Oxygenation Mechanism of Ions in Dynamic Reaction Cell ICP-MS

Transcription

1 ANALYTICAL SCIENCES JULY 2013, VOL The Japan Society for Analytical Chemistry Oxygenation Mechanism of Ions in Dynamic Reaction Cell ICP-MS Tomohiro NARUKAWA and Koichi CHIBA National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 3, Umezono, Tsukuba, Ibaraki , Japan A dynamic reaction cell (DRC) is one of the most effective tools for eliminating spectral interferences caused by polyatomic molecules in inductively coupled plasma mass spectrometry (ICP-MS). Oxygen gas (O 2), by producing oxygenated ions, is very effective in reducing some specific spectral interferences. In this study, the oxygenation of elemental ions (M ) in the DRC was investigated experimentally, and a new explanation for oxygenation based on the enthalpy changes in the oxygenating reactions is proposed. The enthalpy changes of each M were calculated and the possibility of each reaction occurring was evaluated. The calculations were in good agreement with experimental observations. Theoretical and experimental results supported the hypothesis that the enthalpy changes (ΔH) of M O 2 MO O and M O MO and the thermodynamic stability of M O are key factors controlling oxygenation of M in the DRC. Keywords ICP-MS, dynamic reaction cell, enthalpy change, polyatomic molecule, reaction gas, oxygenation mechanism (Received December 14, 2012; Accepted May 14, 2013; Published July 10, 2013) Introduction Since its introduction in the 1980s, inductively coupled plasma mass spectrometry (ICP-MS) has continued to advance and it is now the most widely used technique for elemental analysis. It is of great value to researchers engaged in inorganic analytical chemistry as well as to those performing routine analysis, because ICP-MS has very high sensitivity, a wide dynamic range, isotope detection ability, and a capability for isotope dilution analysis. However, spectral interferences caused by isobaric ions, polyatomic molecules, or doubly charged ions can cause measurement problems, particularly in quadrupole ICP-MS, which is more widely used than the double-convergence method. 1 When an argon (Ar) plasma is used as the excitation source, Ar ions and Ar-based polyatomic molecular ions can cause severe spectral interferences. Many techniques to control and suppress interferences caused by isobaric ions and polyatomic molecular ions have been reported. Avoiding interference by choosing a measurement mass that is free from interferences is possible if the analyte has two or more available isotopes. A correction equation is a common and effective method for eliminating interferences. 2 Interferences caused by coexisting compounds in samples can be avoided by sample pretreatment in which the interfering compounds are removed by, for example, coprecipitation or ion-exchange. 3 5 Uses of a desolvation apparatus for sample introduction, of a low-temperature plasma, and a high-resolution mass spectrometer are effective ways of reducing and removing To whom correspondence should be addressed. tomohiro-narukawa@aist.go.jp interferences In recent years, methods employing collision/reaction gas cells using He, H 2, NH 3, CH 4, or O 2 have been introduced into ICP-MS to remove the interferences of coexisting ions In particular, when He is used as the collision gas, spectral interferences caused by molecular ions are effectively reduced. All ions in the collision/reaction cell lose their kinetic energies in collision with He atoms, but molecular ions lose more kinetic energy than do elemental ions. Therefore, if the electric voltage at the entrance to the MS is correctly adjusted, the interfering molecular ions can be excluded from the MS although elemental ions are introduced. 