Takafumi
Hirata
*a,
Sota
Niki
a,
Shuji
Yamashita
a,
Hisashi
Asanuma
a and
Hideki
Iwano
b
aGeochemical Research Center, The University of Tokyo, Hongo 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail: hrt1@eqchem.s.u-tokyo.ac.jp
bKyoto Fission-Track Co. Ltd, Minamitajiri-cho 44-4, Omiya, Kita-ku, Kyoto 603-8832, Japan
First published on 7th October 2020
206Pb/238U isotopic ratios were measured from transient signals produced by laser ablation for a short time period (4 shots with 60 kHz repetition rates). To investigate the effect of the dwell time on the measured ratios, the signal intensities of 206Pb and 238U were obtained with various dwell times (200, 20, 2, 0.2, 0.1, 0.04, and 0.02 ms). The signal intensities of 206Pb and 238U increase with shortening the dwell time, and the resulting peak signal intensities of 238U for three zircons (Plešovice, GJ-1, and Nancy 91500) exceeded 106 cps with the shortest dwell time (0.02 ms). This is mainly due to moderation of the signal intensity of 238U through averaging the intensity data from neighboring time slices, suggesting that the acquired signal intensity profile does not reflect the actual signal intensities. Based on the moderated signal intensity data, correction of the counting loss due to the detector dead time can be erroneous, and this can be well demonstrated by the changes of the measured 206Pb/238U ratios with different dwell times. With the correction of the counting loss based on the obtained data and a short dwell time (0.02 ms), the resulting 206Pb/238U ratios for zircon and glass standard materials were in good agreement with the literature within the analytical uncertainties achieved in this study (2–3%). The data obtained here demonstrate clearly that the acquisition with a short dwell time is essential to obtain reliable elemental/isotopic data from transient signals emanating from single-shot laser ablation and introduction of nanoparticles.
With single shot laser ablation or laser ablation with short time duration protocols, the resulting signal intensities of 206Pb and 238U isotopes become high (e.g., >106 cps), and thus, the correction of the counting loss due to the detector dead time is very important to obtain reliable isotopic ratio data.9,10 The major problem associated with the transient signals found in single-shot LA or LA with short time durations is the moderation of the peak intensities through averaging the signal intensity data acquired in neighbouring time slices. With the conventional dwell time values (e.g., 2–200 ms), peak signal intensities can be moderated through averaging the signal intensity data of neighbouring time slices. This suggests that, with the conventional dwell time, the measured signal intensity may not reflect the true signal intensity data, suggesting that the magnitude of the counting loss could be underestimated.
In this study, to investigate the effect of the dwell time on the measured signal intensity data, the 206Pb/238U ratios have been measured from transient signals emanating from the laser ablation of short time durations (0.067 ms). The 206Pb/238U ratios were measured from three zircons (Plešovice, Nancy 91500, and GJ-1) with various dwell times (200, 20, 2, 0.2, 0.1, 0,04 and 0.02 ms). Based on the results obtained here, we demonstrate that it is possible to obtain reliable isotopic ratio data even from the transient signals generated by the laser ablation.
An in-house laser ablation system using a Yb:KGW femtosecond laser (Carbide, Light Conversion, Lithuania) was adopted in this study. The ablation pit size used in this study was about 10 μm. Since the laser ablation was conducted with the identical laser parameters (i.e., pit size, fluence, repetition rate, and the number of laser shots) and system setup (i.e., cell geometry, gas flow rates, length and internal diameter of sample transport tubing, and mixing positions of Ar make-up gas), changes in the magnitude of elemental fractionation on the Pb/U ratio among the zircon grains can be neglected. This suggests that the changes in the measured 206Pb/238U ratios obtained with different dwell times reflect the contribution of the counting loss of ion signals.
The 206Pb and 238U signals were obtained through the laser ablation of three natural zircons (Plešovice, GJ-1, and Nancy 91500) and a NIST SRM610 glass standard, using a time-resolved analysis protocol with various dwell times. The analysis sequence begins with analysis of the carrier gas without any laser ablation (the gas blank). This was followed by the five-time repeated analyses of 206Pb and 238U signals for three zircons and NIST SRM610 through laser ablation.
