Hydrogen compounds formation of calcium in an ICP-DRC-MS

Abstract

Isobaric ions overlap an isotope of interest in mass spectrometry, and an inductively coupled plasma-dynamic reaction cell-mass spectrometry (ICP-DRC-MS) analysis also bears the same problem. In this research, processes of ⁴⁰CaH⁺ formation and decreasing methods were studied. Plasma power was swept from 1300 W to 600 W to investigate ⁴⁰CaH⁺ formation in the plasma. 1.90 × 10⁻⁴ of ⁴⁰CaH⁺/⁴⁰Ca⁺ with 1300 W decreased to 1.27 × 10⁻⁴ when 750 W of plasma power was applied. Deionized water of samples was replaced with heavy water to verify if ⁴⁰CaH⁺ is formed in the interface part of the equipment. Normally NH₃ gas is injected into the DRC to remove argide ions for a measurement of calcium samples in an ICP-DRC-MS, but we found NH₃ contributes to form hydrogen compounds in the DRC by introducing a strontium sample, which is the same alkaline earth metal as calcium. 1.59 × 10⁻⁴ of ⁴⁰CaH⁺/⁴⁰Ca⁺ at 0.8 ml min⁻¹ of NH₃ cell gas decreased to 1.72 × 10⁻⁵ with 0.85 ml min⁻¹ of additional O₂cell gas injection. Results verify that hydrogen compounds made from NH₃ in an ICP-DRC-MS could affect ultra trace isotope ratio analysis.

---

1. Introduction

Isobaric ion interference is a troublesome issue in ICP-MS analysis. One of the critical problems is an overlap of hydrogen compound ions on analyte ions because the mass shift due to the hydrogen formation is just one, where another isotope of the same element often exists. For example, ²³⁵UH⁺ overlaps on ²³⁶U⁺ when measuring ²³⁶U/²³⁸U.¹²⁴⁰Pu/²³⁹Pu determination also has the same kind of problems because of ²³⁸UH⁺ and ²³⁸UH₂⁺.²¹²⁷IH₂⁺ formation can become critical for ¹²⁹I/¹²⁷I detection.³ High Resolution ICP-MS, such as double focusing sector-field ICP-MS can be utilized to avoid these isobaric effects but equipment sensitivity is greatly decreased. We are developing a novel apparatus combining an inductively coupled plasma-dynamic reaction cell-mass spectrometry (ICP-DRC-MS) and ion trap-laser cooling technique to analyze trace isotopes. ICP-DRC-MS will be adopted here as an ion source because it enables liquid sample introduction. Laser based isotope selection technique⁴ will be adopted to raise isotope selectivity and sensitivity, and avoid isobaric interferences. ICP-MS requires only simple modifications. A detector of ICP-MS can be replaced with an ion guide to drag ions to trap segments in our apparatus. In the case of double focusing sector field ICP-MS, for example, ion guides have to deal with all trajectories from a magnetic sector. However, only one trajectory should be considered in the case of ICP-MS.

One of our goals is to develop this apparatus to measure a long-lived radioactive ultra trace isotope, such as ⁴¹Ca. Since Raisbeck and Yiou⁵ proposed the use of ⁴¹Ca as a radiometric dating tool because of its low isotope abundance ratio (⁴¹Ca/Ca = 10⁻¹² − 10⁻¹⁶) and long half-life (1.04 × 10⁵ yr),⁶ it has been used in a biomedical study⁷ to perform research on calcium metabolism inside a human body, in a planetary science⁸ investigation using ⁴¹Ca to date a meteorite or speculate a composition of early solar system. ⁴¹Ca has also been proposed as a dating tool in archeology⁹ since calcium is a major element of a bone and its half-life is applicable to date geological features. While Accelerator Mass Spectrometry (AMS), Resonance Ionization Mass Spectrometry (RIMS) and Atom Trap Trace Analysis (ATTA) can be used to measure and analyze ultra trace isotopes,^10–12 low decay energy (3.3 keV)⁶ and natural abundance of ⁴¹Ca makes the measurement difficult. Although only AMS is routinely utilized for ⁴¹Ca analysis, it requires large facilities and complicated pretreatments¹³ of analyte samples. Our apparatus will be able to become one of the candidates to measure trace isotopes such as ⁴¹Ca. Isobars can be avoided with the apparatus since ions of interest are selectively detected with our laser technique,¹⁴ but the amount of isobars is undoubtedly better to be small to raise sensitivity even with the laser technique. We found that with ⁴⁰CaH⁺, the hydrogen compound of calcium ion is formed in the ICP-DRC-MS. Although this can possibly decrease the sensitivity of the apparatus by overlapping on ⁴¹Ca counts, there is no report investigating ⁴⁰CaH⁺ formation in an ICP-MS equipped with a dynamic reaction cell (DRC) as far as we know. Accordingly the ⁴⁰CaH⁺ formation in commercial ICP-DRC-MS should be investigated before modifying the equipment.