17,18 The mechanism is termed the kinetic energy discrimination effect (KED). However, the collision/reaction mechanisms of the other reaction gases are more complicated. Several mechanisms, such as a collisioninduced dissociation (CID), deionization, charge transfer, proton reactions (protonation, proton shift, proton transfer), oxidation reactions, ion-molecule reactions, and cluster generation, have been proposed and they may interrelate with each other. Tanner et al. comprehensively reviewed the history of reaction/collision cells and the advances in ICP-MS. 19 Many reports on the effects of the collision/reaction gas on interferences caused by polyatomic molecular ions have also been published These pioneering studies have provided information on, advanced the usefulness, and extended the applicability of ICP-MS. Molecular ion interferences have been discussed from the aspect of experimental and thermodynamic data, fundamental information on the chemical reactions of analytes and reaction gases has been provided, and measurement techniques to overcome spectral overlaps by using ion-molecule reactions in a quadrupole reaction cell have been proposed. 24 The dependency of the effects of the reaction gas and the behavior of interfering molecules on the DC voltage of Rpa and Rpq, 25 r.f. devise,

2 748 ANALYTICAL SCIENCES JULY 2013, VOL. 29 band-pass, and kinetic reactions have also been reported, 19 and the importance of rf energy noted. 25 This information has contributed to an understanding of interferences and helped in the selection of optimum ICP-MS operating conditions. Explanations of the effects of reaction gases in the dynamic reaction cell (DRC) have been based upon the thermochemistry of the reaction between the ions and the reaction gases. However, statistical explanations for the chemical reaction of ions and O 2 are lacking, although there has been much discussion concerning neutral atoms and the cell gas. Reaction cell mechanisms, based on reference data, have been discussed in terms of thermal energy, kinetic rates, entropy, free energies of the atoms involved, and molecular and chemical reactions Although these recent discussions of reaction cell mechanisms might be thought sufficient, there is a lack of consideration of reaction systems and reaction routes and a reliance on reference data rather than calculation of data specific to the systems under investigation. In addition, although ions are detected in ICP-MS analysis, there is a lack of thermal energy data for the ions. Several mechanisms based upon kinetic rates that use available reference data, entropies and reaction kinetics related to the activation energy of the chemical reaction have sometimes been proposed to discussions of the effect of collision/reaction gases. But, such information cannot be obtained from ICP-MS measurements because only the final products are observed by ICP-MS instrumentation. If we discuss reaction kinetics under these conditions, we must consider the tunneling effect in any proposed system; this has not yet been considered. When free energy is included in a discussion of mechanism, entropy and temperature should be considered together with the other parameters, but these have been ignored. The reaction energy, that is the enthalpy change (ΔH), is the correct parameter to be considered in thermochemical discussions. Some previous studies have explained the effects of reaction gases based on the thermochemistry of the reaction between ions and reaction gases. However, most of the discussions are simplistic, applicable to limited systems, and lacking in rigour. In this paper, we discuss the mechanism of oxygenation of ions in dynamic reaction cell ICP-MS based upon the thermochemistry of the reaction. It is established that when As is measured by ICP-DRC-MS, 75 As 16 O (m/z = 91) is measured instead of 75 As in order to avoid the interference of 40 Ar 35 Cl and/or 40 Ca 35 Cl (m/z = 75) Oxygenation is an effective way of avoiding spectral interferences by producing an oxide ion free from interfering molecular ions or by oxygenating the isobaric interfering ions However, the effectiveness of O 2 has not been evaluated sufficiently and the underlying mechanism of oxygenation in the reaction cell has not been fully elucidated, although the mechanisms underlying the use of NH 3 and CH 4 have been discussed by several researchers. We investigated the mechanism of oxygenation by using the bond energy and the enthalpy change (ΔH) between the ions and O, and demonstrated the usefulness of theoretical calculations for elucidating reaction mechanisms in DRCs. Experimental Apparatus The inductively coupled plasma mass spectrometer used was a NexIon S300 (Perkin Elmer Inc., Waltham, MA) equipped with a DRC. O 2 was used as the reaction gas. A micromist nebulizer and a cyclonic spray chamber were used for sample introduction. Typical operating