High-precision cerium isotope analysis by thermal ionization mass spectrometry using the Ce⁺ technique

Abstract

The ¹³⁸La–¹³⁸Ce isotope system has been regarded as a useful radiogenic tracer for geochronology studies. Compared to the commonly-used CeO⁺ technique, the measurement of Ce isotope ratios as Ce⁺ is more straightforward and more advantageous, but it is challenging due to the severe isobaric interference of ¹³⁸Ba on ¹³⁸Ce and large variations in relative abundances of all Ce isotopes. In this study, a novel method has been developed for high-precision measurement of Ce isotope ratios by thermal ionization mass spectrometry (TIMS) as Ce⁺. A newly-developed film porous ion emitter (FPIE) was used to enhance the ionization of Ce as Ce⁺ ions. The employment of TaF₅ as an activator significantly suppressed the Ba⁺ isobaric interference signal. ¹⁴⁰Ce was proposed to be an alternative reference Ce isotope as there is no isobaric interference on ¹⁴⁰Ce and complicated peak tailing correction can be avoided. The combinations of diverse amplifiers (10¹⁰ Ω, 10¹¹ Ω, 10¹² Ω and 10¹³ Ω) were used for the measurement of Ce isotope ratios as Ce⁺ and ¹³⁷Ba was monitored simultaneously on a 10¹³ Ω amplifier for ¹³⁸Ba interference correction. The reproducibility of Ce isotope ratios obtained was ca. 10-fold better than the previously published Ce⁺ results and even comparable with that obtained using the more laborious CeO⁺ techniques. This method was further applied for the analysis of reference rock samples and uranium ores of world-wide origin. The analytical results demonstrated that Ce isotope ratios could be a promising signature for the nuclear forensic investigation to identify the source of unknown nuclear materials.

---

1. Introduction

Rare earth elements (REEs), which exhibit similar physical and chemical properties, can provide insight into numerous geological and cosmic processes.^1–6 Among REEs, only Ce could be naturally oxidized from a trivalent state to a tetravalent state in oxidizing environments, and insoluble Ce(IV) results in Ce anomalies relative to other REEs(III) eventually.^7–9 The ¹³⁸La–¹³⁸Ce decay system is a highly useful tool for geochronology and a geochemical tracer of a diverse range of geological reservoirs on Earth, especially when combined with other popular isotope systems (¹⁴⁷Sm–¹⁴³Nd and ¹⁷⁶Lu–¹⁷⁶Hf).^2,9–12

As a newly-evolved scientific discipline, nuclear forensics mainly aims at identifying the origin and history of the seized or found nuclear materials. Previous studies have shown that the Ce isotopic composition varies with diverse geological conditions and remains unaltered during metallurgical processes to which uranium ores have been subjected, thus providing information on origin assessment of nuclear-related materials.^13–15 The Ce isotope ratios vary only from −3 to +5 ε-units (ε denotes the parts per ten thousand deviation), therefore, Ce isotope measurements with high precision (2σ ≤ 0.8ε or 2RSD ≤ 80 ppm) are required to identify such small natural isotopic variations.^16,17 However, high-precision measurement of the Ce isotopic composition in geological samples is still challenging to date. Ce has four naturally occurring isotopes (¹³⁶Ce = 0.19%, ¹³⁸Ce = 0.25%, ¹⁴⁰Ce = 88.45%, and ¹⁴²Ce = 11.11%), among which the radiogenic ¹³⁸Ce is a minor isotope relative to ¹⁴⁰Ce and ¹⁴²Ce, and their relative abundances vary by two orders of magnitude. Furthermore, ¹³⁸Ce is severely affected by isobaric interference from ¹³⁸Ba, while Ba is distributed widely in the natural environment and the abundance of ¹³⁸Ba is extremely high (71.7%).

To date, with respect to these limitations, high-precision Ce isotope analysis could only be achieved by mass spectrometry. In a few studies, Ce isotope ratios were measured by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS).^18,19 More studies focused on the measurement of Ce isotopes by TIMS,^{12,16–18,20–31} as it is less affected by isobaric and polyatomic interferences. Prior to TIMS measurement, it is imperative to separate Ce from the matrix and isobaric interferences efficiently, especially Ba and other REEs. However, the separation of Ce from other REEs is difficult because Ce and most REEs occur in a trivalent state, and their chemical behaviors are very similar. The conventional method for Ce separation is cation-exchange chromatography, which employs α-hydroxy-isobutyric acid (α-HIBA) and cation exchange resin to eliminate Ba and most REEs. Nevertheless, the recovery of Ce and operational blank of this method are somewhat unsatisfactory.^12,16,32 An innovative separation method was firstly introduced by Rehkämper et al.,³³ and they used NaBrO₃ as a strong oxidizing reagent to transform Ce(III) into Ce(IV), then Ce was selectively extracted into the organic phase (HDEHP) from other REEs. Recently, the oxidation separation method has been adopted widely in geosciences, and it was further used in combination with the cation exchange resin technique.^{26,28,34–36} The remarkably different chemical behaviors between Ce(IV) and other REEs(III) enable its selective separation, and the oxidation separation method proved to be the most effective method to separate Ce from REEs.

For the determination of Ce with TIMS, as summarized in Table S1,† Ce isotope ratios can be measured by the detection of either CeO⁺ ions or Ce⁺ ions since the ionization energies of both ions are relatively low. High-precision Ce isotope ratio (2RSD < 85 ppm) measurement for relatively large amounts (>1 μg) of reference materials (JMC-304, AMES or other standards) has been achieved with the CeO⁺ technique.^{12,16–18,20–28} The CeO⁺ technique has the advantage that potentially interfering isobaric BaO⁺ is minimized, due to the high first ionization potential of BaO⁺ compared to CeO⁺.^18,27 Most of the methods used double or triple Re filaments for Ce measurement. Double oxidized Ta filaments were also used by Nakamura et al.²² Since oxide molecular ions were measured with the CeO⁺ technique, the oxygen supplement in the source chamber is an important aspect of the method. Unlike Nd isotope analysis as NdO⁺,³⁷ direct introduction of O₂ gas into the source through a gas bleed device is not preferred for the CeO⁺ analytical technique, as poor source vacuum would lead to a severe peak tailing effect. Consequently, phosphoric acid was employed as the major oxygen source. As shown in Table 1, for the CeO⁺ technique, interference on CeO⁺ masses can be generated either by neighboring elements or polyatomic species. Therefore, great care must be taken for the efficient separation of Ce from not only Ba but also other REEs (La, Sm, Gd and Dy). In addition, adequate oxygen isotope correction is also required for obtaining accurate and precise Ce isotope ratios. For this purpose, a constant ¹⁸O/¹⁶O value is assumed in some cases.^1,9,21 However, a gradual change of the ¹⁸O/¹⁶O ratio during the measurement has been observed, thus the importance of in situ¹⁸O/¹⁶O measurement is stressed to obtain precise Ce isotope ratios,^{8,12,18,27,38} leading to a much longer measurement time and extra error.