In this article, ⁴⁰CaH⁺ formation will be studied under various experimental conditions of the ICP-DRC-MS with calcium and strontium samples. First, we will investigate the dependence of ⁴⁰CaH⁺ formation in the plasma on plasma rf power because low energy states of calcium ions form less ⁴⁰CaH⁺ than higher energy states of them. Secondly we will replace the solvent of the samples from deionized water to heavy water to examine ⁴⁰CaH⁺ formation in the interface region. The amount of ⁴⁰CaH⁺ will be expected to decrease because of formation of ⁴⁰CaD⁺. Finally, we will investigate the effect of NH₃ cell gas on hydrogen compounds formation in the DRC using calcium and strontium samples, both of which are alkaline earth metals. Various combinations of NH₃ and O₂cell gases will be analyzed to study ⁴⁰CaH⁺ formation in the DRC.

2. Experiment

Instrumentation

Measurements were performed with an ELAN DRC II (PerkinElmer Sciex, Concord, Ontario, Canada) equipped with a PFA-50 Microflow Nebulizer and Cyclonic Spray Chamber. We warmed up the ICP-MS for 45 min before experiments following the manufacturer's instruction. Dwell time was set to 1 ms to reduce a low frequency noise^15–17 such as a plasma flicker noise, peristaltic pump noise and 1/f noise. Two kinds of cell gases were prepared to inject into the DRC separately or simultaneously. Gas flow unit for the cell gases is Ar-equivalent sccm. All other instrumental parameters to maximize ⁴⁰Ca⁺ intensities according to the ELAN DRC II optimization process are listed in Table 1.

Table 1 Instrumental parameters

rf power	1100 W
Nebulizer gas flow rate	0.99 L min^−1
Auxiliary gas flow rate	1.2 L min^−1
Plasma gas flow rate	15 L min^−1
Lens voltage	Optimized daily for maximum signal
Rpa	0
Rpq	0.25
Cell rod offset	−2 V
Quadrupole rod offset	−7.5 V
Cell path voltage	−19 V
Axial field voltage	275 V
Detector dead time	60 ns
Dwell time	1 ms
Number of sweeps	1000
Number of readings	1
Number of replicates per sample	10

---

Materials and reagents

Purities of argon for a plasma gas, ammonia and oxygen gas for a DRC were all higher than 99.999% (Air Liquide Japan). 100 mg L⁻¹ of AccuTrace reference standard calcium for ICP-MS (Accustandard Inc, New Haven, CT) was diluted ten times (10 mg L⁻¹) and acidified with 1% v/v nitric acid for ultratrace analysis (Wako Chemical, Ltd., Osaka, Japan) to make a calcium solution sample. Strontium solution was prepared along the same step but diluted one hundred times (1 mg L⁻¹) because of its low background scale. All the water used in the experiments was doubly distilled and passed through a Millipore Reagent Water System (18.2 MΩcm). Glass distilled deuterium oxide of 99.9 atom% D was obtained from ISOTEC (Isotec Inc., Miamisburg, OH). Synthetic resin containers were washed with 5% v/v nitric acid to remove residual contaminations. More than two minutes of measurement gap was maintained between each measurement to avoid memory effect.