parameters were as follows: incident rf power was 1600 W, outer Ar gas flow rate 18 L min 1, intermediate Ar gas flow rate 1.0 L min 1, carrier Ar gas flow rate 1.0 L min 1, and Rpq and Rpa were set at 0.65 and 0, respectively. Some previous works 19,24 on reaction gases have reported that the band-pass conditions can affect the efficiency of the elimination of spectral interferences. The Rpq parameter for the band-pass can range from 0.25 to 0.65; however, in the current study, for the most part, the plasma operating conditions were not varied and the Rpq band-pass was kept constant at Thus, we were able to investigate the reactions occurring in the DRC. Procedure Eighteen elements, namely, As, Bi, Cd, Co, Cu, In, Mn, Ni, P, Pb, Re, Rh, Sb, Se, Tl, Y, Yb and Zr, were measured by ICP-DRC-MS with and without introducing O 2 as a reaction gas. All mass numbers were measured. In addition, the mass numbers of each oxygenated molecular ion (m/z = M 16) were monitored. Blank values were also monitored at each mass number and subtracted from the total intensity. Theoretical calculation Gaussian 03W software (Gaussian Inc., Pittsburgh, PA) was used for the calculation of theoretical energy values. The structural optimum energies of the three-dimensional atoms, molecules and ions, and their vibrational and single point energies were calculated with a high precision energy method (density functional theory: DFT). However, the calculation precision for elements of the fifth period of the periodic table was much lower than that for the first four periods. Moreover, the DFT method cannot be applied to the calculations of elements of the sixth period. When an element of the fifth period was included in a target system, another calculation model, less precise than the DFT, was used for calculation of the energies of the whole reaction even though the energy of oxygen could be calculated precisely by DFT. Published thermochemistry data 26 29,44 are shown together with the results in order to validate the calculations. In this study, the enthalpy change between reactants and products was calculated. When the reaction system A B C D is exothermic, the reaction of A B occurs spontaneously without external energy. In other words, when the thermochemical equation is written as A B C D ΔH and when ΔH = H(C D, product) H(A B, reactant) is negative, the reaction occurs without a supply of external energy. On the other hand, when ΔH is positive, the reaction is endothermic and does not occur spontaneously. Results and Discussion Measurement of oxygenated molecules To investigate oxygenating reactions in the DRC, the mass number of each element (M ) and its oxygenated molecule (MO ; M 16) were monitored with and without introducing O 2 into the DRC. The flow rate of O 2 was kept constant at 1.0 L min 1. It is known that As is converted to AsO in ICP-DRC-MS when O 2 is introduced into the DRC as a reaction gas. At first, the production rates of AsO were measured for each O isotope (m/z = 16, 17 and 18). The results are summarized in Table 1. Only As was observed at mass number 75 when no O 2 was supplied. However, when O 2 was supplied, As was not observed at mass number 75, whereas AsO was

3 ANALYTICAL SCIENCES JULY 2013, VOL Table 1 Measurement counts and percentage conversions to oxygenated derivatives of As compounds Analyte Without O 2 gas With O 2 gas Count/cps Abundance, % Count/cps Abundance, % Natural isotope abundance of O, % 75 As 91 AsO 92 AsO 93 AsO ( 16 O: 99.76) ( 17 O: 0.04) ( 18 O: 0.20) observed at mass numbers 91, 92 and 93 with the signal intensities in accord with oxygen isotopic abundances (the isotope abundances of oxygen 16, 17 and 18 are 99.76, 0.04 and 0.20%, respectively). Thus, 100% As is converted to AsO when O 2 is the reaction gas. The conversion percentage was estimated from the oxygenation rate of analyte ions in the DRC. The monoisotopic elements P, Mn, Co, Y, Rh and Bi and their oxygenated ions were measured in the same manner. The percentages converted to MO were 95.6, 14.2, 27.3, 100.0, 1.4 and 0.2%, respectively. The measured conversion rate for P (m/z = 31) of 95.6% was considered to be an underestimation because of spectral interference by NO and NOH ; the actual value must be 100%. Elements with multi isotopes, namely, Se, Zr, Pb, Yb, Sb, Cu, In, Re, Ni, Cd