Laser-Induced Acoustic Desorption Atmospheric Pressure Photoionization via VUV-Generating Microplasmas
- 531 Downloads
We demonstrate the first application of laser-induced acoustic desorption (LIAD) and atmospheric pressure photoionization (APPI) as a mass spectrometric method for detecting low-polarity organics. This was accomplished using a Lyman-α (10.2 eV) photon generating microhollow cathode discharge (MHCD) microplasma photon source in conjunction with the addition of a gas-phase molecular dopant. This combination provided a soft desorption and a relatively soft ionization technique. Selected compounds analyzed include α-tocopherol, perylene, cholesterol, phenanthrene, phylloquinone, and squalene. Detectable surface concentrations as low as a few pmol per spot sampled were achievable using test molecules. The combination of LIAD and APPI provided a soft desorption and ionization technique that can allow detection of labile, low-polarity, structurally complex molecules over a wide mass range with minimal fragmentation.
KeywordsLaser-induced acoustic desorption Atmospheric pressure photoionization Microplasma Microhollow cathode discharge Low-polarity analytes
Over time, the field of mass spectrometry (MS) has expanded in scope because of the introduction of a variety of ion generation methods that extend the range of systems that can be probed. Of particular interest is the development of various MS techniques that are suitable to analyze minute amounts of analytes deposited on surfaces, each having different drawbacks and advantages . Nonpolar molecules have, thus far, been difficult to analyze by surface MS techniques due to both the inherent ionization mechanisms and amount of energy imparted to the molecules during the desorption/ionization events. Owing to their low proton affinities, these compounds do not protonate easily, and are not detected with high sensitivity using techniques such as desorption electrospray ionization (DESI) or matrix-assisted desorption ionization (MALDI). As a consequence, alternative surface analysis techniques that focus on photoionization of neutral desorbates have been developed, including desorption atmospheric pressure photoionization (DAPPI) and laser ablation/desorption photoionization [2, 3, 4]. With DAPPI, a heated solvent jet desorbs analytes from a surface, which undergo atmospheric pressure photoionization (APPI) initiated by a radiofrequency-driven vacuum ultraviolet radiation (rF-VVU) emitting lamp and then swept into the atmospheric pressure inlet of a mass spectrometer. Despite its appealing coverage in terms of analyte polarity, studies have suggested DAPPI may not be best suited for labile high molecular weight compounds because of the relatively large amount of heat required to desorb larger molecules, and the subsequent fragmentation . As for methods involving direct (without matrix) laser ablation/desorption, these are also known to produce fragmented gas-phase molecules because of direct irradiation from the desorption laser as well as above plume/plasma interactions. Direct laser ablation/desorption is also often limited to lower molecular weight desorbates .
Laser-induced acoustic desorption (LIAD) is capable of desorbing analytes in a wide range of molecular weights and has therefore received significant recent attention [7, 8, 9, 10]. LIAD is accomplished by irradiating the backside of a metal foil ~ 5–20 μm thick with a nanosecond laser pulse of power density ~108 W/cm2. After absorption of this pulse, molecular species are ejected from the opposite side of the foil with relatively little energy transferred to the desorbed species [11, 12]. Since LIAD is known to generate mostly neutral and intact molecules, this technique is well suited for softly producing pulses of low-polarity thermally-labile gas-phase molecules [11, 13, 14]. Initial studies describing the LIAD process suggested an acoustic wave mechanism for volatilization . However, subsequent studies suggest a more complex mechanism where laser-induced stresses in the foil cause cracks in deposited molecular crystals and repulsive non-equilibrium electronic surface states, thus creating desorption sites from which analytes are ejected [12, 16]. Though now considered to be misleading, we continue to use the term LIAD here due to familiarity. Applications of LIAD to ambient mass spectrometry surface analysis scenarios demonstrated its viability for desorption of various organic compounds in a mass range extending up to m/z 762 using atmospheric pressure chemical ionization (APCI) as the tool for ion generation .