3. Results and discussions

Isobaric ions of ⁴¹Ca⁺ at mass 41

Various isobaric ions of ⁴¹Ca⁺ can appear in ICP-MS¹⁸ such as ⁴⁰ArH⁺, ⁴⁰CaH⁺, ⁴¹K⁺, ²⁹Si¹²C⁺, ²⁸Si¹³C⁺ and ²³Na¹⁸O⁺ at m/z = 41. Based on our measurement of Si and Na, the total amount of three isobars which are ²⁹Si¹²C⁺, ²⁸Si¹³C⁺ and ²³Na¹⁸O⁺ was estimated to be less than 1% of m/z = 41 count even with an assumption that 100% of minor isotopes contribution is from their composition. Therefore they were considered here negligibly small. ⁴⁰ArH⁺ is removed with NH₃ cell gas, which can be confirmed with slope change of the analyte ion¹⁹ and we affirmed the ArH⁺ suppression was successful by using a blank sample. ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio fluctuated within error bars between 1 mg L⁻¹ to 10 mg L⁻¹ of sample concentration, but its fluctuation became larger at less than 1 mg L⁻¹ of concentration because backgrounds from experimental environment might affect the result. Since ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio is so small (10⁻⁴ − 10⁻⁶) and several ppt orders of backgrounds are difficult to avoid even with clean room facility,¹⁷ we used highly concentrated (10 mg L⁻¹) samples to observe ⁴⁰CaH⁺ formation as a function of experimental parameters clearly. We confirmed ⁴⁰CaH⁺/⁴⁰Ca⁺ had no noticeable dependence on concentration of sample. Low concentration samples could be used by increasing integration time but backgrounds or count fluctuations from analyte samples could become a critical issue in our case. Backgrounds from deionized water and resins could be neglected because they were comparatively smaller than the counts of highly concentrated samples. Then the rest of the main candidates at m/z = 41 can be ⁴¹K⁺ and ⁴⁰CaH⁺. The amount of ⁴¹K⁺ can be estimated from the natural abundance ratio of potassium and we made an assumption that m/z = 39 was wholly ³⁹K⁺. After ⁴¹K⁺ subtraction from the count at m/z = 41, all the experiments were conducted under the assumption that ⁴⁰CaH⁺ was dominant at m/z = 41. Two ways of ⁴⁰CaH⁺ formation can be considered as follows,

Ca⁺ + H(D) ↔ CaH⁺(CaD)⁺ (1)

Ca + H⁺(D⁺) ↔ CaH⁺(CaD)⁺ (2)

Although the reaction (1) would be dominant in ⁴⁰CaH⁺ formation since the degree of ionization of calcium in the plasma is over 95%,²⁰ a small amount of the reaction (2) was also considered to occur since we are interested in measuring very low concentrations. We conducted experiments to investigate where ⁴⁰CaH⁺ formation occurs inside the ICP-DRC-MS.

⁴⁰CaH⁺ formation inside the plasma

Though plasma usually disintegrates molecules to atoms and then ionizes them in ICP, it also forms singly charged polyatomic molecular species.²¹ Since calcium ions of excited states transform into ⁴⁰CaH⁺ at higher probability than those with ground energy states²² and plasma power of ICP affects the energy states of an ion,²³ the ⁴⁰CaH⁺ formation is expected to be varied as a function of the input plasma power. Fig. 1 shows the reaction profiles of ⁴⁰Ca⁺ and ⁴⁰CaH⁺. NH₃ was maintained at 0.8 ml min⁻¹ to suppress argide ions. The counts of the two ions decreased with lowering plasma power since the reduced plasma temperature affected the ionization rate in ICP. With decreasing plasma power from 1300 W, reduction of ⁴⁰CaH⁺ became larger than that of ⁴⁰Ca⁺ and it gave rise to decrease the ratio of ⁴⁰CaH⁺/⁴⁰Ca⁺ gradually. The ratio suddenly increased between 700 to 600 W of the plasma power due to abrupt increase of ⁴⁰CaH⁺. This phenomenon is supposed to be caused by many metastable calcium ions, which are especially feasible to form ⁴⁰CaH⁺, and are created at the specific plasma power. Despite the abrupt ratio increase between 700 to 600 W, a larger ⁴⁰CaH⁺ reduction than ⁴⁰Ca⁺ was maintained on lowering the plasma power. Consequently we could verify ⁴⁰CaH⁺ formation occurs inside the plasma and can be suppressed to some extent by lowering the plasma temperature.

Fig. 1 Normalized counts of ⁴⁰Ca⁺, ⁴⁰CaH⁺ (left axis) and the ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio (right axis) as a function of the rf power measured at optimum conditions (see Table 1).

⁴⁰CaH⁺ formation inside the interface region

If hydrogen in analyte samples affects the composition, it would occur in the cooled interface region located next to the ICP part. In order to verify this, solvent of samples was replaced from deionized water to heavy water. ⁴⁰CaH⁺ was expected to decrease here since several groups^24,25 reported that replacement of H₂O solvent to D₂O showed the mass shift of the hydrogen compound in an ICP-MS. Argide ions were suppressed here by injecting 1 ml min⁻¹ of NH₃ in order to measure calcium compounds clearly.