and Tl, were examined by the same procedure. As examples, the results for Se, Zr and Pb are shown in Table S1 (Supporting Information). Without the addition of O 2, their mass spectra were in accord with their natural isotope abundances; the spectra shifted to 16 masses higher when O 2 was introduced. The percentage conversions to oxygenated molecules with the addition of O 2 were Se 25.6%, Zr 100.0%, Pb 0.2%, Yb 6.7%, Sb 32.0%, Cu 0.3%, In 0.0%, Re 2.6%, Ni 14.3%, Cd 3.6% and Tl 0.0%. On the basis of these results in the actual measurements, the elements were categorized into three groups; As, P, Y, and Zr were 100% oxygenated, Mn, Co, Se, Sb, and Ni were 10 to 30% oxygenated, and Rh, Bi, Pb, Yb, Cu, In, Re, Cd, and Tl were < 5% oxygenated. Effect of oxygen gas flow rate The effect of O 2 flow rate on the production of oxygenated molecules was investigated by varying the flow rate to the DRC from 0.0 to 3.0 L min 1. For elements with a high conversion rate (almost 100%), such as As, P, Y and Zr, the production of their oxygenated derivatives (MO ) increased from 0 to 100% with the O 2 flow increasing from 0 to 0.5 L min 1, and then the rates remained constant at approximately 100% when the flow rate was above 0.5 L min 1. The production rates of oxygenated molecules of the other elements increased to values less than 100% when the O 2 flow rate was increased from 0 to 0.5 L min 1, and then remained constant at those values regardless of the O 2 flow rate when it was above 0.5 L min 1. Mechanism of oxygenation in DRC In previous reports, 19 the effects of NH 3 and CH 4 as reaction gases were discussed in terms of ionization potential (IP). When the IP of the interfering ion is higher than that of the reaction gas, and that of the analyte ion is lower than that of the reaction gas, the charge on the interfering ion transfers to the reaction gas, but that of the analyte ion does not. The interfering ions are thus deionized to neutral atoms or molecules whereas the analyte ions still remain as ions. As a result, the interfering ions are eliminated and the analyte ions are measured without spectral interference. According to this explanation, in which the IP and charge transfer are essential factors, even O 2 has the potential to induce the charge transfer reactions in the DRC because the 1st IP of O 2, ev (reference value, none), falls between those of NH 3 and CH 4. The value ev was higher than that of NH 3 (10.00 ev; reference value, ev), but 1.25 ev lower than that of CH 4 (13.81 ev; reference value, none). The 1st IP of O 2 would thus appear capable of inducing charge transfer. However, the major effect of O 2 as a reaction gas is not based upon charge transfer but upon the production of oxygenated derivatives. Until now, there have been no adequate reports considering the mechanism of oxygenation. In order to understand the mechanism of oxygenation in the DRC, the bond strengths (bond energies) of the mono-oxides and mono-oxide ions were considered. Arsenic was used as an example since the use of the DRC with O 2 is effective in eliminating spectral interference for As. The bond energy of As O is 4.72 ev (reference value, 4.99 ev) and is smaller than that of O=O (5.28 ev, reference value, 5.17 ev). Thus, it is readily apparent that a single O atom prefers to combine with another O atom rather than with As; that is, O is unlikely to react with As. But, experiments show that when O 2 is used in ICP-DRC-MS, As is converted to AsO and the latter is measured. Therefore, the ionization of AsO generated in the DRC cannot be a major process. The reactions between M and O 2 were then investigated because the major chemical species in the DRC are elemental ions (M ), polyatomic ions and O 2. The bond energies of M O were calculated and are given in Table 2. The bond energies of P O, Y O, As O and Zr O were greater than that of O=O and those of the other 14 oxygenated molecular ions (M O) were smaller. As mentioned above, P, Y, As and Zr were completely converted to MO, although Se, Sb, Co, Ni and Mn were oxygenated to less than 30% and the other elements were not oxygenated at all. The oxygenating reactions of elements in the DRC were investigated taking As as an example. The potential reactions of As and As with O 2 are as follows: (1) As O 2 AsO O, (2) As O 2 