In recent years, there is growing interest in ambient ionization mass spectrometry (AIMS) executed under the atmospheric pressure ambient. A typical method is an electro-spray ionization (ESI) method which is carried out by ionizing a charged droplet by application of a high voltage to a sample solution and an atmospheric pressure ionization (APCI) method which is carried out by evaporation. In addition, the DART [1] of helium and the LPI [2] of argon are work actively in the ionization method using excitation light of rare gas. DART was developed by Cody and Larameel et al in 2005, and LPI was developed by Tsuchiya et al in 1974. Both can ionize most organic compounds [3] (IE: 7 to 11 eV). The big difference lies in the energy difference of the excitation gas used.?To exciting energy (19.6 eV and 20.6 eV[4]) of the helium used in DART, the argon used in LPI is 11.83 eV [4] which smaller than those the values. DART can generate molecular ions and / or fragment ions [6] shown in Equations 1, 2 and 3 according to the Penning ionization mechanism [5] even for inert alkanes.

As in the case of electron capture ionization (EI), the appearance of fragment ions possible with DART is a powerful weapon for structural analysis of stable alkane molecules. Equations 1 to 3 occur with helium gas with large excitation energy, but in LPI, the excitation energy of argon is too small, so spectrum of these ion groups does not appear. On the case of the objective is to make only molecular ions appear, LPI is also advantageous in that no fragmentation of molecules occurs, which is caused by miscellaneous spectrum. DART performs with heated and excited helium gas to the sample for efficient desorption and/or ionization of the sample. Recently it seems unnecessary to do strong heating to damage the sample. The LPI method is carried out by a method in which corona discharge and a directly high DC voltage of positive electrode [7] (+1000 V) for assisting it are added to the sample. Furthermore, a rapid and programmed heating method is possible to assist ionization and desorption of the sample. LPI, which can further control heating, can also cut molecules of heat sensitive compounds such as peroxide [8]. Protonation of molecules is done with hydronium ions generated which ionization by DART and/or LPI of water in the atmosphere. The ions are generated from one molecule (or two molecules) of water for DART and two molecules of water for LPI, which were shown in Equations 4 ~ 8, respectively.

Production of hydronium ions necessary for the formation of protonated molecules has two processes occurring from water of monomer (equation 6) and dimer (equation 7). And the ionization energy is smaller for the dimer forming to cluster (11.21 eV [9]) than the monomer (12.6 eV [10]). Therefore, the formation of hydronium ions by excited argon (EE : 11.83 eV [4]) is from the water forming the dimer [9]. Protonated molecules are also generated by Penning ionization, but LPI make not have this process as molecular ions are not produced.

The amount of hydronium ion required to form protonated molecules of organic compounds is important for the formation of stable protonated molecules. Therefore, it is requiered to control the amount of water to generate hydronium ions. However, in the LPI method using an open-system corona discharge tube where the discharge part is in contact with the atmosphere, it is impossible to contain water in the argon gas flowing into the discharge tube due to the problem of energization. Therefore, as a method of exciting argon gas containing water to ionize it, we focused on argon gas for assisting startup of low pressure mercury lamp. The photoionization mass spectrometry developed by Rossing and Tanaka in 1956 [11] uses resonance lines (IE; 10.03 and 10.64 eV) of rare gas Kr. and, this was also referred to. In other words, it was thought that if the argon gas containing moisture was flowed outside the lamp and excited with the excited argon gas in the lamp, the problem of energization could be solved. Ionization by mercury excitation light (EE: 4.9 eV, 254 nm) [12] contained in a low-pressure mercury lamp seems to be irrelevant as it cannot ionize compounds with higher ionization energy. In addition, being concerned about the analytical operation by using low pressure mercury lamp is the generation of harmful ozone. However, that problem does not occur. This is because oxygen does not exist in the argon gas, and it is also decomposed by the contained water. As an example using an argon resonance lamp, Chen et. Al [13] is using as a detector of GC to analysis of an organic compounds. However, this method is not an indirect method. Ionization process is the same as LPI and is performed by both hydronium ions generated by excited argon and needle-shaped electrodes (sample stage) supplied with high voltage.

Experimental Method and Equipment

The entire ion source was shown in Figure 1 and its details has also described.