Although the shift of the mass ⁴⁰CaH⁺ to ⁴⁰CaD⁺ was expected due to the solvent replacement, no evidence of the mass shift was found. Fig. 2(a) shows ⁴⁰CaH⁺/⁴⁰Ca⁺ grew as increasing the ratio of heavy water in the sample. The reason why the clear mass shift which as Vais and Zoriy showed^24,25 did not appear is supposed to be due to the interferences from heavy water contributing to the formations of isobaric ions at m/z = 41. Fig. 2(b) shows that the amounts of Na and Si ions increases linearly as the amount of heavy water in the sample grows. They were possibly transformed into ²³Na¹⁸O⁺, ²⁸Si¹³C⁺ and ²⁹Si¹²C⁺. So we could not evaluate if ⁴⁰CaH⁺ formation occurs in the interface region. Furthermore, investigating ⁴⁰CaD⁺ formation to find ⁴⁰CaH⁺ variation indirectly was impossible since the amount of ⁴⁰CaD⁺ composition is so small compared to ⁴²Ca⁺. A D₂O purification system adapted to the same process as deionized water might be required to remove this isobaric effect because there was no successful solution which removed the interferences completely as far as we tested (99.96 atom% D of Aldrich (Aldrich, St. Louis, MO) and 99.9 atom% D of CIL (Cambridge Isotope Laboratories, Andover, MA) were also examined).

Fig. 2 Reaction profiles of ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio (a) and the amount of Na ions and Si ions (b) as a function of the concentration of heavy water in the sample. Experiment (b) was exclusively conducted under the conditions of 100 ms of dwell time, 10 times of sweep, 5 times of replicate.

Hydride compound formation of Ca and Sr in the DRC

NH₃ cell gas injection into the DRC is inevitable when analyzing calcium with an ICP-DRC-MS to remove argide ions, but NH₃ could possibly offer hydrogen ions forming ⁴⁰CaH⁺ in the DRC even though Ca⁺ and NH₃ reacts endothermically.²⁶ Thus ⁴⁰CaH⁺ formation was examined as a function of NH₃ gas flow. Fig. 3 shows the normalized counts of ⁴⁰Ca⁺, ⁴⁰CaH⁺ and ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio. 1.5 ml min⁻¹ of NH₃ gas flow achieved the lowest ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio of 3.9 × 10⁻⁵, possibly due to the physical scattering with gas molecules. Because an average energy of ions in ICP is between 5 and 30 eV,²⁷ the collisional cross section between analyte ions and gas molecules will depend on their physical radii.¹⁹ So the amount of ⁴⁰CaH⁺ which has bigger physical radii than ⁴⁰Ca⁺ is expected to decrease faster than ⁴⁰Ca⁺. Fig. 3 indeed shows the higher decreasing rate of ⁴⁰CaH⁺. In Fig. 3, ⁴⁰Ar⁺ starts to be removed from 0.2 ml min⁻¹ of NH₃ gas flow, and then a collisional focusing¹⁹ appears from the same gas flow rate. However 0.8 ml min⁻¹ of NH₃ was required to observe the pure count of ⁴⁰CaH⁺ because of its small count scale. Thus it is difficult to observe count variation of ⁴⁰Ca⁺ and ⁴⁰CaH⁺ directly when the NH₃ injection was indeed started in order to evaluate the NH₃ contribution to the ⁴⁰CaH⁺ formation.

Fig. 3 Normalized counts of ⁴⁰Ca⁺, ⁴⁰CaH⁺ (left axis) and ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio (right axis) as a function of the NH₃ gas flow.