AsO 2, (3) As O 2 As O 2, (4) As O 2 AsO O, (5) As O 2 AsO O, (6) As O 2 AsO O e, (7) As O 2 AsO 2 e. The reactions (1), (2), (6) and (7) are not relevant in ICP-DRC-MS since As atoms are not a main component in the DRC. Only ions, such as As and Ar, are introduced into the DRC through the ion lens of the instrument, whereas neutral elements and molecules are eliminated at the ion lens. Although the reactions (1), (2), (6) and (7) are not expected to occur in the DRC, the enthalpies of all the reactions were calculated. The thermochemical reaction equations and their calculated enthalpy changes are shown in Table 3-1. As can be seen, the most energetically favorable product is AsO because only reaction (4) is exothermic and the others are endothermic. The conversion of O 2 to 2O requires the bond dissociation energy of 5.28 ev (reference value, 5.17 ev), which is less than the bond

4 750 ANALYTICAL SCIENCES JULY 2013, VOL. 29 Table 2 Calculated bond energies of M O Chemical bond Y O P O As O Zr O (O=O) Se O In O Sb O Co O Ni O Mn O Cd O Cu O Rh O Cu O Bi O Pb O Yb O Re O Tl O Bond energy/ev (5.28) (singlet) Fig. 1 Relationship between percentage oxidation and ΔH in the thermochemical equations of M O 2 = MO O ΔH. Table 3-1 Thermal chemical reaction formulas of As, As and O 2 As O 2 = AsO O 0.56 ev As O 2 = AsO ev As O 2 = As O ev As O 2 = AsO O 0.94 ev As O 2 = AsO O 4.74 ev As O 2 = AsO O e 8.93 ev As O 2 = AsO 2 e 7.75 ev (1) (2) (3) (4) (5) (6) (7) Table 3-2 Thermal chemical reaction formulas of M and O 2 Y O 2 = YO O 8.92 ev P O 2 = PO O 2.44 ev Zr O 2 = ZrO O 6.56 ev Se O 2 = SeO O 1.37 ev In O 2 = InO O 2.93 ev Sb O 2 = SbO O 0.12 ev Co O 2 = CoO O 2.80 ev Ni O 2 = NiO O 2.30 ev Mn O 2 = MnO O 2.82 ev Cd O 2 = CdO O 0.54 ev Rh O 2 = RhO O 0.23 ev Cu O 2 = CuO O 3.28 ev Cu O 2 = CuO O 5.22 ev (singlet) dissociation energy of As O. Therefore, once AsO is formed in the DRC, it is more stable than O 2 and becomes a major chemical species in the DRC. These calculated results agreed with the experimental observation that AsO was observed by ICP-DRC-MS when O 2 was introduced and As was not. The thermochemical reaction equations (M O 2 MO O) of other elements and the calculated enthalpy changes are summarized in Table 3-2 and Fig. 1. The reactions for P, Y and Zr, which were almost 100% oxygenated, were exothermic, and therefore favorable like that for As. In contrast, the reactions of the elemental ions giving low or no yields of oxygenated products were endothermic and were therefore unlikely to occur spontaneously. Table 4 Thermal chemical reaction formulas (M O MO /M O MO ) Co O = CoO 3.20 ev Ni O = NiO 2.98 ev Cu O = CuO 2.00 ev Cu O = CuO 0.06 ev* Mn O = MnO 2.47 ev Rh O = RhO 0.73 ev Sb O = SbO 3.23 ev Se O = SeO 3.91 ev In O = InO 3.42 ev Ar O = ArO 2.43 ev *, Singlet. Co O = CoO ev Ni O = NiO 8.57 ev Cu O = CuO 8.53 ev Cu O = CuO 6.59 ev* Mn O = MnO 9.15 ev Rh O = RhO 2.71 ev Sb O = SbO 1.88 ev Se O = SeO 8.13 ev In O = InO 5.06 ev Co, Ni, Mn, Se and Sb were oxygenated to between about 15 and 30% although their reactions were endothermic. This suggests that another reaction process was facilitating their oxygenation to MO. The enthalpy changes for the reactions M O MO and M O MO were calculated and are given in Table 4. As can be seen, both reactions for Co, Ni, Mn, Se and Sb are exothermic, and therefore are favorable and progress spontaneously. M O MO is less likely to occur than M O MO because the neutral atom M does not pass through the ion lens into the DRC. Possibly, a very small amount of M is generated by collision between M and a third species in the DRC. The concentration of O is extremely low in the DRC. However, there must be an adequate amount of O atoms in the DRC, and M is introduced into the DRC through the ion lens. Where does the O come from? When large amounts of highly reactive ions such as As and P are present, reactions such as As O 2 AsO O can supply O. However, in this case, the production rate of O depends on the quantity of As and only a small amount of O would be produced. Thus, this may not be the main route to O in the DRC. However, large amounts of N, O, H and OH are always introduced into the DRC because ICP instrumentation is used in a normal atmosphere and samples are usually supplied to the instrument as aqueous solutions.