Table 1 Possible interferences on Ce⁺ and CeO⁺ and their abundances (%)

Analytical mode	Ce^+		CeO^+
Mass	138	140	154	156
Ions required	^138Ce^+(0.25)	^140Ce^+(88.45)	^138Ce(0.25)^16O(99.76)^+	^140Ce(88.45)^16O^+(99.76)
Possible interferences	^138Ba^+(71.70), ^138La^+(0.09)	—	^136Ba(7.85)^18O(0.20)^+	^138Ce(0.25)^18O(0.20)^+ 138Ba(71.70)^18O(0.20)^+
			^136Ce(0.19)^18O(0.20)^+, ^138Ba(71.70)^16O(99.76)^+	^138La(0.09)^18O(0.20)^+ 139La(99.91)^17O(0.04)^+
			^137Ba(11.23)^17O(0.04)^+	^156Gd(20.47)^+, ^156Dy(0.06)^+
			^138La(0.09)^16O(99.76)^+
			^154Gd(2.18)^+, ^154Sm(22.75)^+

---

Compared to the CeO⁺ technique, the Ce⁺ technique seems to be a better method as the Ce isotopic composition is directly measured with much less isobaric interference and oxygen isotope corrections are completely avoided. Only a few preliminary attempts, however, were made to measure Ce isotope ratios with the Ce⁺ technique, and the reproducibility obtained could only reach 335–517 ppm,^18,29–31 which was inadequate for application in nuclear forensics research. The poor reproducibility was mainly attributed to relatively weak and unstable Ce⁺ ion beams obtained. In recent years, a PIE (porous ion emitter) was developed by incorporating equal mass of platinum and rhenium powder to form a porous Pt/Re alloy, and it has been reported to significantly enhance the ionization efficiency of trace quantities of actinides.^39–43 More recently, we further proposed a FPIE material, which was more uniform in thickness and easier to prepare, and it has been proven to lead to a relatively constant measurement state and consequently a more stable ion beam in precise measurements of trace neodymium isotopes as Nd⁺ ions by TIMS.⁴⁴

When using the Ce⁺ technique, the main challenge remains the interference of ¹³⁸Ba on ¹³⁸Ce. Ba exists naturally in large quantity and is more prone to be ionized into metal ions than Ce. Another analytical challenge encountered in the Ce⁺ technique arises from the low abundance of ¹³⁸Ce relative to the main isotopes ¹⁴⁰Ce and ¹⁴²Ce. The accuracy and precision of low signal analysis are limited by the detection system of the instrument, and the noise inherent in the resistors used in the feedback loop is the dominant limiting factor in TIMS. The newly developed high ohm (10¹² Ω and 10¹³ Ω) resistor amplifiers make it possible to analyze minor ion beam intensities on Faraday cups. They can be used alone, or in combination with 10¹⁰ Ω and/or 10¹¹ Ω current amplifiers. Additionally, combinations of diverse ohm amplifiers could be applied for simultaneously collecting isobaric interfering species, thus providing accurate isobaric interference correction. High ohm resistor amplifiers have been used in earlier TIMS and MC-ICP-MS studies for various isotope systems (B, U, Nd, Sr, Ta, and Pb).^45–50 However, application and comprehensive evaluation of high ohm resistor amplifiers for Ce isotope analysis as Ce⁺ ions has not yet been reported.

The aim of this study is to develop a new TIMS measurement method for Ce isotope ratio measurement based on the Ce⁺ technique. A newly-developed FPIE filament is used and the ionization behaviors of Ce⁺ and CeO⁺ ions are investigated. We propose the employment of TaF₅ as an activator, combined with the use of amplifiers with diverse ohm resistances (10¹⁰ Ω, 10¹¹ Ω, 10¹² Ω and 10¹³ Ω) to measure Ce isotope ratios as Ce⁺ with high precision. Ce isotope ratios in reference material JMC-304 and reference rock samples are measured for method validation. The applicability of Ce isotope ratios as a possible signature for nuclear forensics is tested by the measurement of several uranium ores of world-wide origin as well.

2. Materials and methods

2.1. Reagents and materials

High-purity water (18 MΩ cm⁻¹) was prepared using a Milli-Q (Millipore) water purification system. The AG 50W-X12 resin (50–100 μm particle size) and LN resin (50–100 μm particle size) were supplied by Bio-Rad Company and TRISKEM International Company, respectively, and they were cleaned before the experiment with 8 M HCl to remove residual Ce. All reagents (HNO₃, HF, HCl and H₃BO₃) used for sample preparation were of ultrapure grade. A certified Ce isotope reference material JMC-304, which was prepared by Tanaka and Masuda,²⁰ was used for method validation. As there is no universally accepted uranium-based reference material against which to report Ce isotopic data nowadays, five reference rock samples: BCR-2 (Basalt), JA-2 (Andesite), JB-3 (Basalt), JG-1 (Granite) and JR-1 (Rhyolite), obtained from Geological Survey of the United States (USGS) and Geological Survey of Japan (GSJ) were selected for testing the Ce separation procedure and isotope ratio measurement. These reference rock materials were different in lithology, and their matrix composition varied significantly. Additionally, eight uranium ore samples with different origins were analyzed as well. The detailed information and photographs of these uranium ore samples are shown in Table 2 and Fig. 1, respectively.

Table 2 The relevant information on eight uranium ore samples

Sample	Ore type	Country	Mine	Size (cm)
1# UO	Pitchblende	The United States	Ruggles	5.1 × 4.8 × 4.1
2# UO	Pitchblende	The United States	Happy Jack	10.5 × 7.5 × 3.6
3# UO	Saleeite	Australia	Radium Hill	11.9 × 5.4 × 5.7
4# UO	Guilleminite	Congo	Musonoi	6.0 × 4.1 × 2.5
5# UO	Metatorbernite	Congo	Musonoi	5.7 × 3.8 × 1.9
6# UO	Metatorbernite	Congo	Musonoi	3.8 × 4.8 × 3.8
7# UO	Samirestie	Madagascar	Ambolotara	1.9 × 1.9 × 1.0
8# UO	Boltwoodite	Namibia	Goanikontes	4.4 × 2.9 × 2.2

---

Fig. 1 Photos of eight uranium ore samples of different origins. The coin for comparison is 25 mm in diameter.

2.2. Sample digestion and separation

Bulk uranium ore samples were crushed and then powdered with the help of a mortar grinder (Fritsch Pulverisette 2) in advance. For Ce isotopic analysis of reference rock samples and uranium ore samples, approximately 0.5 g of the sample powder was weighed into a PFA vessel. The samples were dissolved with an acid mixture (9 mL of 15 M HNO₃ + 6 mL of 20 M HF + 3 mL of 12.4 M HClO₄) on a hotplate at 190 °C for 48 h. After evaporating the sample solutions to dryness, they were treated by dissolving with 2 mL of concentrated HNO₃ and evaporating to dryness three times. Then the sample was re-dissolved with 2 mL of 6 M HCl at 120 °C and heated to dryness again. After that, the sample was treated overnight at 90 °C with 2 mL of 2.5 M HCl (mixed with 3% H₃BO₃) to decompose the fluoride gels that may form during the digestion procedure. Finally, the target fraction was re-dissolved in 2 mL of 1 M HNO₃ to be prepared for the following separation procedure.