Because direct measurement of NH₃ effect on hydrogen compound formation was impossible due to argide ions, a strontium sample was introduced. Strontium is expected to show a similar reaction profile compared to that of calcium since both of them are the same alkaline earth metals. Therefore, the reaction profiles of calcium and strontium samples with O₂cell gas were investigated. O₂ was used to form OH compounds of each sample. Fig. 4 shows OH compound reaction profiles of the calcium and strontium with O₂cell gas. OH compound formation was saturated at 0.8 ml min⁻¹ of the gas and decreased by scattering with gas molecules in both experiments. The reason why ⁴⁰Ca⁺ count was larger than that of ⁸⁸Sr⁺ when O₂ gas flow was small is ⁴⁰Ar⁺ at the same m/z. ⁴⁰Ar⁺ was gradually removed by O₂, which is capable of removing argide ions the same as NH₃¹⁹ since the ionization energy²⁸ of oxygen (12.07 eV) is located between those of argon (15.76 eV) and calcium (6.11 eV), the same as ammonia (10.07 eV). However, because the reaction coefficient between O₂ and argon ions (k = 2.3 − 9.00 × 10⁻¹¹) is lower than that of NH₃ (k = 1.60 − 1.84 × 10⁻⁹),²⁹ 1 ml min⁻¹ of gas flow was needed to remove ⁴⁰Ar⁺ and ⁴⁰ArH⁺ though 0.8 ml min⁻¹ was enough in the case of NH_3.⁸⁸Sr⁺ showed pure counts without O₂ since no argide ions interfered on the count. The count was increased by collisional focusing, and decreased from 0.6 ml min⁻¹ of gas flow rate since scattering with gas molecules became dominant. We concluded that the calcium and strontium show similar reaction aspects from this result and investigated reaction profiles of strontium as a function of each cell gas to evaluate that of calcium.

Fig. 4 OH compounds reaction profiles of the calcium and strontium samples as a function of O₂cell gas. Concentration of both samples is 1 mg L⁻¹.

Fig. 5 shows the reaction profiles of strontium with two different gases. Increase of all analyte ions at small gas flow rate here is due to the collisional focusing. ⁸⁸SrH⁺ notably increased with the NH₃ cell gas injection compared to O₂. If the same chemical reaction occurs in the case of calcium, it can be assumed that NH₃ might contribute to form hydrogen compounds in the DRC. Although calcium ions and NH₃ are not considered to react because the reaction is known to be endothermic, ⁴⁰CaH⁺ formation cannot be neglected in our experiments. We suppose metastable calcium ions formed in the plasma drifted into the DRC and formed ⁴⁰CaH⁺. ⁴⁰CaH⁺ formation in the DRC was further investigated to look for the possibility of suppression by injecting two cell gases simultaneously.

Fig. 5 Reaction profiles of ⁸⁸Sr⁺, ⁸⁸SrH⁺ in the ICP-DRC-MS as a function of O₂ and NH₃ gas flow rate respectively. Concentration of the sample is 1 mg L⁻¹.

⁴⁰CaH⁺ formation in the DRC

Complicated chemical reactions occur inside the DRC.¹⁹ Two types of reaction were assumed to occur when O₂ was additionally injected into the DRC which already created ⁴⁰CaH⁺ inside with NH₃. First is hydrogen reduction. The amount of hydrogen which reacts with calcium ions is supposed to decrease since the reaction constant of oxygen with hydrogen (k = 1.00 × 10⁻⁹) is much higher than that of calcium (k = 2.00 × 10⁻¹⁴).²⁹ Therefore O₂ reacts with hydrogen before calcium ions react with it. For example, the following reaction²⁹ can be considered,

H⁺ + O₂ → Products (3)

Second is decomposition or transformation of ⁴⁰CaH⁺. O₂ injection is expected to make ⁴⁰Ca⁺ by stealing hydrogen ions from ⁴⁰CaH⁺ or form ⁴⁰CaOH⁺ by offering oxygen ions to ⁴⁰CaH⁺. For example, the following reaction can be considered,

⁴⁰CaH⁺ + O₂ → ⁴⁰Ca⁺ + Products (4)

or

⁴⁰CaH⁺ + O₂ → ⁴⁰CaOH⁺ + Products (5)

In any case, ⁴⁰CaH⁺ is reduced if the reaction successfully occurs.

