Scientific). The ion maximum injection time of the linear ion trap was set at 10 ms. The capillary temperature was maintained at 275 °C. The tube lens voltage and capillary voltage were set at 110 and 35 V, respectively. MS/MS experiments were used to identify reaction intermediates and products using the default parameters, except the collision voltage, which was adjusted in the range of 15–35 V for best signal-to-noise ratio.
Results and Discussion
Characterization of TiO2-Coated Cover Slip
In contrast to the conventional cover slip surface, the TiO2-coated surface is hydrophilic. Therefore, in SCI-DSI, both the droplet form of the sample solution loaded onto the modified surface and the spray performance may change compared with the previously reported DSI performance.
Droplet diameters of the sample solution loaded onto the TiO2-coated and raw slip surface were investigated to confirm the transformation of the hydrophobic surface to a hydrophilic surface. As shown in Figure 3c, the 10-μL sample solution loaded onto the raw cover slip formed a diameter of approximately 4.1 mm. However, the sample solution loaded onto the TiO2-coated slip generated a larger droplet diameter, which was 12.5 mm. Thus, the TiO2-coated surface significantly enhanced the hydrophilicity of the slip surface.
Spray Performance of SCI-DSI
To investigate the effect of TiO2-coated surface on the spray performance of SCI-DSI, both the spray time and signal intensity with different sample volumes were characterized. The signals were obtained by monitoring protonated 4-aminophenol (5 × 10−5 mol/L). Each experiment was conducted in triplicate. The TiO2 uncoated area was at the edge of the slip and had a width of 1.5 mm. This uncoated area maintained the spray tip of a raw cover slip; thus the spray stability and tolerance to varied source position of TiO2-coated slip were similar to that of DSI.
Figure 3e is a bar plot of both spray times and signal intensities versus sample volumes. As shown in Figure 3e, the spray time increased with increasing sample volumes. At approximately 70 μL of sample volume, the spray time reached a maximum, above which the spray time is approximately constant. The overall trend of SCI-DSI compared well with that of DSI , but compared to DSI, the SCI-DSI spray time decreased. This trend is due to two factors. First, the hydrophilic and porous TiO2-coated surface increases the surface area of the droplet. Second, the blackbody radiation from the UV lamps increases the surrounding temperature of the droplet. Both factors accelerated the evaporation of the sample solution.
In addition, a slight decrease of signal intensities was observed with increasing the sample volume (Figure 3e). The signal intensity decreased with larger sample volume; this may be related to inefficient spray characteristics. Larger sample volumes loaded onto the cover slip corner accelerate the sample solution being sucked into the MS inlet and hinder the formation of fine spray droplets. Notably, the fluctuant range of the signal intensities was small, and y-error bars regarding the signal intensities and spray time shown in Figure 3e are relatively small. This confirms that the spray of SCI-DSI is relatively steady.
In Situ Analysis and Real-Time Monitoring of the Removal of CP by TiO2/UV
As shown in Figure 4b, the signals for protonated CP at m/z 261 were observed after loading CP into the spray at 10 s. Figure 4b also shows the signals of m/z 283 and 299, corresponding to [M + Na]+ and [M + K]+, respectively. The chlorine isotopic envelope sufficiently confirmed these assigned peaks.
Figure 4b and c illustrate a decrease in intensities over time of the CP signals (m/z 261, 283, and 299) and concurrent increase in intensities of 1 (m/z 277) and 2 (m/z 259), respectively. The 1 was assigned as a hydroxylation product, 4-hydroxycyclophosphamide . The 2 was identified as iminophosphamide, a nontoxic human metabolite of CP . However, 2 was not detected in Pt-TiO2/UV system of Ofiarska et al. . The reason for this phenomenon may be that the Pt-TiO2 surface enhanced the reactivity of photogenerated electrons to the adsorbed oxygen species, and resulted in mainly hydroxylation of CP.
Figure 4c also gave signals of m/z 275, 142, and 138, although they were recorded at relatively low peak ratios. As the reaction proceeded, an increase in intensities over time of these ions was observed (Figure 4c and d). The signals at m/z 275 (3) was reported by Lutterbeck et al. , and was identified as 4-ketocyclophosphamide, an inactive metabolite of CP . Ions of m/z 138 (4) were assigned as (1-aminocyclopropyl)phosphonic acid by Lutterbeck et al. . The 4 may be formed by cleavage of the N–P bond in -P-N(CH2CH2Cl)2 group from 3 with loss of bis(2-chloroethyl)amine (m/z 142).
Figure 4a also shows that different reaction dynamics and mass transfer are needed to explain the ion abundance of reagents, intermediates, and products. As seen from Figure 4a, the signal intensity of CP (m/z 261) rapidly decreased and the intensities of 1 (m/z 277) and 2 (m/z 259) increased. Later, the intensities of 3 (m/z 275) and 4 (m/z 138) increased. This proved that SCI-DSI can be used for in situ analysis and real-time monitoring of CP removal under TiO2-based photocatalytic reactions, and determining the reaction mechanism in future studies. The extracted ion chromatogram (EIC) of reagent and intermediate ions can be used to reflect the ion abundances in solution, although it is not an accurate method for determining ion concentrations in solution.