5 ANALYTICAL SCIENCES JULY 2013, VOL Ar / 52 ArO was The ion count ratio of 36 Ar / 52 ArO increased to 6486 when O 2 was supplied. Thus, the net ion count of 36 Ar increased and that of 52 ArO decreased when O 2 was supplied. These results indicated that ArO O Ar O 2 occurs mainly in the DRC. Experimental observations in ICP-DRC-MS agree well with our proposed explanation based upon enthalpy change. The signal intensity of ArO as measured by ICP-DRC-MS is consistent with a large amount of ArO forming in the argon plasma and passing through the DRC to the mass spectrometer. When the reaction of Cu with O is considered, the triplet spin state of CuO appears to be relevant. However, calculation results of the energy of CuO in the singlet state are better in agreement with the results of ICP-MS. Therefore, the spin state of Cu may change in the high energy plasma. Details of this process are currently being investigated. Fig. 2 Relationship between percentage oxidation and ΔH in the thermochemical equations of chemical reaction model: M O = MO ΔH. Solid line, Cu singlet was used as the ΔH value of Cu ; dotted line, Cu triplet was used as the ΔH value of Cu. The reaction N O 2 NO O 6.3 ev and O O 2 O O ev are exothermic, although the reactions H O 2 HO O 0.3 ev, OH O 2 HOO O 3.3 ev, NO O 2 NO 2 O 2.1 ev and NO 2 NO O 3.2 ev are endothermic. Therefore, the former two reactions are likely to be the main processes to supply enough O to oxygenate the above elements. The enthalpy change in M O MO for Cu is exothermic, but the change is too small to yield an adequate amount of MO. For In, whose conversion was 0%, both In O InO and In O InO are endothermic; so it is clear that In is not oxygenated in the DRC. Figure 2 shows the relationship between the reaction yields obtained in this experiment and the calculated ΔH values for the reaction M O MO. There is a strong positive linear correlation between the reaction rates and ΔH when ΔH was greater than 0. The concentration of O in the DRC, from neutralized O from the plasma, can be no more than 2 ppm. 19 Reaction with O may be difficult because of an insufficient total number of collisions. The production of Ar O from Ar was therefore investigated. The bond dissociation energy of Ar O (2.43 ev) is smaller than that of O=O (5.28 ev; reference value, 5.17 ev), which means Ar O is less stable than O=O. In addition, the reaction of Ar O 2 ArO O is endothermic and is not favorable. Thus ArO is unlikely to be generated by the reaction between Ar and O 2 in the DRC. However, Ar has a chance to react with O to produce Ar O when O is available as a by-product of reactions in the cell, since the reaction is exothermic. ArO O Ar O 2 is also exothermic. Therefore, ArO is dissociated by collision with O, even if it is generated by the collision of Ar with O. To investigate reactions involving Ar, ArO and O 2 in the DRC, 36 Ar (m/z = 36, isotope abundance 0.337%) and 36 Ar 16 O (m/z = 52) were monitored with and without introducing O 2 into the DRC. Without O 2, the conversion of ArO to Ar was 0.02%, and the ion count ratio of Advantages and problems of using oxygen When O 2 is used as the reaction gas in the DRC, the oxygenation rate varies from one element to another; that is, some ions are oxygenated to 100%, some to a certain intermediate percentage, and others not at all. However, the oxygenation rate is constant for each element and the repeatability of measurements is excellent under a particular set of conditions. Therefore, if an element is subject to spectral