The ionization section consists of the following next four parts:  a discharge tube equipped with a low-pressure mercury lamp in a stainless steel tube equipped with a means for supplying water-containing argon gas,  argon gas generator containing moisture,  method of supplying to heating and DC high voltage to tungsten bar serving both as sample holder and electrode, and  it is constituted by a supply section of a adduct reagent which can be added as necessary. The lamp used was a Spectronics Corporation (New York, USA) model 11 SC-1 Short Wave UV lamp low pressure mercury lamp (Light dimension (mm): 6.5 ^φ x 19.1) which provides continuous lighting time of 5000 hrs. For the high voltage required for lighting, 200 V was supplied from Spectronics Corporation model SCT-1A power supply.

Figure 1: Overall view of the ionization method using a low-pressure mercury lamp (discharge tube) installed in a water-containing argon gas flow.

Argon gas containing a certain amount of water was supplied by combining the following parts. An empty vial D, a small vial E with water, valves B and C for controlling the amount of argon gas to be supplied, and a valve A for controlling the amount of argon gas contained in water.

Figure A exposes all of light-emitting portion, and Figure B shows a structure with its length restricted. Figure C and D show a structure in which gas is flowed throughout the light emitting portion and the portion in contact with the atmosphere is limited. Figure E and F show the structure combining Figure A to D. Figure 2 shows the various discharge tube shapes studied, and the discharge tube adopted has the structure shown in Figure 2-E. Figure 2 shows the various discharge tube shapes studied and studied, and the actual analysis was carried out with the structure shown in Figure 2-F.

Mass spectrometry was carried out by the method described below by the ionization unit shown in Figure 1 fabricated which was attached to a quadrupole mass spectrometer (Extra EXM-2000). The sample (10 all in controlled almost 10% concentration) was placed on a tungsten rod (tip is conical and flatly processed) also serving as a needle electrode and installed at a distance of about 5 mm from the pin hole plate of the mass spectrometer . Then, a voltage of 1 kV of DC positive charge necessary for sample removal and/or activate was supplied. Distance and supply voltage are related. The needle electrode can be heated with a heater wrapped around a quartz tube. Also, a temperature sensor (CA) is attached to the tip. Tungsten which is resistant to oxidation was used as the material of the needle electrode.

Reagents

The water was used collected from a high-purity refiner manufactured by Millipore immediately before use. In addition, 99.9% or more quality products were used for gases.

Alcohol, acetone and amine reagents (ethylamine, diethyl amine, triethylamine and pyridine) of special grade analytical grade were purchased from Wako Pure Chemical Industries, Ltd. In consideration of the production of protonated molecules and the action of water other than ionization, the target sample selected peroxides containing oxygen atoms and being sensitive to heat. t-butyl peroxylaurate is manufactured by Tokyo Oil & Fat Co., Ltd. and contains a large amount of water as a stabilizer.

Structure of Discharge Tubes, Water Content, and Gas Species

First, we tried the possibility of ionization of gas-phase water by exposing the entire light-emitting part of the lamp to the atmosphere (Figure 2A), but the spectrum of the ion group did not appear. Rather, ozone formation and decomposition (185 nm: 6.7 eV) by excitation light (254 nm: 4.9 eV) of mercury was harmful to the environment. Even when the length of the light emitting portion was limited to 5 mm (Fig.2B), no ion group appeared. Next, the light emitting part exposed in the argon gas flow was placed (Figure 2C to 2F). A small amount of ions (noise) was observed when only the tip of the light emitting part touched the atmosphere (Figure 2C). For that 5 mm (Figure 2D) these signal groups expanded further. So it seems that there is some relation between the light of the lamp and the flowing argon gas from the appearance of spectrum-like signals. In the discharge tube shown in Figure 2E where the outer tube, the inner tube, and the tip of the light emitting part are aligned, the ion intensity was very weak. By the light emitting part expanded to 5 mm (Figure 2F), 1 to 2 protonated molecules which corresponding to number of molecules water was observed (Figure 3 Ar - 0). The appearance of two protonated water molecules may have been caused by the tip of the light emitting part ionizing gas-phase water in the atmosphere.

The production of hydronium ions necessary for the formation of protonated molecules has two processes resulting from water of the monomer of formula 6 and the dimer of Fig. 7. However, at the excitation energy of argon (EE : 11.83 eV [4]), the monomer with higher ionization energy (IE : 12.6 eV [10]) cannot be ionized. But you can do it if it is a dimer (IE : 11.21 eV [9]). In addition, the LPI [7] that adds a high DC voltage to the needle electrode increases the possibility. The high voltage application of the needle electrode is important, and ions did not appear unless electricity was applied. We think that the protonated molecule of the dimer shown in Figure 3 Ar-0 was formed not by direct ionization of two molecules of water but by the hydronium ion ionizing the dimer water.