As shown in Table 1, compared to the CeO⁺ technique, which needs sufficient removal of several interferents (Ba, La, Sm, Gd and Dy), the main focus in the Ce⁺ technique lies on the effective separation of only Ba and La. Therefore, the chemical separation procedure for the Ce⁺ technique could become relatively simpler and less time-consuming. A two-stage separation procedure (Fig. S1†), which was modified from the methods reported previously,^{26,28,34–36} was used. Firstly, the sample solution was loaded into a Teflon column filled with 1 g AG 50W-X12 resin, which was pretreated with 4 mL of 1 M HNO₃. Main matrix elements were washed out with 8 mL of 5 M HCl. Ba was then eluted with 5 mL of 2 M HNO₃, and the light rare earth elements (LREEs) were subsequently eluted with 10 mL of 5 M HNO₃. The LREE fraction obtained from this column separation proved to be entirely free of matrix elements, but still contained small amounts of Ba, as an effect of column tailing. Secondly, the collected LREE solution was introduced into the second Teflon column filled with 1 g LN resin, which was pretreated with 4 mL of 10 M HNO₃, and the resin top was covered with a porous 30 mm polyethylene filter with the aim of preventing the flotation of resin particles. LREEs(III) were washed out with 6 mL of 10 M HNO₃ containing 20 mM KBrO₃. In this step, Ce(III) was oxidized to Ce(IV) by the strong oxidizing reagent KBrO₃ and thus it remained on the resin. The remaining Ce was then eluted with 10 mL of 6 M HCl solution containing 20 mM H₂O₂ to ensure effective recovery of Ce. The potential interfering element Ba was effectively removed due to its high decontamination factor (ca. 4 × 10⁶) achieved, and the final fraction enriched in Ce was obtained. The eluted fraction was collected for TIMS analysis. It should be noted that a new column was used for every sample to avoid the risk of cross-contamination. The chemical yield of Ce was about 78% and the total Ce operational blank was less than 60 pg, which was negligible compared to the large quantity of Ce presented in the real samples.

2.3 TIMS analysis

High-purity zone-refined Re filaments with single filament geometry were used. We prepared a FPIE material for the accurate measurement of Ce isotope ratios. Equal weight portions of rhenium (0.5 g, Alfa Aesar 325 mesh) and platinum metal powders (0.5 g, Alfa Aesar 325 mesh) were used. The detailed procedures for the preparation of FPIE have been described elsewhere.⁴⁴ The FPIE stock was then sintered to the center of the filament and heated at ∼1800 °C using a Thermo Scientific degas unit, with a vacuum of ∼10⁻⁷ mbar. The resulting FPIE material was a Pt/Re alloy with a porous structure. The optical and scanning electron microscopy (SEM) images are shown in Fig. S2.†

1 μL of the Ce analyte was loaded onto the FPIE filament in 2.5 M HCl, then 1 μL of 0.2 M TaF₅ was loaded as an activator after the sample was evaporated to dryness in air. Once the sample was dried, the filament current was gently increased to 2.5 A, and then immediately turned down. In order to enhance the ionization efficiency, great care was taken to load both the sample and activator right on the center of the FPIE. The Ce isotope measurements were performed as Ce⁺ ions on a Thermo Scientific Triton TIMS at State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences. A multi-collector acquisition routine was employed using collectors L4 to H2, and the configuration of the Faraday cups is given in Line 1 of Table 3. The center cup was connected to a 10¹⁰ Ω amplifier for collection of the large ¹⁴⁰Ce⁺ ion beam, while all other Faraday cups were connected to conventional 10¹¹ Ω amplifiers. Isobaric contribution from ¹⁴²Nd⁺ on ¹⁴²Ce⁺ was monitored by measuring ¹⁴³Nd⁺ ions in cup H2 during data acquisition, while ¹³⁷Ba⁺ and ¹³⁹La⁺ ion beams were also collected by cups L3 and L1 to correct the isobaric interferences of ¹³⁸Ba⁺ and ¹³⁸La⁺ on ¹³⁸Ce⁺, respectively. Due to large variations between the relative abundances of Ce isotopes, the peak tailing of the major isotope ¹⁴⁰Ce severely interferes with the determination of neighboring minor isotopes. To correct the peak tailing effect, half-mass intensities were measured at magnet settings at −0.5 amu offset (Line 2 of Table 3). The correction of low-mass tailing of the ¹⁴⁰Ce peak was performed using the exponential law, which was proposed by Willbold.¹⁸ Then the tail-corrected intensities of ¹³⁸Ce and ¹³⁶Ce were obtained by subtracting the peak-tail of ¹⁴⁰Ce from the measured raw intensities at mass 138 and 136. Afterwards, the mass fractionation correction was performed off-line, and the fractionation factor (β) was determined using the exponential mass fractionation law.

where (¹³⁶Ce/¹⁴²Ce)_m is the measured and tail corrected ¹³⁶Ce/¹⁴²Ce ratio and (¹³⁶Ce/¹⁴²Ce)_t is set to 0.01688.⁵¹ M(¹³⁶Ce) and M(¹⁴²Ce) are the atomic masses of ¹³⁶Ce and ¹⁴²Ce, respectively. Then the mass fractionation corrected ¹³⁸Ce/ⁱCe isotope ratio can be calculated using the equationwhere ⁱCe is the reference Ce isotope (¹³⁶Ce, ¹⁴⁰Ce or ¹⁴²Ce), and (¹³⁸Ce/ⁱCe)_m is the tail corrected ¹³⁸Ce/ⁱCe measured ratio.

Table 3 Faraday cup configurations of multi-collector mode

Faraday cup	L4	L3	L2	L1	C	H1	H2	H3	H4
Line 1	^136Ce	^137Ba	^138Ce	^139La	^140Ce	^142Ce	^143Nd
Line 2	135.5	136.5	137.5	138.5	139.5
Line 3	^142Ce				146				^142Ce^16O

---

Each measurement run consisted of 16 blocks with 11 cycles per block, and the integration time for Line 1 and Line 2 of Table 3 were both set to 4.914 s. The amplifier gain of each Faraday cup and the baseline were calibrated prior to Ce isotope measurement every day. After the FPIE filament was kept at ∼1200 °C for ca. 40 minutes, the ion beam was centered and carefully focused until an optimum intensity was reached, then the measurement of Ce isotope ratios could be started when the vacuum in the ion source became better than 1.1 × 10⁻⁷ mbar.

3. Results and discussion

3.1. Enhancing the ionization of Ce⁺ ions with the FPIE

Ce ionizes as Ce⁺ and CeO⁺ ions in the thermal ion source of TIMS, and the values of Ce⁺/CeO⁺ would vary significantly with different experimental conditions, including sample-loading techniques, filament material and assembly, temperature of filaments etc. To optimize the conditions for Ce isotope ratio measurement, the ionization behaviors of Ce with the FPIE were investigated. Approximately 1 μg of Ce was loaded on FPIE filaments in 2.5 M HCl together with 1 μL of 0.2 M TaF₅. The Faraday cup configuration adopted is shown in Line 3 of Table 3, with all Faraday cup collectors coupled to 10¹¹ Ω resistors, and the integration time was set to 4.914 s. ¹⁴²Ce⁺ and ¹⁴²Ce¹⁶O⁺ ion beams were collected simultaneously by cups L4 and H4, respectively. Increasing damage to the Faraday cups, which can affect the cup efficiency and response linearity, should be taken into account for the routine measurement at high ion beam intensities. Thus the ¹⁴⁰Ce⁺ and ¹⁴⁰Ce¹⁶O⁺ ion beams, which might overload the corresponding amplifier, were not collected in the measurement.