Recent developments in microplasma ion sources for ambient/atmospheric MS have demonstrated their viability as vacuum ultraviolet (VUV) sources for APPI techniques . This specific class of plasma device, the micro-hollow cathode discharge (MHCD), is characterized by a small hollow in the cathode of the device in which Pendel electrons experience multiple collisions with the plasma gas . For a neon:hydrogen admixture discharge gas, these interactions are known to produce neon excimers that through near-resonant energy transfer dissociate H2 leading to Lyman-α (10.2 eV) photon emission . The emission spectra from these devices are also known to be dependent upon the discharge gas composition, flow rates, and discharge current/voltage [18, 21, 22].
In this work, the generation of VUV light and the potential utility of MHCDs as miniature, low power consumption, atmospheric pressure photoionization sources for applications in surface desorption measurements is demonstrated. Specifically, we report the first combination of LIAD and APPI for the analysis of low-polarity compounds deposited on titanium foil. This combination utilizes a MHCD with a LiF exit window that seals the plasma cavity, blocks the emission and release of reactive plasma species, and allows the transmission of VUV to mid-infrared photons. This LIAD-APPI approach provides the ability to detect picomoles of low-polarity substances on a foil surface, and is expected to be general and particularly applicable for larger molecules that are prone to unimolecular decay and fragmentation if desorbed by other means.
Materials, Chemicals, and Sample Preparation
Titanium foil 12.7 μm thick (99.8% metals basis) for the LIAD experiments was purchased from Alfa Aesar (Ward Hill, MA, USA). The low-polarity compounds studied (α-tocopherol, perylene, cholesterol, phenanthrene, phylloquinone, and squalene), solvents (toluene, hexane, ethanol, dichloromethane), and molybdenum foil (100 μm thick) for the plasma device electrodes were purchased from Sigma Aldrich (St. Louis, MO, USA). The LiF window used in the MHCD device (10 mm diameter, 1 mm thick) was purchased from Crystran (Poole, Dorset, UK). Samples were deposited on the 12.7 μm Ti foil using the iMatrixSpray open source deposition system (Grubenacker, Switzerland). Solutions ranging from 10 mM to 1 μM were deposited by the iMatrixSpray system through 10 coating cycles, which leads to a pixel to pixel coefficient of variance of 7% or less .
Lyman-α light (10.2 eV) generation was accomplished by operating the MHCD with the admixtures of 3%–7% hydrogen in neon as previously reported in similar devices . For all mass spectrometry experiments, unless otherwise noted, a nitrogen purge in the device cavity (2 SLM) was allowed to run for 10 min and shut off before plasma ignition. The plasma gas mixture of 4% H2 in Ne then flowed through the device at 40 sccm to enhance excimer formation and thus VUV emission . Current delivered to the device was limited between 10 and 11 mA by the Stanford Research Systems model PS350 power supply (Sunnyvale, CA, USA). The corresponding applied voltages ranged between 1100 and 1200 V. Flow rates were controlled by two FMA-5500 series mass flow controllers (MFCs) from Omega Engineering (Stamford, CT, USA). For VUV energy-resolved spectroscopy experiments, the MHCD cavity was purged with argon and a 7% H2 in Ne plasma gas flowed into the device at 70 sccm, while the current was varied between 5 and 10 mA. Though the VUV emission and LIAD-APPI experiments were carried out under slightly different relative gas mixing ratios, the emission is not expected to vary drastically, especially under constant current/discharge conditions.
VUV Monochromator Instrumentation
Measurements of the Lyman-α output of the device were made using a 0.2 m VUV monochromator (Acton VM-502) equipped with a sodium salicylate scintillator and photomultiplier tube detector at the Cryogenic Chemistry Laboratory at NASA’s Jet Propulsion Laboratory (Pasadena, CA, USA). Calibration of the instrument was accomplished using a Hg lamp, and tuning of the Lyman-α line resolution was performed with a rf argon/hydrogen lamp. Using the emission lines of Hg at 253.7 and 184.5 nm, a resolution of 5 Å could be estimated. For measurements on the plasma device, the spectrometer was operated at low vacuum (0.1 Torr) and the device was sealed to the monochromator inlet by Apiezon putty. This was done to reduce further attenuation of the Lyman-α emission intensity by windows external to the device.