The fact that three or more molecules of cluster ions do not appear will be extremely small amount of gas phase water contained in the atmosphere. Indoor humidity and temperature were 20% and 22?, respectively. Tsuchiya [14] and others have observed water clusters starting from one water molecule as the background in the atmosphere, but the ion intensities of protonated water molecules 1 and 2 did not show specific ionic strength. Fenn and Searcy [15] discussed the appearance state of water clusters under vacuum with extremely low moisture content, but ions of protonated molecules of one and two molecules did not show special ionic strength. Vapor phase water concentration in the atmosphere seems to be related to the number of clusters produced. Since water forms clusters by intermolecular hydrogen bond even in liquid state [16], the appearance state of cluster ions due to moisture content in argon gas was also investigated in this experiment. In the condition of Ar-10 in Figure 3 in which water was added to argon gas, cluster ions of three or more molecules appeared accompanied by protonated molecules of 2 molecules or less. In Ar - 20 (Figure 3), where the moisture content was further increased, protonated molecules with no more than 2 water molecules disappeared, and the other cluster group extended to the high region. As the water content further increased (Figure 3, A-30), the cluster was further expanded and moved to a higher area.

LI method [17] which impregnated absorbent cotton with water to vaporize high concentration water at the tip of the corona discharge tube also presented the same mass spectrum pattern as ours. It was clear that the content of water in the argon gas was related to the protonation of the compound and the formation of a water addition protonated molecule. In the case of helium gas, two molecules of water protonated and hydronium ion did not appear. However, cluster ions of water with weak ionic strength and narrow spectral width appeared in the hydrated state of Figure 3, He-10 and Figure 3, He-20. It was unexpected that a cluster ion of water appeared, since the ionization energy of helium gas is larger than the excitation energy of argon. It seems that there was a factor that excites helium gas in light radiation from the lamp. The cluster ion did not expand to the mass of the high region even when the water content further increased. It is unlikely that helium gas destroys the water clusters, so it seems to be due to synchrotron radiation. Therefore, it seems unsuitable for increasing hydronium ion. It was suggested that the clusters shown in Ar-20 and Ar-30 in Fig.3 were generated by excitation of argon outside the lamp from the excitation of helium enclosed in the lamp. Manufacturers suggested that different gases other than argon gas are enclosed. In the case of nitrogen gas, water-related ions were not observed even when it was anhydrous or even containing water. The energy of the nitrogen gas excited with the excited argon was as small as 6.3 eV [18], which was too small to ionize the water.

Figure 3: Effect of gas species and moisture content on hydronium ion or water cluster formation

He and Ar in the figure represent helium and argon gas species, respectively. Wet gases (argon or helium) were prepared by mixing water-contained gas (numbers in the figure) with a water-free gas and setting the total aeration rate to 100 ml / min.

Alcohols

An alcohol-containing gas generated by bubbling argon gas into alcohol [19] (IE ; methanol:10.6 eV, ethanol:10.5 eV) was flowed into the discharge tube. As shown in Figure 4, alcohol having the lower energy than the energy of excited argon did not show molecular ions by Penning ionization. Instead, hydronium ions and water molecule 2 protonated ions appeared due to moisture contained in alcohol and gas phase moisture in the atmosphere, and one molecule of protonated alcohol appeared by hydronium ion. However, since the amount of water present is very small, cluster ions are not observed. Methanol also showed a spectrum pattern similar to ethanol (Figure 4). It is obvious that water contributes to the protonation of the compound, so we tried it by flowing hydrated alcohol with different concentrations into the discharge tube. The spectrum of ethanol added with water was shown in Figure 5. The alcohol cluster was formed with 1/1 ethanol water (50%: concentration of whiskey). In 1/30 ethanol water (3%: beer concentration) with a further increased water content, many of the ions added with water appeared besides alcohol clusters. The influence of water was large, and in each case, the amount of ions increased by the proton transfer reaction of hydronium ion more than in the case of anhydrous case. This method shows that it is an excellent ionization method to investigate the relationship between alcohol and water cluster. Wakisaka [19] carried out examining clusters of alcohol, water, alcohol water similarly using ESI. But, it gave different spectra from us. It seems to be a difference in acquisition of ions to be generated forcibly or gently.