The filament current was raised at a constant rate of 60 mA min⁻¹, and the data acquisition procedure was carried out until the signals of both ¹⁴²Ce⁺ and ¹⁴²Ce¹⁶O⁺ ions dropped below 10 mV. The ¹⁴²Ce⁺ and ¹⁴²Ce¹⁶O⁺ ion beams increased with the filament temperature at the beginning, and their profiles are shown in Fig. 2a. The peaks of ¹⁴²Ce¹⁶O⁺ (4.12 V) and ¹⁴²Ce⁺ (19.72 V) ion signals appeared at approximately 1050 °C and 1460 °C, respectively. It can be recognized that the integrated intensity of Ce⁺ was much higher than that of CeO⁺. Similar trends were observed for the samples containing 2 μg and 0.1 μg Ce as well. In contrast, the phenomenon of the traditional double Re filaments was quite different (Fig. 2b). The peaks of ¹⁴²Ce¹⁶O⁺ (10.17 V) and ¹⁴²Ce⁺ (8.19 V) ion signals located at ca. 1470 °C and 1620 °C, respectively. Notably, CeO⁺ ionizes more preferentially than Ce⁺ with traditional double Re filaments. The main reason for the differences was that plenty of CeO⁺ ions would be broken up by frequent collisions with the porous structure when they migrated through the FPIE material. It was confirmed by the fact that the ratios of Ce⁺/CeO⁺ increased with heating of the FPIE filament.

Fig. 2 Intensities of ¹⁴²Ce⁺ and ¹⁴²Ce¹⁶O⁺ as a function of the temperature: (a) FPIE filament; (b) traditional double Re filaments. Approximately 1 μg Ce was loaded on the filament, and the Faraday cup configuration is shown in Line 3 of Table 3.

Under the optimized conditions for Ce isotope ratio measurements, further experiments were performed with FPIE filaments loaded with 1 μg Ce. Steady ¹⁴²Ce⁺ ion intensities up to ca. 18 V (with a maximum fluctuation of ±5%) could last for over 2–3 h. The remarkable improvement in the stability of the Ce⁺ ion beam with FPIE was attributed to the combined effect of the porous structure of the FPIE filament. Firstly, the pervasion of the sample solution into the porous structure of the FPIE was achieved by gradually heating the FPIE filament. As a result, the free flow of the sample solution was prevented during TIMS analysis. Therefore, the optimal state of the ion lens was kept unaltered to extract a satisfactory ion beam. Moreover, due to the decrease of the sample surface area, a relatively slow evaporation rate of samples during the heating of the filament was ensured, and a long acquisition time could be obtained, accordingly.

3.2. Isobaric interference of Ba

The isobaric interference of ¹³⁸Ba on ¹³⁸Ce is the major challenge for accurate and high-precision Ce isotope measurements with the Ce⁺ technique. Contributions from interfering elements onto isotopes of interest can be corrected by monitoring another isotope of the interfering species and its known natural abundance. The Ba interference on mass 138 is normally corrected by monitoring ¹³⁷Ba, for the reason that ¹³⁷Ba is free of isobaric interferences and its abundance is relatively high. However, as Ba is typically more abundant than Ce in the environment,²⁶ and ¹³⁸Ba is the most abundant Ba isotope (71.7%), while the abundance of ¹³⁸Ce is only 0.25%, the presence of even a tiny amount of Ba contamination could result in inaccurate interference correction for ¹³⁸Ce. In this study, the ion exchange based chemical procedure enabled effective separation of Ba from Ce. Moreover, the FPIE filaments were heated at ∼1200 °C for ca. 40 minutes after sample loading for further eliminating possible residual Ba from the loaded sample. However, these two practices were inadequate for high-precision Ce isotope ratio measurement.

Another inherent source of Ba comes from the degassed filaments and materials used for the production of the FPIEs, and it is difficult to overcome the Ba interference even under the conditions of high temperature. We performed the measurements of Ba signals in newly-prepared FPIE filaments after being degassed. The FPIE filaments were prepared with different batches of single Re filaments (n = 5 each batch). ¹³⁸Ba⁺ ion beams were collected in the central Faraday cup, which was connected to a 10¹³ Ω amplifier. Small signals (0.5–0.9 mV) of ¹³⁸Ba⁺ were observed in degassed FPIE filaments (Table 4), and they were a little higher than those reported by Willbold,¹⁸ the difference may be attributed to different batches of the Re filaments. Such an amount of ¹³⁸Ba could not be negligible for high-precision measurement of Ce isotope ratios.

Table 4 Average values of measured ¹³⁸Ba ion beams for different batches of FPIE filaments before (¹³⁸Ba) and after (¹³⁸Ba*) being loaded with TaF₅. Five single Re filaments were included in each batch

Manufacturer	Batch	^138Ba (μV)	^138Ba* (μV)	Improvement factor
Thermo Fisher Scientific	001-0509444	742	68	10.9
Thermo Fisher Scientific	003-0702444	695	87	8.0
Thermo Fisher Scientific	004-2002444	838	65	12.9
Thermo Fisher Scientific	006-2804444	504	52	9.7

---

In order to further reduce the influence of Ba, we employed TaF₅ as an activator to consume Ba by forming BaF⁺ ions, and the BaF⁺ peaks at masses 153, 154, 155, 156 and 157 were detected during the preliminary heating procedure (Fig. 3). Since fluorine is a monoisotopic element, the intensities of BaF⁺ ion beams are proportional to the abundances of Ba isotopes. There was an exception at mass 156, which suffered polyatomic interference of ¹⁴⁰Ce¹⁶O⁺ on ¹³⁷Ba¹⁹F⁺. The appearance of the peak on mass 158, which consisted of ¹⁴⁰Ce¹⁸O⁺ and ¹⁴²Ce¹⁶O⁺, further confirmed the ¹⁴⁰Ce¹⁶O⁺ interference on mass 156. The signal at mass 159 was also measured to detect the possible ¹⁴⁰Ce¹⁹F⁺ ion beam and no signals were observed. Therefore, the formation of Ce fluoride during the measurement was negligible. The intensity of BaF⁺ ion beams increased rapidly with increasing FPIE filament temperature, and the Ba⁺ peaks tended to diminish in the meantime. Blanks of Ba in FPIE filaments were measured again after being loaded with 1 μL of 0.2 M TaF₅, and the results showed that ¹³⁸Ba⁺ ion beams were significantly reduced by a factor of ca. 10 (Table 4).

Fig. 3 The formation of BaF⁺ ions at masses 153, 154, 155, 156 and 157 with FPIE filament (T = 1200 °C).

3.3. Selecting the proper reference Ce isotope

The variation of radiogenic isotope ¹³⁸Ce in various natural samples is expressed in terms of isotope ratios (the ratios of the ¹³⁸Ce isotope to a stable Ce isotope). The selection of the reference Ce isotope, therefore, is very important. As shown in Table S1,† the most commonly-used reference isotope was ¹⁴²Ce. Recently, ¹³⁶Ce was also proposed to be a new reference isotope.¹⁸ However, the accurate determination of ¹⁴²Ce and ¹³⁶Ce suffers from isobaric interferences from ¹⁴²Nd and ¹³⁶Ba, respectively. Moreover, measurements of these two Ce isotopes are both affected by the peak tailing of the major isotope ¹⁴⁰Ce. All these factors further complicate Ce isotope ratio measurements with high-precision. Compared to ¹⁴²Ce and ¹³⁶Ce, ¹⁴⁰Ce seems to be a possible alternative choice as there is no isobaric interference on ¹⁴⁰Ce and the complicated peak tailing correction can be avoided.