In this experiment, only O₂ gas flow rate was swept in the DRC but NH₃ was not since investigation of O₂ effect was more important here. Two NH₃ gas flow rates explained in Fig. 3 were chosen. One is 0.8 ml min⁻¹, which is the lowest flow rate for removing both ⁴⁰Ar⁺ and ⁴⁰ArH⁺. Another is 1.5 ml min⁻¹, which realizes the largest ⁴⁰CaH⁺ suppression by the physical scattering. Fig. 6 shows both ⁴⁰Ca⁺ and ⁴⁰CaH⁺ ions decreased with increasing O₂ flow in all cases, because the scattering between ions and gas molecules increased. However the decreasing slopes of the ⁴⁰CaH⁺ when 0.8 ml min⁻¹ of NH₃ injection was steeper than the case of 1.5 ml min⁻¹ of NH₃. O₂ gas ratio in the cell when NH₃ was maintained at 0.8 ml min⁻¹ is higher than the case of 1.5 ml min⁻¹. So the ion's axial energy damping is expected to be more active when NH₃ was maintained at 0.8 ml min⁻¹ since heavier collisional gas is more effective at damping ion's energy.¹⁹ As ⁴⁰CaH⁺ damping increases, their residence time in the cell becomes longer and the chemical reaction probability increases. Therefore, a steeper slope of ⁴⁰CaH⁺ reduction is expected at 0.8 ml min⁻¹ of NH₃ with O₂cell gas. Whereas ⁴⁰Ca⁺ was reduced at constant speed regardless of NH₃ flow rate, ⁴⁰CaH⁺ slope depended on the cell gas ratio combinations. We assume the ⁴⁰CaH⁺ reduction reactions occurred here.

Fig. 6 Reaction profiles of ⁴⁰Ca⁺, ⁴⁰CaH⁺ as a function of O₂ gas flow rate in the ICP-DRC-MS. NH₃ gas flow was fixed at 0.8 and 1.5 ml min⁻¹ respectively.

Fig. 7 is a reproduction of Fig. 6 to show ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio as a function of the total cell gas flow rate. O₂ flow was swept with fixed NH₃ gas flows at 0.8 and 1.5 ml min⁻¹, respectively. The ratio depending on the gas combinations was observed. The lowest ratio was achieved at 0.8 ml min⁻¹ of NH₃ and 0.85 ml min⁻¹ of O₂ flow rate combination. We are convinced this achievement benefits from the maximal use of multiple collision and selective reaction mechanism²³ of the DRC. 0.8 ml min⁻¹ of NH₃ successfully removed interfering argide ions but prevented ⁴⁰Ca⁺, ⁴⁰CaH⁺ from extra physical scattering. Additionally injected O₂ made the residence time of ⁴⁰CaH⁺ longer in the cell by collision and chemically reacted with ⁴⁰CaH⁺. 0.85 ml min⁻¹ was verified as the most effective O₂ flow rate. The fact that simultaneous multiple cell gas injection could alter hydrogen compound formation in the DRC was realized from these results.

Fig. 7 Reaction profiles of the ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio as a function of the combination of cell gas flow rate. Traces adopting triangle markers start from fixed NH₃ flow rate because O₂ was added from the flow rate.

Bandura³ could decrease ¹²⁷IH₂⁺/¹²⁷I⁺ by increasing post-cell energy barriers and in-cell retarding field strength but we could not. We suppose that the collision effect is dominant in their experiment but additional hydrogen compound formation by chemical reaction with NH₃ was dominant in our case.

4. Conclusions

In this paper ⁴⁰CaH⁺ formation in the ICP-DRC-MS was studied. ⁴⁰CaH⁺ formation depends on the plasma power and could be suppressed by decreasing plasma power to some extent. The solvent of samples was replaced from deionized water to heavy water to find ⁴⁰CaH⁺ variation in the interface region. However it was not able to observe the reactions since too many impurities from heavy water had interfered calcium analysis. Highly purified D₂O may be needed to investigate the reactions properly. Although the reactions of calcium ions with NH₃ are not usually considered, we found NH₃ can possibly affect the reaction of hydrogen compound with calcium and strontium in the DRC when determining isotope ratio of 10⁻⁴ to 10⁻⁵ from the experiment. Hydrogen compound formation in the DRC was varied according to the cell gas combination even at the same total cell gas flow rate. Furthermore, additional injection of O₂ into the cell could achieve the lowest ⁴⁰CaH⁺/⁴⁰Ca⁺ ratio.

Acknowledgements

The authors appreciate the technical assistance with ICP-DRC-MS of K. Kobayashi, the pure deionized water supplied by D. Hiroishi and valuable discussions with M. Kitaoka.