interference by polyatomic molecular ions and if it can be oxygenated to some extent in the DRC, it can be readily determined by measuring the oxygenated derivative ions instead of the original ion. Furthermore, if there is a large difference in the oxygenation of the analyte and the isobaric interfering ion, the isobaric interference can be eliminated. Selective oxygenation is sometimes useful for simultaneous measurement of elements present in a wide concentration range under the same ICP-MS operating conditions. If the oxygenation rate of an element present at a high concentration is very low, a very small amount of its oxide ion will be generated in the DRC. Therefore, we can measure such oxide ions together with the elements that are present in similarly low concentrations. Thus, elements present in low and high concentrations can be simultaneously measured under the same operating conditions, and this is of considerable practical value. However, there are also risks of unexpected spectral interferences because unanticipated MO ions can be generated by introducing O 2 into the DRC. For example, MoO and ZrO will interfere with Cd in ICP-MS analysis when O 2 is introduced into the DRC, and NiO will interfere with the Se measurement when a large amount of Ni is present in a sample, although the oxygenation rate of Ni to NiO is only about 14%. Conclusions In this study, the mechanism of oxygenation of ions in ICP-DRC-MS was elucidated by using theoretical calculations based on enthalpy changes (ΔH) in reactions between elemental ions (M ) and O 2. The reaction M O 2 MO O is the main process for oxygenating chemical species in the DRC when the reactions are exothermic. For example, As, P, Y, and Zr are oxygenated by O 2 to MO almost completely. The other possible process is the reaction M O MO, although the oxygenation rates are lower than those for M O 2 MO O. For example, Se, Sb, Ni, Mn, Co and Sb are partially oxygenated through the above reaction and their oxygenation rates are less than 30%. Such information is very useful for selecting a reaction gas for the DRC that will

6 752 ANALYTICAL SCIENCES JULY 2013, VOL. 29 eliminate spectral interferences and allow for the prediction of new ones. In addition, we are studying the enthalpy changes between metal ions and other gases such as NH 3 and CH 4. Information obtained will be additionally useful in selecting a reaction gas. In actual measurements, the experimental equipment or operating conditions such as those of r.f. devices and band-pass DC voltage will influence the production rate of oxygenated ions, but they are not the essential factors governing selective oxygenation. The removal of interference fundamentally depends on the enthalpy changes in the oxygenation reaction involved. The improvement of calculation precision, especially for the elements of the fourth period of the periodic table, will help us to understand the behavior of the chemical species in the DRC and facilitate the effective removal of spectral interferences. The work reported here demonstrates that technological advances in ICP-MS can be brought about by experimental observation in conjunction with theoretical calculations. Supporting Information The production rates of oxygenated molecules of Se, Zr and Pb. This material is available free of charge at analsci/. References 1. S. Tanner and V. Baranov, At. Spectrosc., 1999, 20, E. H. Van Veen, S. Bosch, and M. T. C. De Loos-Vollebregt, Spectrochim. Acta, Part B, 1994, 49B, C. F. Brown, P. E. Dresel, K. N. Geiszler, and O. T. Farmer, J. Anal. At. Spectrom., 2006, 21, K. Inagaki, A. Takatsu, A. Uchiumi, A. Nakama, and K. Okamoto, J. Anal. At. Spectrom., 2001, 16, Y. Luo, E. Dabek-Zlotorzynska, V. Celo, D. C. G. Muir, and L. Yang, Anal. Chem., 2010, 82, S. D. Tanner, J. Anal. At. Spectrom., 1995, 10, J. R. Bacon, J. S. Crain, L. Van Vaeck, and J. G. Williams, J. Anal. At. Spectrom., 1999, 14, J. Sabine Becker and H.