The LIAD substrate foils were secured tightly over a metal tube to ensure a flat foil surface and placed at a fixed position 5 mm from the inlet of a Bruker micrOTOF-Q II mass spectrometer (Bruker, Billerica, MA, USA). The instrument was operated in positive mode in the range of 50 to 1000 m/z with the following settings: +50 V capillary bias, –50 V end plate offset, 3.0–4.0 L min–1 N2 dry gas at 34 °C, 3.3 mbar foreline pressure, and a 5.81 × 10–7 mbar TOF pressure. A vial of high pressure liquid chromatography grade toluene was placed beneath the LIAD foil surface where the vapor present due to normal evaporative loss would act as an APPI dopant. The liquid level was checked and replenished periodically to ensure a consistent atmospheric concentration between experiments. A 1064 nm Nd:YAG laser (Continuum, San Jose, CA, USA) with a spot size of 0.11 cm2 was aligned directly with the mass spectrometer inlet capillary at a power density of 0.6 × 108 W/cm2, and operated at a pulse frequency of 10 Hz. This laser strikes the backside of the foil, ejecting molecules via the LIAD process into the region containing toluene photoions generated by Lyman-α photons from the MHCD. These molecules are then ionized via APPI mechanisms near and flowing towards the inlet of the mass spectrometer and are then collected. This process is schematically shown in Figure 2. Before all experiments, the MHCD was on with toluene present and signal was monitored. Base peak intensity from toluene oxidation products was observed to be between 1 and 1.5 × 104 counts and only significantly deviated from that amount if the photon producing discharge was observed to be unstable. Typical compound analyses were accomplished by taking 1 min of spectra in the 50–800 mass range, starting the laser pulse sequence after 6 s of collection with only the MHCD running. This length of data acquisition was done to observe signal rise time and signal consistency over time. Cholesterol and α-tocopherol were studied in decreasing surface concentrations by depositing more dilute solutions of each on foil via the iMatrixSpray system.
Results and Discussion
Microplasma Lyman-α Output
LIAD-APPI of Low-Polarity Organics
To extend the applications of microplasmas in MS and to demonstrate a novel means for desorption and mass analysis of low-polarity species, α-tocopherol, perylene, cholesterol, phenanthrene, phylloquinone, and squalene were chosen for analysis by LIAD-APPI. For demonstration purposes, these analytes were deposited on the titanium foil in surface coverages of 10 nmol/cm2. Control analyses were performed by recording spectra without the MHCD operating and laser firing. In every case, no analyte signal was observed. This may be due to absorption by atmospheric water dropping the 10.2 eV photon flux by two orders of magnitude over a few centimeters in the region of the LIAD plume combined with the neutral number density of desorbates being too low to yield any appreciable amount of photoionized products. Thus, the production and transport of toluene photoions within and towards the LIAD plume region has significant impact on ionization efficiency.