After the correction of the system backgrounds by subtracting the background ion counts from 206Pb and 238U data, the U–Pb signal intensity data were then used for the calculations of the 206Pb/238U ratio. The details of the instrumentation and operational settings were identical to those employed in our previous studies.9,10
Transient signals (count rates of both the 206Pb and 238U being >106 cps with a time duration of about 1 s) were produced by laser ablation with 4 shots of 60 kHz repetition rate (i.e., time duration of 0.067 ms). The 206Pb and 238U signals were acquired for 15 s with various dwell times (200, 20, 2, 0.2, 0.1, 0.04, and 0.02 ms). The background was corrected by subtracting the signal intensity data obtained without laser ablation (gas blank). The correction of the counting loss due to the detector dead time was made based on the non-extendable protocol.10 The resulting signal intensity data for 206Pb and 238U were used to calculate the 206Pb/238U ratio.
Fig. 1 Signal intensity profiles of 206Pb and 238U obtained with laser ablation for Plešovice zircon with a short time duration. |
After the correction of the counting loss due to the detector dead time, the signal intensity data for 206Pb and 238U were used to calculate the 206Pb/238U ratios. The 206Pb/238U ratios for two zircons (GJ-1 and Nancy 91500) and the glass standard increased about 4% with shortening the dwell time (Fig. 2). To evaluate the accuracy of the 206Pb/238U ratio data, both the mass bias effect and gain of the Daly detectors were calibrated by normalising the 206Pb/238U ratio for Plešovice which is 0.05372.12 Hence, the 206Pb/238U ratios for Plešovice were measured with dwell times of 200, 20, 2, 0.2, 0.1, 0.04, and 0.02 ms (Table 1).
Sample | U (ppm) | Dwell time (ms) | 206Pb/238Ua | |||||
---|---|---|---|---|---|---|---|---|
Measured | % 2SD | Bias factorb | Corrected | Refc | % RD | |||
a Counting loss due to the detector deadtime was corrected based on the non-extended type formula. b Bias factor was defined by normalising the 206Pb/238U ratio for Plešovice which is 0.05372 (Sláma et al., 2008).12 c Literature values for GJ-1, Nancy 91500, and NIST SRM610 were 0.09761 (Jackson et al., 2004),13 0.1793; (Wiedenbeck et al., 1995),14 and 0.257 (Stern and Amelin, 2003; Jochum and Nohl, 2008).15,16 | ||||||||
Plešovice (zircon for normalisation) | 700 | 200 | 0.03466 | 1.7 | 1.550 | — | 0.05372 | — |
20 | 0.03464 | 1.551 | — | — | ||||
2 | 0.03461 | 1.552 | — | — | ||||
0.2 | 0.03438 | 1.563 | — | — | ||||
0.1 | 0.03417 | 1.572 | ||||||
0.04 | 0.03377 | 1.591 | ||||||
0.02 | 0.03353 | 1.602 | — | — | ||||
GJ-1 (zircon) | 350 | 0.02 | 0.0609 | 2.4 | — | 0.0943 | 0.09761 | −3.4 |
— | 0.0944 | −3.3 | ||||||
— | 0.0945 | −3.2 | ||||||
— | 0.0951 | −2.6 | ||||||
— | 0.0957 | −2.0 | ||||||
— | 0.0968 | −0.8 | ||||||
— | 0.0975 | −0.1 | ||||||
Nancy 91500 (zircon) | 80 | 0.02 | 0.1116 | 2.8 | — | 0.173 | 0.1793 | −3.5 |
— | 0.173 | −3.4 | ||||||
— | 0.173 | −3.4 | ||||||
— | 0.174 | −2.7 | ||||||
— | 0.176 | −2.1 | ||||||
— | 0.178 | −0.9 | ||||||
— | 0.179 | −0.2 | ||||||
NIST SRM610 (glass) | 450 | 0.02 | 0.1591 | 1.8 | — | 0.247 | 0.257 | −4.0 |
— | 0.247 | −4.0 | ||||||
— | 0.247 | −3.9 | ||||||
— | 0.249 | −3.3 | ||||||
— | 0.250 | −2.7 | ||||||
— | 0.253 | −1.5 | ||||||
— | 0.255 | −0.8 |
The resulting 206Pb/238U ratios for GJ-1, Nancy 91500, and NIST SRM610 are listed in Table 1. For easier comparison, in Fig. 2, the percent (%) relative deviation of the measured 206Pb/238U ratios from the literature values was plotted against the dwell times. The reported 206Pb/238U ratios for GJ-1, Nancy 91500, and NIST SRM610 are 0.09761,13 0.17928,14 and 0.257,15,16 respectively. Uncertainties given in each data point were the two-times standard deviation (2SD) calculated based on five repeated measurements. With the conventional data acquisition with dwell times of 2–200 ms, the measured 206Pb/238U ratios for GJ-1, Nancy 91500, and NIST SRM610, were about 3.5–4% lower than the literature values. In contrast, with a shorter dwell time (i.e., 0.04 to 0.02 ms), the measured 206Pb/238U ratios for the zircons (GJ-1 and Nancy 91500) and glass standard (NIST SRM610) agreed with the literature values within the analytical uncertainties achieved in this study (about 2–3%).