-J. Dietze, J. Anal. At. Spectrom., 1999, 14, J. Sabine Becker and H.-J. Dietze, J. Anal. At. Spectrom., 1998, 13, C. Sariego Muñiz, J. M. Marchante-Gayón, J. I. García Alonso, and A. Sanz-Mede, J. Anal. At. Spectrom., 1998, 13, M. Moldovan, E. M. Krupp, A. E. Holliday, and O. F. X. Donard, J. Anal. At. Spectrom., 2004, 19, D. W. Koppenaal, G. C. Eiden, and C. J. Barinaga, J. Anal. At. Spectrom., 2004, 19, C. F. Harrington, R. Clough, L. R. Drennan-Harris, S. J. Hill, and J. F. Tyson, J. Anal. At. Spectrom., 2011, 26, A. Taylor, S. Branch, M. P. Day, M. Patriarca, and M. White, J. Anal. At. Spectrom., 2011, 26, J. Olesik, C. Hensman, S. Rabb, and D. Rago, Spec. Publ. -R. Soc. Chem., 2001, 267, D. Bandura, S. Tanner, V. Baranov, G. Koyanagi, V. Lavov, and K. Bohme, Spec. Publ. -R. Soc. Chem., 2001, 267, D. R. Bandura, V. I. Baranov, and S. D. Tanner, Fresenius J. Anal. Chem., 2001, 370, D. J. Douglas, J. Am. Soc. Mass Spectrom., 1998, 9, S. Tanner, V. Baranov, and D. Bandura, Spectrochim. Acta, Part B, 2002, 57B, H. Gurleyuk, R. Brunette, C. Howard, C. Schneider, and R. Thomas, Spectroscopy, 2005, 30, V. Epov, D. Lariviere, E. Epova, and D. Evans, Geostand. Geoanal. Res., 2005, 28, F. Vanhaecke, L. Balcaen, I. Deconninck, I. De Schrijver, C. Almeida, and L. Moens, J. Anal. At. Spectrom., 2003, 18, B. Hattendorf and D. Gunther, Spectrochim. Acta, Part B, 2003, 58B, J. Olesik and D. Jones, J. Anal. At. Spectrom., 2006, 21, V. I. Baranov and S. D. Tanner, J. Anal. At. Spectrom., 1999, 14, Database, Diethard Bohme, York University, chem.yorku.ca/profs/bohme/research/selection_table.html. 27. V. G. Anicich, An Index of the Literature for Bimolecular Gas Phase Cation-Molecule Reaction Kinetics, JPL Publication 03-19, National Aeronautics and Space Administration, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California. 28. G. K. Koyanagi, D. Caraiman, V. Blagojevic, and D. K. Bohme, J. Phys. Chem. A, 2002, 106, D. R. Bandura, V. I. Baranov, A. E. Litherland, and S. D. Tanner, Int. J. Mass Spectrom., 2006, 255, M. Grotti and R. Frache, J. Anal. At. Spectrom., 2007, 22, D. Wallschläger and J. London, Environ. Sci. Technol., 2008, 42, M. He, B. Hu, Y. Zeng, and Z. Jiang, J. Alloys Compd., 2005, 390, F. Ardini, F. Soggia, F. Rugi, R. Udistib, and M. Grotti, J. Anal. At. Spectrom., 2010, 25, J. Jarrett, G. Xiao, K. Caldwell, D. Henahan, G. Shakirova, and R. Jones, J. Anal. At. Spectrom., 2008, 23, V. Taylor, R. Evans, and R. Cornett, Anal. Bioanal. Chem., 2007, 387, S.-F. Kan and P. Tanner, Environ. Chem., 2006, 3, J. Ejnik, T. Todorov, F. Mullick, K. Squibb, M. McDiarmid, and J. Centeno, Anal. Bioanal. Chem., 2005, 382, C.-H. Yang and S.-J. Jiang, Spectrochim. Acta, Part B, 2004, 59B, C.-C. Chang, H.-T. Liu, and S.-J. Jiang, Anal. Chim. Acta, 2003, 493, V. Vais, C. Li, and J. Cornett, Anal. Bioanal. Chem., 2003, 377, V. Baranov, Z. Quinn, D. Bandura, and S. Tanner, J. Anal. At. Spectrom., 2002, 17, D. Bandura, V. Baranov, and S. Tanner, Anal. Chem., 2002, 74, L. Simpson, M. Thomsen, B. Alloway, and A. Parker, J. Anal. At. Spectrom., 2001, 16, R. C. Weast, Handbook of Chemistry and Physics, 76th ed., 1995, CRC Press.