Mechanisms of LIAD-APPI and Detection Limits
Mechanisms of APPI contributing to the effective ionization of low-polarity molecules desorbed via LIAD may include direct photoionization of the analyte, charge exchange with photoionized dopant molecules introduced in the sample region when the IP of the dopant is greater than that of the analyte, and protonation of higher proton affinity analytes via reactive dopant photoions . No signal was observed in the absence of the toluene dopant, which is consistent with the observation that analyte ion-formation is heavily dependent upon the ion-molecule reactions with dopant photoions [28, 29]. Owing to the relatively low IP of toluene (IP = 8.8 eV) compared with other high vapor pressure dopant molecules [e.g., chlorobenzene and acetone (IP ≥ 9 eV)], we expect this to be a relatively soft means of charge transfer ionization [30, 31, 32]. The efficient formation of the molecular radical cation for the samples shown can be attributed to their effective ionization potentials. For α-tocopherol, phenanthrene, perylene, and squalene with IPs equal to 6.7 eV, 7.9 eV, 6.96 eV, and 6.9 eV, respectively, it can be concluded that the molecular radical cation peak is dominant due to all IPs being lower than that of toluene by at least 0.9 eV [33, 34, 35, 36]. This allows for a very favorable charge exchange. With cholesterol (IP = 8.5 eV) and phylloquinone (estimated IP = ~9.3 eV compared with IPs of vitamin K2 derivatives), the IPs lie closer to or exceed that of toluene. This leads to reduced analyte ion formation from the charge exchange reaction [37, 38]. Early studies on cholesterol fragmentation support this observation using electron ionization at 70 and 16 eV, where the molecular radical cation was found to be the dominant peak in the mass spectrum and also demonstrates its sensitivity to internal vibrational temperature causing increased fragmentation [27, 39].
Estimating the toluene number density in the air to be ~1016 mL–1 due to vapor diffusion and using a cross section of 3.0 × 10–17 cm2 for 10.2 eV toluene photoionization , the number density of toluene photoions can be estimated to be on the order of 105–106 mL–1. This assumes a Lyman-α flux of 2.0 × 109 photons s–1sr–1 from an estimated 5% of total radiative output seen in Lyman-α lamps [18, 41]. This number is much smaller than the estimated number density of the desorbed products Ndes > 109 mL–1. This estimate assumes an average uniform upper bound for the toluene photoion density in the region of the LIAD plume, whereas the actual photoion density may be localized about the window of the MHCD, thus reducing the number density in the region of the plume even further. Thus, the sensitivity for this source geometry is enhanced by the forward scattering of the desorbates through the LIAD process and the pressure-driven flow at the inlet of the mass spectrometer but limited by the amount of toluene ions present lowering the ionization efficiency. Optimizing the setup to maximize collection consisted of observing base peak intensity of α-tocopherol spectra versus foil distance from the inlet. The result led to an optimum distance of 5 mm relative to the mass spectrometer sampling port. The higher collection at the optimum distance is mainly attributed to closer proximities cutting off photon flux, thereby reducing ions generated and further distances being too far for most desorbates to reach the inlet.
Lowest Detected Amounts of Cholesterol and α-tocopherol. Foil Coverage Values were Calculated Using Parameters from the iMatrixSpray System and Relative Error Based on Inhomogeneity of Sample and Uncertainty in Laser Spot Size
Solution Concentration [μM]
Foil Coverage (pmol/cm)
Lowest Detected Amount (pmol/spot)
Relative Error (±pmol/spot)
m/z 368 [M – H2O]+, m/z 369 [M – OH]+
m/z 430 M+ ⋅
The first application of laser-induced acoustic desorption and atmospheric pressure photoionization as a mass spectrometric method for detecting adsorbed low-polarity organic molecules has been demonstrated. The unique approach involved the use of a Lyman-α generating microhollow cathode discharge microplasma photon source in conjunction with the addition of a gas-phase molecular dopant. This combination provides both a soft desorption method and a relatively soft ionization technique. The total energy inherent in the desorption step is low and the total energy deposited in the product ion is low enough to produce spectra with minimal fragmentation for most molecules. Detectable surface concentrations as low as a few pmol per spot were achievable using test molecules such as α-tocopherol and cholesterol. Efforts are underway to improve the ion transport and toluene photoion density parameters, which should extend the mass range to higher molecular weights and allow detection of lower concentrations than shown here. These include the use dopant sprays to aid in transport and increase ionization efficiency of molecules while still using LIAD as the main desorption mechanism. The sample preparation and analysis technique together may prove to be useful in detecting reaction products generated from drying aerosol sprays on metal surfaces. If successful, LIAD-APPI may prove to be a useful approach to detecting larger and more labile molecules, molecular assemblies, and aggregates than shown here.