A number of publications involving only a specific Ce isotope ratio (¹³⁸Ce/¹⁴²Ce or ¹³⁸Ce/¹³⁶Ce) are available, but no comprehensive comparison on different Ce isotope ratios has been found to date. Therefore, we investigated different Ce isotope ratios of reference rock sample BCR-2 with FPIEs in Ce⁺ mode. The measurements were performed on large sample loads, i.e. FPIE filaments loaded with 5 μg Ce, and the external precisions (2RSD) of different Ce isotope ratios at various ¹³⁸Ce beam intensities are illustrated in Fig. 4. All the isotope ratios were corrected with isobaric interferences, and the ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios were tail-corrected as well. It is obvious that the external precisions of all these Ce isotope ratios improved with the increasing of ion beam intensities. The external precisions were similar for ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios, and they were comparable to those determined in Ce⁺ mode at similar beam sizes.^20,31–33 The external precision of the ¹³⁸Ce/¹⁴⁰Ce ratio showed a remarkable improvement, and it was a factor of 1.2–1.8 better compared to those of ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios at the same ¹³⁸Ce beam intensities. Therefore, the ¹³⁸Ce/¹⁴⁰Ce ratio was regarded as a promising choice to monitor the anomalies of ¹³⁸Ce in various geological materials.

Fig. 4 External precisions (2RSD) of different Ce isotope ratios as a function of ¹³⁸Ce⁺ beam intensity. The measurements were performed in Ce⁺ mode with FPIE filaments loaded with 5 μg Ce.

The main reason for a better reproducibility achieved by the new target Ce isotope ratio is the avoiding of the isobaric interference correction and the peak-tailing effect. The less data correction work, the better precision of the data would be obtained. To our knowledge, the ¹³⁸Ce/¹⁴⁰Ce ratios for JMC-304 and the chondritic uniform reservoir (CHUR) were not certified, thus we obtained them using the certified ¹³⁸Ce/¹⁴²Ce values and a constant ¹⁴⁰Ce/¹⁴²Ce value of 7.941.^8,24,25 With (¹³⁸Ce/¹⁴²Ce)_JMC-304 = 0.0225762–0.0225799 (ref. 25, 29, 38 and 52–56) and (¹³⁸Ce/¹⁴²Ce)_CHUR = 0.0225652,^16,19,26,57 the ¹³⁸Ce/¹⁴⁰Ce ratio values were calculated to be 0.002842992–0.002843458 for JMC-304 and 0.002841607 for CHUR, respectively.

Nearly all Ce isotopic data in the literature have been reported in terms of ¹³⁸Ce/¹⁴²Ce or ¹³⁸Ce/¹³⁶Ce, thus it is inconvenient for a direct comparison. Therefore, Ce isotope ratios were translated to the ε notation in our discussion. It is defined as

where ⁱCe refers to different Ce reference isotopes (¹³⁶Ce, ¹⁴⁰Ce or ¹⁴²Ce), and (¹³⁸Ce/ⁱCe)_CHUR refers to the ¹³⁸Ce/ⁱCe ratio for CHUR at time T (T = 0 for present). Take the ratios of ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹⁴⁰Ce as an example, although their absolute values vary greatly, there is no distinction between ε(¹³⁸Ce/¹⁴⁰Ce) and ε(¹³⁸Ce/¹⁴²Ce) for the constant conversion factor of ¹⁴⁰Ce/¹⁴²Ce = 7.941.

3.4. Using diverse ohm amplifiers for the measurement of the ¹³⁸Ce/¹⁴⁰Ce ratio

For high-precision measurement of the ¹³⁸Ce/¹⁴⁰Ce ratio, ¹³⁸Ce and ¹⁴⁰Ce ion beams should be collected simultaneously. The ¹⁴⁰Ce isotope is much more abundant compared to the ¹³⁸Ce isotope (88.45 at% versus 0.25 at%), thus measuring ion beams over a large dynamic range is needed. This issue is partly overcome on new generation mass spectrometers with the development of amplifiers with diverse ohm resistances (10¹⁰ Ω, 10¹² Ω and 10¹³ Ω). The use of high ohm resistor amplifiers (10¹² Ω or 10¹³ Ω) is an option for the measurement of low beam intensities with high precision, and they can be used alone, or in combination with conventional 10¹⁰ Ω and/or 10¹¹ Ω amplifiers.^{45–50,58–60} While the 10¹² Ω resistors provide a 10 times higher gain of the amplifiers, the Johnson noise of the resistor only increases by , thus resulting in an approximately 3-fold theoretical improvement in the signal-to-noise ratio. Similarly, the theoretical improvement obtained in the signal-to-noise ratio is 10 times for 10¹³ Ω amplifiers. The gain calibration for each resistor was performed prior to the mass spectrometric measurements each day. Gain calibration factors for the 10¹⁰ Ω, 10¹¹ Ω and 10¹² Ω resistors were determined using the regular software controlled gain procedure, which uses a stable reference current (3.33333 V). However, this procedure is not implemented for 10¹³ Ω resistors, because this used reference current will overload the 10¹³ Ω resistors. Instead, gain calibration factors for the 10¹³ Ω resistors were determined with the calibration protocol described by Kimura et al.,⁶¹ and then manually updated in the system table. Baselines were measured in a 15 min routine (900 cycles of 1s integration) before and after the Ce isotopic measurements. The gain corrected baseline noise values for 10¹¹ Ω, 10¹² Ω and 10¹³ Ω amplifiers were ∼9.7 μV, ∼2.2 μV and ∼0.8 μV (1SD), respectively.

For comparison, Ce isotope ratio analysis with FPIEs was performed using different combinations of amplifiers (Table 5) in Ce⁺ mode. All analyses consisted of 176 cycles of 8.389 s integration each, and 2 μg Ce standard (JMC-304) was loaded and measured. The ¹⁴⁰Ce⁺ intensities were kept at about 280 V. For convenient comparison, gain corrected intensities relative to the default 10¹¹ Ω amplifiers were reported in this paper.

Table 5 Different combinations of amplifiers in the Faraday cup configuration of multi-collector mode

Faraday cup	L4	L3	L2	L1	C	H1	H2
Ions required	^136Ce	^137Ba	^138Ce	^139La	^140Ce	^142Ce	^143Nd
Amp 1	10^11 Ω	10^11 Ω	10^11 Ω	10^11 Ω	10^10 Ω	10^11 Ω	10^11 Ω
Amp 2	10^11 Ω	10^11 Ω	10^12 Ω	10^11 Ω	10^10 Ω	10^11 Ω	10^11 Ω
Amp 3	10^11 Ω	10^12 Ω	10^11 Ω	10^11 Ω	10^10 Ω	10^11 Ω	10^11 Ω
Amp 4	10^11 Ω	10^13 Ω	10^11 Ω	10^11 Ω	10^10 Ω	10^11 Ω	10^11 Ω
Amp 5	10^11 Ω	10^13 Ω	10^12 Ω	10^11 Ω	10^10 Ω	10^11 Ω	10^11 Ω