The slightly biased 206Pb/238U ratios found in zircons can be attributed to the erroneous correction of the counting loss based on the moderated signal intensities of 238U on the Plešovice zircon, that has the highest U contents among the samples analysed in this study (about 700 μg g−1 (ref. 12)). With the moderated signal intensity data obtained with the longer dwell times, the magnitude of the counting loss on 238U signals would be underestimated, resulting in biased 206Pb/238U ratios from the true value. Even with the shortest dwell time (0.02 ms), the resulting 206Pb/238U ratios for NIST SRM610 were slightly (about 1%) lower than the literature value reported by Jochum and Nohl,15 and Stern and Amelin.16 Since the 206Pb/238U ratios for two zircons (GJ-1 and Nancy 91500) agreed well with the literature values, the residual discrepancy found in NIST SRM610 could be originating from the systematical deviations in the reported 206Pb/238U ratios for NIST SRM610. In fact, the reported abundance values for Pb and U in NIST SRM610 were 413.3–428 μg g−1 and 447–461.5 μg g−1, respectively, and thus the overall uncertainty in the calculated 206Pb/238U ratio based on the Pb and U abundances, together with Pb and U isotopic ratios, exceeds 5%. This suggests that the resulting 206Pb/238U ratios for NIST SRM610 were in good agreement with the reported values within the uncertainties. However, it should be noted that the matrix effect can be a cause of discrepancy in the 206Pb/238U ratio between the present data and the literature values, and thus, this remains a possibility.
One may consider that the erroneous measurements due to the moderation can be avoided by reducing the signal intensities of the analytes. The lower signal intensity, however, induces both the higher contribution of the counting statistics and the lower signal-to-noise ratio. The data obtained here indicate that great consideration must be given to the underestimation of the counting loss due to the detector dead time in the case of elemental/isotopic ratio measurements being performed from transient signals.
The technique can also be applied for the high-resolution depth profiling analysis of trace-elements and the elemental/isotopic analysis of nanoparticles (NPs), where the signal intensities of analytes can change dramatically within a short time period.17–19 In fact, since the signal intensity data obtained from large-sized NPs (e.g., >50 nm) can exceed 10 Mcps, the incomplete correction of the counting loss can result in erroneous size and elemental/isotopic ratio data. Recently, we reported the elemental/isotopic ratio data from individual NPs based on data acquisition with short dwell times (e.g., 0.02–0.03 ms). After the correction of the counting loss, the measured isotopic ratios for Os and Pt showed good agreement with the literature values.20,21 To derive reliable size and number concentration data, as well as elemental/isotopic signatures from the NPs, data acquisition with short dwell times is highly desired.
With a shorter dwell time (i.e., 0.02 ms), the resulting 206Pb/238U ratios for two zircons (Nancy 91500, and GJ-1) and the glass standard material (NIST SRM610) agreed well with the literature values within the analytical uncertainties achieved in this study (about 2–3%, 2SD, N = 5). The technique developed here can become a powerful tool to derive reliable U–Pb age data from the outer thin-layer of the minerals (<1 μm), that can tell us the thermal formation sequence of the minerals underlying various geological events, as well as carry out elemental/isotopic analyses of nanoparticles (NPs).
This journal is © The Royal Society of Chemistry 2021 |