Tandem Mass Spectrometry
Fragmentation of 1 led to the fragment ions at m/z 259, 221, and 142 (Figure 5a). The fragment ions at m/z 259 were formed by the loss of H2O from the parent ions. Ions of m/z 221 were formed by the loss of C3H4O from the parent ions. Cleavage of N–P bond of the parent ions with loss of C3H6NO3P generated fragment ions at m/z 142, consistent with [C4H9Cl2N + H]+. As shown in Figure 5b, the MS/MS result of 2 (m/z 259) produced fragment ions at m/z 243, 223, 175, 147, and 101, by loss of O, HCl, CH2Cl2, C3H6Cl2, and C4H10Cl2NO, respectively. As seen from Figure 5c, fragmentation of 3 led to the fragment ions at m/z 239, 221, 159, 142, and 124. Loss of HCl from the parent ions gave the fragment ions at m/z 239 with low ion abundances. Ions of m/z 221 were formed by the loss of C3H2O from the parent ions. Loss of C2H3Cl and H2NO2P from ions of m/z 221 gave the fragment ions at m/z 159 and 142, respectively. Ions of m/z 124 were assigned as [C2H3ClNOP + H]+, which was formed by loss of C5H10ClNO2 from the parent ions.
The MS/MS results of 1, 2, and 3 compared well with previous results . This agreement indicates that SCI-DSI is a suitable tool for detection and identification of reactive intermediates in real time.
Real-time monitoring of the roles of reactive species for CP removal
The 2-propanol molecule and the F– anion are scavengers and were used to investigate the roles of the reactive species (free ·OH, adsorbed ·OH, and h+) for the removal of CP in TiO2/UV. The 2-propanol molecule is known as an electron donor, which can react with free ·OH, adsorbed ·OH, and h+ . The F– anion is commonly used to inhibit the production of adsorbed ·OH and enable the generation of free ·OH because the F- anion shows strong adsorption on the TiO2 surface and can completely occupy the reaction sites on the TiO2 surface .
As depicted in Figure 6b, the signal intensity of CP in TiO2/UV continually decreased before adding scavengers. However, there was significant difference on signal intensity of CP after adding the scavengers. In detail, the signal intensity of CP decreased slowly in presence of 2-propanol, but the signal intensity of CP rapidly decreased in presence of F-. This result supports the free OH· conclusion of Ofiarska et al.  and compared well with previous investigations of Lin et al.  that the free ·OH in solution was responsible for the removal of CP in TiO2/UV. To the best of our knowledge, this result is the first monitoring of active radicals in real time and in situ.
A novel method named substrate-coated illumination droplet spray ionization (SCI-DSI) was developed. SCI-DSI has advantages of in situ analysis and real-time monitoring of photocatalytic reactions. By sampling in situ and monitoring in real time, SCI-DSI addresses many of analytical limitations of ESI such as dilution prior to analysis, expensive pulled glass capillaries, and reaction-monitoring delay. SCI-DSI also provides great advantages in simplifying the analysis process for photocatalytic reactive intermediates and products, and can detect photocatalytic reactive intermediates in less than 1 s.
The spray performance of SCI-DSI was investigated, and the utility of SCI-DSI for photocatalytic reaction monitored was successfully demonstrated by in situ analysis and real-time monitoring of the removal of CP in TiO2/UV. Proposed reactive intermediates were captured and identified by tandem mass spectrometry. The free ·OH was confirmed as the main radicals for CP removal. These results show that SCI-DSI is not only a powerful tool to directly track CP removal under TiO2-based photocatalytic reactions, to detect and study reaction intermediate, and to elucidate the reaction mechanism, but also a potential tool to assess the roles of radicals during reactions.
In SCI-DSI, the solvent is very important. It must be a suitable spray solvent to form a steady spray, and also have slight/no influence on the reaction. The solvent used in the current work has no significant influence on the reaction as the proposed reactive intermediates were observed. However, for a special reaction, the solvent should be carefully selected. In the future, we will replace the spray platform with an airtight micro-reactor to overcome the drawback of solvent evaporation and address batch monitoring. Moreover, SCI-DSI can be used for in situ analysis and real-time monitoring of photocatalytic reactions with other surface-modified slip covers (e.g., Pt-TiO2, Si-TiO2, etc.), and is a new tool for testing self-cleaning glass.
This work was supported by the 13th Five-Year Key Research and Development Program of China (no. 2016YFF0100302), Natural Science Foundation of Shandong (ZR2016ZM11, ZR2016BP01), and Key Research and Development Program of HITWH (no. 2013DXGJ01, HIT.NSRIF.201710).
- 7.Bianco Prevot, A., Fabbri, D., Pramauro, E., Baiocchi, C., Medana, C.: High-performance liquid chromatography coupled to ultraviolet diode array detection and electrospray ionization mass spectrometry for the analysis of intermediates produced in the initial steps of the photocatalytic degradation of sulfonated azo dyes. J. Chromatog. A 1202, 145–154 (2008)CrossRefGoogle Scholar
- 11.Kouloumbos, V.N., Tsipi, D.F., Hiskia, A.E., Nikolic, D., van Breemen, R.B.: Identification of photocatalytic degradation products of diazinon in TiO2 aqueous suspensions using GC/MS/MS and LOMS with quadrupole time-of-flight mass spectrometry. J. Am. Soc. Mass Spectrom. 14, 803–817 (2003)CrossRefGoogle Scholar
- 29.Lutterbeck, C.A., Wilde, M.L., Baginska, E., Leder, C., Machado, E.L., Kummerer, K.: Degradation of cyclophosphamide and 5-fluorouracil by UV and simulated sunlight treatments: assessment of the enhancement of the biodegradability and toxicity. Environ. Pollut. 208, 467–476 (2016)CrossRefGoogle Scholar