---

The measured ¹³⁸Ce/¹⁴⁰Ce ratios for 2 μg JMC-304 using different combinations of amplifiers with the FPIE method in Ce⁺ mode are illustrated in Fig. 5, wherein the error bars represent the internal precision (2SE) of an individual sample. Firstly, a combination of 10¹⁰ Ω and 10¹¹ Ω amplifiers was used where the ¹⁴⁰Ce⁺ beam was measured on a 10¹⁰ Ω amplifier and the other isotopes on 10¹¹ Ω amplifiers. The external precision (2RSD) of ¹³⁸Ce/¹⁴⁰Ce ratios obtained was 312 ppm, and it was improved to 281 ppm and 178 ppm with 10¹² Ω amplifiers connecting to the Faraday cups L2 and L3, respectively. These improvements are mainly explained by the high signal-to-noise ratio of high ohm amplifiers. Compared to ¹³⁸Ce, the beam intensity of the interference monitor ¹³⁷Ba was extremely low (∼10 μV), therefore the application of a high ohm amplifier to collect ¹³⁷Ba resulted in a better analytical precision. In spite of the improvements, none of them could fulfill the requirements for Ce isotope ratio measurement in nuclear forensic research. Furthermore, a 10¹³ Ω amplifier was employed to collect the ¹³⁷Ba ion beam, meanwhile 10¹¹ Ω and 10¹² Ω amplifiers were employed on ¹³⁸Ce, successively. The weighted average ¹³⁸Ce/¹⁴⁰Ce ratios were 0.00284305 ± 14 (2SD, n = 10) and 0.00284296 ± 10 (2SD, n = 10) for the 10¹³–10¹¹–10¹⁰ Ω and 10¹³–10¹²–10¹⁰ Ω amplifier combinations, respectively. The consistency of the experimental data with the reference value of JMC-304 (0.002842992–0.002843458) demonstrated the accuracy of the newly-developed method. Moreover, it suggested that the contribution of the operational blank for Ce isotope analysis is negligible. The obtained reproducibility of ¹³⁸Ce/¹⁴⁰Ce ratios for the 10¹³–10¹²–10¹⁰ Ω combination was 36 ppm, showing a factor of ca. 10 improvement compared to that obtained using solely 10¹¹ Ω amplifiers on ¹³⁷Ba and ¹³⁸Ce, and it was even comparable with those obtained using the CeO⁺ technique for similar Ce loading levels.^{12,16–18,20–28} The reproducibility of ¹³⁸Ce/¹⁴⁰Ce ratios using the 10¹³–10¹¹–10¹⁰ Ω combination (2RSD of 48 ppm, n = 10) was slightly worse compared to that obtained using the 10¹³–10¹²–10¹⁰ Ω combination. However, they were both adequate for resolving variations of ¹³⁸Ce/¹⁴⁰Ce ratios within the eighth decimal place for geological samples.

Fig. 5 ¹³⁸Ce/¹⁴⁰Ce ratios measured for 2 μg JMC-304 using different combinations of amplifiers with the FPIE method as Ce⁺. The error bars in each measurement are given in 2SE. Amp 1–5 refer to different combinations of amplifiers in the Faraday cup configuration listed in Table 5.

With the employment of TaF₅ as an activator, the ¹³⁷Ba⁺ ion beam was consequently suppressed to levels of ∼10 μV, corresponding to ¹³⁸Ba/¹³⁸Ce ≈ 0.00008 as the ¹³⁸Ce⁺ ion beam could reach ca. 80 mV on 2 μg loads with FPIEs. The baseline noise on the 10¹³ Ω feedback resistors was measured to be ca. 0.8 μV (1 SD), which introduced ca. 8% bias for the ¹³⁷Ba and ¹³⁸Ba signals. As the ¹³⁸Ba/¹³⁸Ce ratio was about 0.00008, the bias introduced for this ratio from the correction of ¹³⁸Ba was ca. 7 ppm, which was within the external precision (2RSD = 36 ppm) of ¹³⁸Ce/¹⁴⁰Ce ratios using the 10¹³–10¹²–10¹⁰ Ω combination of current amplifiers. Therefore, with the application of a 10¹³ Ω resistor for the ¹³⁷Ba signal detection, accurate interference correction of ¹³⁸Ba on ¹³⁸Ce was possible, and high-precision Ce isotope ratios with the Ce⁺ technique could be obtained.

Idle time defines the waiting time prior to measuring the new line in the cup configuration table during dynamic measurements, and the majority of the idle time is needed for decay of the current amplifier. Compared to conventional 10¹¹ Ω amplifiers, the idle time of high ohm amplifiers is much longer for their slower response time, accordingly, a longer data acquisition time is required. Therefore, the use of high ohm amplifiers requires a more stable ion beam to obtain high-precision Ce isotope ratios. Typically, the ¹⁴⁰Ce⁺ ion beam could reach 280 V after being fully focused on 2 μg loads with FPIEs, and it could last for more than 200 minutes, which was superior to that when loading with conventional Re filaments. The excellent stability of Ce⁺ ion beams obtained with FPIEs supported our method to be applied in high-precision measurements of Ce isotope ratios.

Recently, Bonnand et al.³⁶ determined Ce isotope ratios using diverse ohm amplifiers with the CeO⁺ technique, and they measured the ¹³⁶Ce¹⁶O⁺, ¹³⁸Ce¹⁶O⁺ and ¹⁴²Ce¹⁶O⁺ ions together with the interference oxide ions. As the number of interferences collected was limited, the less abundant ¹³⁴Ba (2.4%) was used instead of ¹³⁷Ba (11.2%) as a Ba interference monitor. In our work, the combination of diverse ohm amplifiers was applied for Ce measurement as Ce⁺ ions. By using the high ohm amplifier, the small signal of ¹³⁷Ba (which has no isobaric interference) could be directly detected, thus the interference of a trace amount of ¹³⁸Ba could be corrected more precisely.

3.5. Ce isotope analysis of reference rock samples and uranium ore samples

Replicate analysis was done for five reference rock samples with different sample matrices to evaluate the reproducibility of the newly-developed TIMS method. The amount of Ce introduced into the mass spectrometer for a single analysis was typically 2 μg Ce, which was sufficient to obtain a large and stable Ce⁺ ion beam. The results are shown in Table 6 and Fig. 6, and the bars below the εCe values of reference rock samples represent the reported εCe values in Table 6. The external precisions of ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios were comparable to those previously reported in CeO⁺ mode (Table S1†). By using our method, the external precision of the ¹³⁸Ce/¹⁴⁰Ce ratio was slightly better than those of ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios. Moreover, the εCe values obtained for most reference rock samples were in good agreement with literature values, indicating high reliability of our method in geological samples.