Figure 4: Mass spectra of methanol and ethanol.

Figure 5: Mass spectra of ethanol water with different water contents.

The mixing ratio of ethanol and water was represented by the ratio of the volume.

t-Butyl Peroxylaurate (Heat Sensitive Substance)

The characteristic of ionization using a low pressure mercury lamp is that unlike the corona discharge method, water necessary for production of protonated molecules can be supplied in the immediate vicinity of the excitation light. Utilizing this characteristic, ionization was attempted by gradually heating t-butyl peroxylaurate (molecular weight 272) sensitive to heat by adding water to argon gas. In the case of water-free argon (Figure 5A), a protonated molecule (m/z: 273) was also observed, but it was thermally decomposed to generate a fragment ion of m/z:201 as a base peak.

Figure 6: Effect of moisture content in argon affecting pyrolysis of t-Butyl peroxy laurate.

Water content in argon gas: A is a nonaqueous system, and B is the same condition as in Figure 3 Ar-10.

Therefore, when argon was hydrated (under the same condition as in Figure 3 Ar-10) in order to suppress decomposition, the fragment ions decreased and the ionic strength of the protonated molecule remarkably increased (Figure 5B). Water acts to suppress the decomposition of t-butyl peroxylaurate, consistent with what we expected to contribute to the production of protonated molecules. This approach can be used to determine the presence and decomposition position of the functional group of a thermosensitive molecule by controlling the moisture content of argon. When the moisture content was further increased, the fragment ion (m/z :201) completely disappeared and ions with numerous water appeared.

Effect of Addition Reagents

Water Free Argon Gas

The formation of protonated molecules of peroxides and formation of addition reagent ions has been studied by the LPI method [21]. As in the previous report, addition reagents were examined with higher electronegativity than oxygen of the functional group of peroxide (PBL). Next, as shown in Figure 1, the supply method of the addition reagents was performed by impregnating quartz wool from the bottom of the sample holder. Water-containing argon suppresses thermal decomposition, but many ions groups with water clusters added to one molecule of PBL also appear (Figure 5B). However, used of addition reagents, those ion groups disappeared (Figure 6). The proton affinity (Pas) of the additive reagents [20] was larger than that of water, so it was a natural result. Incidentally, the proton affinity (Pas) is 166.5 kcal/mol for water and 180 kcal/mol or more for others. Acetone had no effect. Rather, it promoted thermal decomposition of PBL. Acetone had no effect. Rather, it promoted thermal decomposition of PBL (Figure 6-Ac). It seems that protons (H⁺) connecting the molecule and the addition reagent were lost by acetone by blocking the action of water. Ethyl alcohol presented the same spectrum (Figure 6 EtOH) as water. The cluster ion added to the compound turned alcohol instead of water. The effect of the number of alkyl groups of the molecule linked to the proton was the ethylamine with the least hydrogen (Pas: 217.0 kcal / mol) presented the cleanest spectrum. As the number of alkyl chains increased from diethylamine (Pas: 220.6 kcal / mol) (Figure6 DEA) to triethylamine (Pas: 225.1 kcal / mol) (Figure 6 TA) the reagent added ionic strength decreased. Pyridine (Pas: 220.3 kcal / mol) (Figure 6 Py) also presented a simple spectrum like ethylamine.

Figure 7: Effect of addition reagent on hydrated argon gas.

Supply of addition reagent is the same method as in Figure 6.

 Argon Containing Water

The effect of water was remarkable and the synergistic effect of water and addition reagent showed a clean spectrum of fewer ions group appearance than waterless argon (Figure 7).

In the case of acetone, decomposition of PBL was suppressed by water in argon, acetone acted as an addition reagent, and [M + Ac + H]⁺ became the base peak. Acetone used as an addition reagent seems to be an effective method for structural analysis of thermal hypersensitivity compounds by comparing anhydrous and hydrated of argon. Acetone used as an addition reagent seems to be an effective method for structural analysis of thermal hypersensitivity compounds by comparing anhydrous and hydrated of argon. Supply of addition reagent is the same method as in Figure 6. Acetone used as an addition reagent seems to be an effective method for structural analysis of thermal hypersensitivity compounds by comparing anhydrous and hydrated of argon.

Figure 8: Effect of addition reagent on hydrated argon gas.

Supply of addition reagent is the same method as in Figure 6.