References

1. S. F. Boulyga and J. S. Becker, Fresenius J. Anal. Chem., 2001, 370, 612–617.
2. I. Rodushkin, P. Lindahl, E. Holm and P. Roos, Nucl. Instrum. Methods Phys. Res., Sect. A, 1999, 423, 472–479.
3. D. R. Bandura, V. I. Baranov and S. D. Tanner, J. Am. Soc. Mass Spectrom., 2002, 13, 1176–1185.
4. S. Hasegawa, L. Matsuoka, Y. Fukushima, H. Osaki and Y. Hashimoto, J. Nucl. Sci. Technol., 2006, 43, 300–304.
5. G. M. Raisbeck and F. Yiou, Nature, 1979, 277, 42–44.
6. M. Paul, I. Ahmad and W. Kutschera, Zeitschrift für Physik A Hadrons and Nuclei, 1991, 340, 249–254.
7. R. L. Fitzgerald, D. J. Hillegonds, D. W. Burton, T. L. Griffin, S. Mullaney, J. S. Vogel, L. J. Deftos and D. A. Herold, Clin. Chem., 2005, 51, 2095–2102.
8. G. Srinivasan, S. Sahijpal, A. Ulyanov and J. Goswami, Geochim. Cosmochim. Acta, 1996, 60, 1823–1835.
9. D. Fink, J. Klein and R. Middleton, Nucl. Instrum. Methods Phys. Res., Sect. B, 1990, 52, 572–582.
10. D. Fink, R. Middleton, J. Klein and P. Sharma, Nucl. Instrum. Methods Phys. Res., Sect. B, 1990, 47, 79–96.
11. K. Wendt and N. Trautmann, Int. J. Mass Spectrom., 2005, 242, 161–168.
12. I. D. Moore, K. Bailey, Z. Lu, P. Müller, T. P. O'Connor and L. Young, Nucl. Instrum. Methods Phys. Res., Sect. B, 2003, 204, 701–704.
13. K. Nishiizumi, M. W. Caffee and D. J. DePaolo, Nucl. Instrum. Methods Phys. Res., Sect. B, 2000, 172, 399–403.
14. Y. Hashimoto, D. Nagamoto and S. Hasegawa, Int. J. Mass Spectrom., 2009, 279, 163–169.
15. S. Sturup, L. Bendahl and B. Gammelgaard, J. Anal. At. Spectrom., 2006, 21, 297–304.
16. I. S. Begley and B. L. Sharp, J. Anal. At. Spectrom., 1994, 9, 171–176.
17. D. R. Bandura, V. I. Baranov and S. D. Tanner, J. Anal. At. Spectrom., 2000, 15, 921–928.
18. J. Becker, K. Füllner, U. Seeling, G. Fornalczyk and A. Kuhn, Anal. Bioanal. Chem., 2008, 390, 571–578.
19. S. D. Tanner, V. I. Baranov and D. R. Bandura, Spectrochim. Acta, Part B, 2002, 57, 1361–1452.
20. P. A. D. Groot, Handbook of Stable Isotope Analytical Techniques, Volume I, Elsevier Science, 2005, p. 694.
21. H. E. Taylor, Inductively coupled plasma-mass spectrometry: practices and techniques, Academic Press, 2001, pp. 12–13.
22. A. Boutalib, J. P. Daudey and M. El Mouhtadi, Chem. Phys., 1992, 167, 111–120.
23. D. R. Bandura, V. I. Baranov, A. Litherland and S. D. Tanner, Int. J. Mass spectrom., 2006, 255–256, 312–327.
24. V. Vais, C. Li and J. Cornett, J. Anal. At. Spectrom., 2004, 19, 1281–1283.
25. M. V. Zoriy, L. Halicz, M. E. Ketterer, C. Pickhardt, P. Ostapczuk and J. S. Becker, J. Anal. At. Spectrom., 2004, 19, 362–367.
26. Scott D. Tanner and Vladimir I. Baranov, At. Spectrosc., 1999, 20.
27. J. P. Guzowski Jr and G. M. Hieftje, J. Anal. At. Spectrom., 2001, 16, 781–792.
28. S. G. Lias, “Ionization Energy Evaluation” in NIST Chemistry WebBook, NIST Standard Reference Database Number 69, ed. P. J. Linstrom and W. G. Mallard, National Institute of Standards and Technology, Gaithersburg MD, 20899, http://webbook.nist.gov, (retrieved October 31, 2010).
29. V. G. Anicich, An index of the literature for bimolecular gas phase cation–molecule reaction kinetics, NASA/Jet Propulsion Laboratory, Report #03-19, 2003.

---