This work was jointly supported by NSF and the NASA Astrobiology Program, under the NSF Center for Chemical Evolution, CHE-1504217. Work at JPL on VUV detection was supported by a joint Jet Propulsion Laboratory-Georgia Institute of Technology Strategic University Research Partnership (JPL-GIT SURP) grant.
- 2.Haapala, M., Pól, J., Saarela, V., Arvola, V., Kotiaho, T., Ketola R.A., Franssila, S., Kauppila, T.J., Kostiainen, R.: Desorption atmospheric pressure photoionization. Anal. Chem. 79, 7867–7872 (2007)Google Scholar
- 8.Sezer, U., Wörner, L., Horak, J., Felix, L., Tüxen, J., Götz, C., Vaziri, A., Mayor, M., Arndt, M.: Laser-induced acoustic desorption of natural and functionalized biochromophores. Anal. Chem. 87, 5614–5619 (2015)Google Scholar
- 9.Jia, L., Weng, J., Zhou, Z., Qi, F., Guo, W., Zhao, L., Chen, J.: Note: Laser-induced acoustic desorption∕synchrotron vacuum ultraviolet photoionization mass spectrometry for analysis of fragile compounds and heavy oils. Rev. Sci. Instrum. 83, 026105 (2012)Google Scholar
- 10.Cheng, S., Cheng, T.-L., Chang, H.-C., Shiea, J.: Using laser-induced acoustic desorption/electrospray ionization mass spectrometry to characterize small organic and large biological compounds in the solid state and in solution under ambient conditions. Anal. Chem. 81, 868–874 (2009)CrossRefGoogle Scholar
- 17.Nyadong, L., McKenna, A.M., Hendrickson, C.L., Rodgers, R.P., Marshall, A.G.: Atmospheric pressure laser-induced acoustic desorption chemical ionization Fourier transform ion cyclotron resonance mass spectrometry for the analysis of complex mixtures. Anal. Chem. 83, 1616–1623 (2011)CrossRefGoogle Scholar
- 27.Wyllie, S.G., Amos, B.A., Tökés, L.: Electron impact induced fragmentation of cholesterol and related C-5 unsaturated steroids. J. Org. Chem. 42, 725–732 (1977)Google Scholar
- 30.Lu, K.-T., Eiden, G.C., Weisshaar, J.C.: Toluene cation: nearly free rotation of the methyl group. J. Phys. Chem. 96, 9742–9748 (1992)Google Scholar
- 31.Baer, T., Tsai, B.P., Smith, D., Murray, P.T.: Absolute unimolecular decay rates of energy selected metastable halobenzene ions. J. Chem. Phys. 64, 2460–2465 (1976)Google Scholar
- 32.Traeger, J.C., McLouglin, R.G., Nicholson, A.J.C.: Heat of formation for acetyl cation in the gas phase. J. Am. Chem. Soc. 104, 5318–5322 (1982)Google Scholar
- 34.Thantu, N., Weber, P.M.: Dependence of two-photon ionization photoelectron spectra on laser coherence band width. Chem. Phys. Lett. 214, 276–280 (1993)Google Scholar
- 35.Shchuka, M.I., Motyka, A.L., Topp, M.R.: Two-photon threshold ionization spectroscopy of perylene and Van der Waals complexes. Chem. Phys. Lett. 164, 87–95 (1989)Google Scholar
- 38.Ishihara, M., Sakagami, H.: QSAR of molecular structure and cytotoxic activity of vitamin K2 derivatives with concept of absolute hardness. Anticancer Res. 27, 4059–4064 (2007)Google Scholar
- 39.Amirav, A.: Electron impact mass spectrometry of cholesterol in supersonic molecular beams. J. Phys. Chem. 94, 5200–5202 (1990)Google Scholar