Conventionally, ionization under atmospheric pressure by a low-pressure mercury lamp has been considered not to ionize compounds in consideration of excitation energy of mercury. It was found that the protonated molecule of the compound appears by flowing argon gas to the outside of the low pressure mercury lamp. When water was contained in the gas, the amount of ions became more pronounced and could be controlled. This method means that ionization is possible even in an environment containing a large amount of water. It is ionization different from enrichment of droplets of protogenic solvent of ESI. Like ESI, it can be applied to thermal hypersensitivity compounds. It is ionization different from enrichment of droplets of protogenic solvent of ESI. Like ESI, it can be applied to thermal hypersensitivity compounds. In particular, unlike ESI, additional reagents can be used, so this method has versatility.

This ionization method has applied for a patent from anagawa University.

1. Cody RB, Laramée JA, Durst HD. Versatile new ion source for the analysis of materials in open air under ambient conditions. Anal Chem. 2005; 77: 2297-2302.
2. Tsuchiya M, Taira T. A new ionization method for organic compounds--atmospheric pressure/liquid ionization using penning effect and chemical ionization. J Mass Spctrom Soc Jpn. 1978; 26: 333-342.
3. Jiri Sevclk. Detectors in Gas Chromatography, Elsevier Scientific Publishing Company, Amsterdam (1976), p.130 and NIST NIST Chemistry Webbook. 
4. Harada Y. J Vac Soc Jpn. 2005; 48: 397.
5. Penning FM. Uber ionisation durch metastabile atome. Naturwissenshaften. 1927; 15: 818.
6. Cody RB. Observation of molecular ions and analysis of nonpolar compounds with the direct analysis in real time ion source. Anal Chem. 2009; 81: 1101-1107.
7. Tsuchiya M. Development, principle and application of liquid ionization mass spectrometry. Anal Sci. 1998; 14: 661-676.?
8. Mochida Y, Yokoyama Y, Kawai T, Nakamura S, Masahiko M. Liquid Ionization mass spectrometry of methyl hydroperoxyoleates and their related compounds. J Mass Spectrom Soc Jpn. 1994; 42: 207-215.

9. Tachikawa H, Takada H. Proton transfer rates in ionized water clusters (H₂O)_n (n = 2–4). Low Temperature Science. 2005; 64: 31.

10. Snow KB, Thomas TF. Mass spectrum, ionization potential, and appearance potentials for fragment ions of sulfuric acid vapor. Int J Mass Spectrom Ion Processes. 1990; 96: 49-68.

11. Lossing FB, Tanaka I. Photoionization as a source of ions for mass spectrometry. J Chem Phys. 1956; 25: 1031.
12. Franck J, Hertz G. Discrete electron states in an atom: The Franck-Hertz experiment. Verhand Deut Physik Ges. 1914; 16: 457.
13. Genuit W, Chen HN, Boerboom AJH, Los J. Selective ion source for trace gas analysis. Int J Mass Spectrom Ion Phys. 1983; 51: 207-213.
14. Tsuchiya M, Kuwabara H. Cluster ions observed by liquid ionization mass spectrometer. J Mass Spectrom Soc Jpn. 1982; 30: 305-
15. Searcy JQ, Fenn JB. Clustering of water on hydrated protons in a supersonic free jet expansion. J Chem Phys. 1974; 61: 5282.
16. Lee SW, Cox H, Goddard WA, Beauchamp JL. Chemistry in Nanodroplets:  Studies of Protonation Sites of Substituted Anilines in Water Clusters Using FT-ICR. J Am Chem Soc. 2000; 122: 9201-9205.
17. Tsuchiya M, Aoki E, Kuwabara H. Clusters of water under atmospheric pressure studied by field ionization and liquid ionization mass spectrometry. Int J Mass Spect Ion processes. 1989; 90: 55 -70.
18. Litttlewood AB. Gas Chromatography second edition. 1970; 325.
19. CRC Handbook of Chemistry and Physics, 73, D. R. Lides(ed), CRC Press 1993.
20. Wakisaka, BUNSEKI KAGAKU, Vol. 59, p. 743 (2010).
21. Mochida Y, Yokoyama Y, Kawai T, Nakamura S, Tsuchiya M. Liquid Ionization Mass Spectrometry of Methyl Hydroperoxyoleates and Their Related Compounds. J Mass Spectrom Soc Jpn. 1994; 42: 207-215.