Table 6 Analytical results of Ce isotope ratios of reference rock samples

Sample	^138Ce/^142Ce	^138Ce/^136Ce	^138Ce/^140Ce	ε(^138Ce/^140Ce)CHUR	Reported ε(^138Ce/^142Ce)CHUR
BCR-2	0.0225739	1.33729	0.00284173	+0.43
	0.0225714	1.33718	0.00284166	+0.19
	0.0225711	1.33721	0.00284162	+0.05
	0.0225728	1.33742	0.00284177	+0.57
	0.0225745	1.33728	0.00284185	+0.85
	0.0225716	1.33734	0.00284171	+0.36
Mean ± 2SD	0.0225726 ± 14	1.33729 ± 9	0.00284172 ± 16	+0.41 ± 0.56	0.08–0.79 (ref. 12, 19, 27 and 28)
JA-2	0.0225701	1.33704	0.00284147	−0.48
	0.0225710	1.33717	0.00284153	−0.27
	0.0225715	1.33711	0.00284158	−0.09
	0.0225691	1.33709	0.00284141	−0.69
	0.0225727	1.33727	0.00284150	−0.37
	0.0225719	1.33718	0.00284156	−0.16
Mean ± 2SD	0.0225711 ± 13	1.33714 ± 8	0.00284151 ± 12	−0.34 ± 0.42	−0.19 to −0.30 (ref. 19)
JB-3	0.0225673	1.33694	0.00284103	−2.03
	0.0225684	1.33697	0.00284117	−1.54
	0.0225696	1.33702	0.00284125	−1.26
	0.0225668	1.33696	0.00284109	−1.82
	0.0225689	1.33699	0.00284122	−1.36
	0.0225677	1.33712	0.00284119	−1.47
Mean ± 2SD	0.0225681 ± 11	1.33700 ± 6	0.00284116 ± 9	−1.58 ± 0.32	−1.22 to −1.96 (ref. 2, 16, 19 and 26)
JG-1	0.0225741	1.33731	0.00284202	+1.45
	0.0225749	1.33737	0.00284207	+1.63
	0.0225762	1.33746	0.00284199	+1.35
	0.0225737	1.33725	0.00284186	+0.89
	0.0225721	1.33727	0.00284179	+0.64
	0.0225744	1.33736	0.00284195	+1.20
Mean ± 2SD	0.0225742 ± 14	1.33734 ± 8	0.00284195 ± 13	+1.20 ± 0.46	0.05–0.52 (ref. 19)
JR-1	0.0225698	1.33711	0.00284141	−0.69
	0.0225692	1.33705	0.00284125	−1.26
	0.0225674	1.33698	0.00284115	−1.61
	0.0225721	1.33704	0.00284130	−1.08
	0.0225683	1.33682	0.00284118	−1.50
	0.0225706	1.33695	0.00284122	−1.36
Mean ± 2SD	0.0225696 ± 17	1.33699 ± 10	0.00284125 ± 19	−1.25 ± 0.67	−0.6 to −1.47 (ref. 2 and 23)

---

Fig. 6 The εCe values of reference rock samples and uranium ore samples. The bars below the εCe values of reference rock samples represent the reported εCe values in Table 6, and the vertical line represents εCe = 0.

In addition to multiple analysis of reference rock samples, further Ce isotope measurements were performed for eight uranium ore samples (1#–8#) to test the developed method for nuclear forensics purposes. The results are depicted in Fig. 6 and presented in Table 7 and Table S2.† Clear differences, ranging from −3 to +3, in the εCe values of the uranium ores can be observed. The εCe value of the 4# uranium ore sample was considered to overlap within error with those of 5# and 6# samples originating from the same uranium mine from Congo, while a similar phenomenon was observed for 1# and 2# uranium ore samples mined from the USA. Meanwhile, the εCe values of 3#, 7# and 8# uranium ore samples originating from different countries shared no overlap with each other. These results indicated the Ce isotopic composition might act as an informative clue for assessing the origin of uranium ores and it would be helpful to establish a worldwide database of Ce isotopic composition with high precision to serve nuclear forensics purposes.

Table 7 Analytical results of Ce isotopic ratios of uranium ore samples^a

Sample	Ce concentration (ppm)	ε(^138Ce/^140Ce)CHUR	Mean ± 2SD
a Ce concentrations of uranium ore samples were measured by ICP-MS (Agilent 7700x).
1# UO	964	−1.36, −1.54, −1.26, −1.19, −1.43, −1.36	−1.36 ± 0.25
2# UO	72.9	−1.57, −2.03, −2.42, −1.64, −1.92, −2.21	−1.96 ± 0.67
3# UO	19.1	+0.43, +0.33, +0.71, +0.36, +0.29, +0.22	+0.39 ± 0.35
4# UO	2.04	−1.54, −1.89, −1.26, −1.82, −1.64, −1.19	−1.55 ± 0.56
5# UO	141	−0.45, −0.83, −0.90, −0.38, −0.69, −0.90	−0.69 ± 0.46
6# UO	77.6	−0.76, −1.04, −0.83, −0.90, −1.365, −1.22	−1.02 ± 0.46
7# UO	127	+1.95, +2.16, +1.42, +1.24, +2.09, +1.59	+1.74 ± 0.74
8# UO	92.8	−1.01, −1.47, −0.76, −0.38, −1.22, −0.69	−0.92 ± 0.77

---

However, we also observed that there were overlaps of the εCe values of some uranium ore samples originating from different countries. Consequently, the Ce isotopic composition should not be regarded as an exclusive proof in origin assessment, instead, it was supposed to perform better by the combination with other characteristic information (such as Nd isotopic composition, REE concentrations etc.) to reliably trace the origin of unknown nuclear materials or significantly narrow down to a few possibilities.

4. Conclusions

A novel method has been developed and validated for the measurement of Ce isotope ratios as Ce⁺ ions in geological and uranium ore samples by TIMS. Ce was oxidized and effectively separated from interfering matrix components with an oxidation separation procedure. The newly-developed FPIE was used for TIMS measurement and Ce was observed to ionize more efficiently as Ce⁺ ions than CeO⁺ ions compared to the methods using traditional double Re filaments. The isobaric interference of Ba was effectively suppressed by one order of magnitude with the employment of TaF₅ as an activator, and it further ensured the high-precision Ce isotope ratio analysis with the Ce⁺ technique. ¹⁴⁰Ce was found to be a promising reference isotope to monitor the anomalies of ¹³⁸Ce in geological materials. The external precision of the ¹³⁸Ce/¹⁴⁰Ce ratio showed a remarkable improvement, and it was a factor of 1.2–1.8 better compared to those of ¹³⁸Ce/¹⁴²Ce and ¹³⁸Ce/¹³⁶Ce ratios at the same beam intensities. For the analysis of Ce isotopic ratios as Ce⁺ ions, the performances of diverse ohm current amplifiers were investigated and an optimized combination of current amplifiers was suggested. With the 10¹³–10¹²–10¹⁰ Ω amplifier combination, the obtained weighted ¹³⁸Ce/¹⁴⁰Ce ratio for a reference material (JMC-304) containing 2 μg Ce was 0.00284296 ± 10 (2SD, n = 10). The reproducibility of the ¹³⁸Ce/¹⁴⁰Ce isotope ratio was a factor of ca. 10 better compared to that obtained using the previously published Ce⁺ techniques. Furthermore, the εCe values of the reference rock samples of this study agreed well with those previously reported, indicating the reliability of the newly-developed method. This method was used for the determination of the ¹³⁸Ce/¹⁴⁰Ce ratios in uranium ores of world-wide origin. The results showed significant variations between different samples, and the obtained analytical precision successfully met the requirements for sample source identification.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 11805172 and 11605172) and partly supported by the Science Challenge Project (No. TZ2016004).

References

1. S. Nakai, H. Shimizu and A. Masuda, Nature, 1986, 320, 433–435.
2. T. Tanaka, H. Shimizu, Y. Kawata and A. Masuda, Nature, 1987, 327, 113–117.
3. A. P. Dickin, Nature, 1988, 333, 403–404.
4. N. Bellot, M. Boyet, R. Doucelance, P. Bonnand, I. P. Savov, T. Plank and T. Elliott, Chem. Geol., 2018, 500, 46–63.
5. M. Tanimizu and T. Tanaka, Geochim. Cosmochim. Acta, 2002, 66, 4007–4014.
6. R. Nakada, Y. Takahashi and M. Tanimizu, Geochim. Cosmochim. Acta, 2016, 181, 89–100.
7. H. J. W. De Baar, M. P. Bacon and P. G. Brewer, Nature, 1983, 301, 324–327.
8. M. Tanimizu, Phys. Rev. C: Nucl. Phys., 2000, 62, 017601.
9. H. Shimizu, M. Amano and A. Masuda, Geology, 1991, 19, 369–371.
10. A. Bouvier, J. D. Vervoort and P. J. Patchett, Earth Planet. Sci. Lett., 2008, 273, 48–57.
11. H. Tazoe, H. Obata and T. Gamo, Geochem., Geophys., Geosyst., 2011, 12, 04004.
12. R. Doucelance, N. Bellot, M. Boyet, T. Hammouda and C. Bosq, Earth Planet. Sci. Lett., 2014, 407, 175–186.
13. E. Keegan, S. Richter, I. Kelly, H. Wong, P. Gadd, H. Kuehn and A. Alonso-Munoz, Appl. Geochem., 2008, 23, 765–777.
14. S. H. Han, Z. Varga, J. Krajkó, M. Wallenius, K. Song and K. Mayer, J. Anal. At. Spectrom., 2013, 28, 1919–1925.
15. Z. Varga, M. Wallenius and K. Mayer, Radiochim. Acta, 2010, 98, 771–778.
16. H. Tazoe, H. Obata, H. Amakawa, Y. Nozaki and T. Gamo, Mar. Chem., 2007, 103, 1–14.
17. A. Makishima and E. Nakamura, Chem. Geol., 1991, 94, 1–11.
18. M. Willbold, J. Anal. At. Spectrom., 2007, 22, 1364–1372.
19. C. Schnabel, C. Münker and E. Strub, J. Anal. At. Spectrom., 2017, 32, 2360–2370.
20. T. Tanaka and A. Masuda, Nature, 1982, 300, 515–518.
21. H. Shimizu, T. Tanaka and A. Masuda, Nature, 1984, 307, 251–252.
22. N. Nakamura, M. Tatsumoto and K. R. Ludwig, J. Geophys. Res., 1984, 89, 438–444.
23. H. Amakawa, Y. Nozaki and A. Masuda, Chem. Geol., 1996, 131, 183–195.
24. M. Tanimizu, T. Hayashi and T. Tanaka, J. Mass Spectrom. Soc. Jpn., 2004, 52, 177–181.
25. T. Hayashi, M. Tanimizu and T. Tanaka, Precambrian Res., 2004, 135, 345–357.
26. H. Tazoe, H. Obata and T. Gamo, J. Anal. At. Spectrom., 2007, 22, 616–622.
27. N. Bellot, M. Boyet, R. Doucelance, C. Pin, C. Chauvel and D. Auclair, Geochim. Cosmochim. Acta, 2015, 168, 261–279.
28. M. Willig and A. Stracke, Chem. Geol., 2018, 476, 119–129.
29. H. Shimizu, H. Sawatari, Y. Kawata, P. N. Dunkley and A. Masuda, Contrib. Mineral. Petrol., 1992, 110, 242–252.
30. Y. K. Xiao, W. G. Liu and Y. M. Zhou, Int. J. Mass Spectrom. Ion Processes, 1994, 136, 181–189.
31. T. L. Chang, Q. Y. Qian, M. T. Zhao, J. Wang and Q. Y. Lang, Int. J. Mass Spectrom. Ion Processes, 1995, 142, 125–131.
32. A. Makishima and A. Masuda, Chem. Geol., 1994, 118, 1–8.
33. M. Rehkämper, M. Gärtner and S. L. Goldstein, Chem. Geol., 1996, 129, 201–208.
34. C. F. Li, X. C. Wang, Y. L. Li, Z. Y. Chu, J. H. Guo and X. H. Li, J. Anal. At. Spectrom., 2015, 30, 895–902.
35. S. Kagami and T. Yokoyama, Anal. Chim. Acta, 2016, 937, 151–159.
36. P. Bonnand, C. Israel, M. Boyet, R. Doucelance and D. Auclair, J. Anal. At. Spectrom., 2019, 34, 504–516.
37. C. F. Li, F. K. Chen and X. H. Li, Int. J. Mass Spectrom., 2007, 266, 34–41.
38. A. Masuda, H. Shimizu, S. Nakai, A. Makishima and S. Lahti, Earth Planet. Sci. Lett., 1988, 89, 316–322.
39. M. G. Watrous, J. E. Delmore and M. L. Stone, Int. J. Mass Spectrom., 2010, 296, 21–24.
40. M. G. Watrous and J. E. Delmore, Int. J. Mass Spectrom., 2011, 303, 1–5.
41. M. L. Baruzzini, H. L. Hall, K. J. Spencer and F. E. Stanley, Int. J. Mass Spectrom., 2018, 430, 57–62.
42. M. L. Baruzzini, H. L. Hall, M. G. Watrous, K. J. Spencer and F. E. Stanley, Int. J. Mass Spectrom., 2017, 412, 8–13.
43. F. E. Stanley, K. J. Spencer, D. S. Schwartz, M. G. Watrous and J. E. Delmore, J. Radioanal. Nucl. Chem., 2014, 299, 1447–1452.
44. X. P. Shao, W. T. Bu, K. M. Long, L. Tang, X. M. Liu, H. Yan and F. H Hao, Spectrochim. Acta, Part B, 2019, 159, 105656.
45. J. M. Koornneef, C. Bouman, J. B. Schwieters and G. R. Davies, J. Anal. At. Spectrom., 2013, 28, 749–754.
46. J. M. Koornneef, C. Bouman, J. B. Schwieters and G. R. Davies, Anal. Chim. Acta, 2014, 819, 49–55.
47. A. Trinquier and P. Komander, J. Radioanal. Nucl. Chem., 2016, 307, 1927–1932.
48. A. V. Quadt, J. F. Wotzlaw, Y. Buret, S. J. E. Large, I. Peytcheva and A. Trinquier, J. Anal. At. Spectrom., 2016, 31, 658–665.
49. M. Pfeifer, N. S. Lloyd, S. T. M. Peters, F. Wombacher, B. M. Elfers, T. Schulze and C. Münker, J. Anal. At. Spectrom., 2017, 32, 130–143.
50. N. S. Lloyd, A. Y. Sadekov and S. Misra, Rapid Commun. Mass Spectrom., 2018, 32, 9–18.
51. A. Makishima, H. Shimizu and A. Masuda, J. Mass Spectrom. Soc. Jpn., 1987, 35, 64–72.
52. H. Shimizu, S. Nakai, S. Tasaki, A. Masuda, D. Bridgewater, A. P. Nutman and H. Baadsgaard, Earth Planet. Sci. Lett., 1988, 91, 159–169.
53. H. Shimizu, N. Umemoto, A. Masuda and P. W. U. Appel, Geochim. Cosmochim. Acta, 1990, 54, 1147–1154.
54. H. Amakawa, J. Ingri, A. Masuda and H. Shimizu, Earth Planet. Sci. Lett., 1991, 105, 554–565.
55. H. Shimizu, S.-G. Lee, A. Masuda and M. Adachi, Geochem. J., 1996, 30, 57–69.
56. S.-G. Lee, A. Masuda, H. Shimizu and Y.-S. Song, Geochem. J., 2001, 35, 175–187.
57. A. Makishima and A. Masuda, Chem. Geol., 1993, 106, 197–205.
58. D. Wielandt and M. Bizzarro, J. Anal. At. Spectrom., 2011, 26, 366–377.
59. A. Makishima and E. Nakamura, J. Anal. At. Spectrom., 2012, 27, 891–895.
60. T. Schulz, C. Muenker and S. T. M. Peters, Earth Planet. Sci. Lett., 2013, 362, 246–257.
61. J. I. Kimura, Q. Chang, N. Kanazawa, S. Sasaki and B. S. Vaglarov, J. Anal. At. Spectrom., 2016, 31, 790–800.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ja00420c

---