Technical note: ²¹Ne in the CoQtz-N quartz standard material

Intercomparison standards produced from mineral samples exposed to the natural cosmic ray flux are routinely used for interlaboratory comparison and normalization of measurements of the cosmogenic noble gases ³He and ²¹Ne. This effort is facilitated by the availability of multiple standard materials with a wide range of cosmogenic nuclide concentrations. The “CoQtz-N” quartz standard, which was originally produced and distributed as an intercomparison standard for cosmogenic ¹⁰Be and ²⁶Al, has a significantly lower cosmogenic nuclide concentration than other mineral standards used for cosmogenic noble gas analysis and is therefore potentially useful for assessing the linearity of interlaboratory offsets across a wide range of concentrations. This paper reports ²¹Ne analysis of 13 aliquots of CoQtz-N, interspersed with 14 aliquots of the commonly used CRONUS-A quartz standard, in three analytical sessions on two noble gas mass spectrometer systems. The excess ²¹Ne concentration in CoQtz-N, normalized to the accepted value of 320 × 10⁶ atoms g⁻¹ for CRONUS-A, is 13.87 ± 0.24 × 10⁶ atoms g⁻¹. Alternatively, the ratio of excess ²¹Ne concentrations in CoQtz-N and CRONUS-A is 0.04330 ± 0.00077.

1 Intercomparison standards for cosmogenic nuclides produced in quartz

Several cosmic-ray-produced nuclides, most importantly ¹⁰Be, ¹⁴C, ²⁶Al, and ²¹Ne, are produced in the common mineral quartz when it is exposed to the cosmic ray flux at or near the Earth's surface, and concentrations of these nuclides are routinely measured for various applications in geochronology and Earth surface process studies. In an effort to improve accuracy, precision, and reproducibility of such measurements, a number of homogenized quartz standard materials have been produced and distributed among laboratories active in this research area. These materials are naturally occurring samples without any independent knowledge of their irradiation histories, so they are not primary or certified concentration standards but are mainly intended for interlaboratory comparison. Quartz intercomparison standards that have been distributed include the CRONUS-A and CRONUS-N standards, which are described by Jull et al. (2015) and originally intended for interlaboratory comparison of ¹⁰Be, ¹⁴C, and ²⁶Al measurements; the CREU-1 standard described by Vermeesch et al. (2015) and intended for ²¹Ne intercomparison measurements; and the CoQtz-N standard described by Binnie et al. (2019) and intended for ¹⁰Be and ²⁶Al intercomparison. CRONUS-A has also been used for interlaboratory comparison of ²¹Ne measurements (Vermeesch et al., 2015; Balco et al., 2019).

Interlaboratory comparison of cosmogenic noble gas measurements by means of these standard materials, both for ²¹Ne using CREU-1 and CRONUS-A and for ³He using the CRONUS-P pyroxene standard (Schaefer et al., 2016; Blard et al., 2015), has been valuable in revealing systematic offsets among ²¹Ne and ³He measurements performed at different laboratories and on different analytical systems (e.g., Balco et al., 2019; Balter-Kennedy et al., 2023). These differences often exceed nominal absolute concentration uncertainties claimed by the various laboratories. Although a number of reasons for this overdispersion have been proposed (Blard et al., 2015; Vermeesch et al., 2015), some data indicating that inter-system/interlaboratory offsets are constant across multiple samples and a wide range of ${}^{\mathrm{21}}\mathrm{Ne}{/}^{\mathrm{3}}\mathrm{He}$ concentrations (Balco et al., 2019; Balter-Kennedy et al., 2023) imply that the offsets are most likely the result of undetected inaccuracies in pressure/volume calibrations used to prepare primary gas standards. This, in turn, implies that ²¹Ne and ³He measurements can be standardized among laboratories by applying constant normalization factors derived from analysis of mineral intercomparison standards. Even if the true concentration of an intercomparison standard is not known from first principles, standardization to a common value reconciles measurements from different laboratories used for production rate calibration and exposure age determination, thus improving the accuracy and precision of exposure dating and other surface-process-related applications.

Verifying that interlaboratory offsets are not concentration- or sample-dependent, of course, requires multiple intercomparison standards with a wide range of cosmogenic noble gas concentrations. This paper reports ²¹Ne measurements on the CoQtz-N standard, which has significantly lower cosmogenic nuclide concentrations than CRONUS-A and CREU-1, to facilitate its use for interlaboratory comparison for ²¹Ne as well as ¹⁰Be and ²⁶Al as originally intended. The CoQtz-N standard was prepared by Binnie et al. (2019) from a vein quartz boulder exposed in a lag deposit near the locality of Hakos, Namibia, and has ¹⁰Be and ²⁶Al concentrations of 2.5 × 10⁶ and 15.6 × 10⁶ atoms g⁻¹, respectively. Although the precise location at which the sample was collected was not reported, this locality is near 23.25° S latitude and at approximately 1850 m elevation. Given assumptions of steady-state erosion, this estimated location, and a sample thickness of 15 cm estimated from Fig. 1 of Binnie et al. (2019), the ¹⁰Be and ²⁶Al concentrations imply an erosion rate near 2.2 m Myr⁻¹ and an expected cosmogenic ²¹Ne concentration near 12 × 10⁶ atoms g⁻¹. As the CRONUS-A and CREU-1 ²¹Ne intercomparison standards have extremely high ²¹Ne concentrations (320 and 348 × 10⁶ atoms g⁻¹, respectively), CoQtz-N is potentially useful for interlaboratory comparison at less extreme concentrations.

2 Analytical methods

The data reported here consist of measurements of Ne isotope concentrations in the CRONUS-A and CoQtz-N quartz standards, measured in three analytical sessions on two noble gas mass spectrometers and associated extraction systems at the Berkeley Geochronology Center (BGC) and the Lawrence Livermore National Laboratory (LLNL) Noble Gas Lab, such that multiple aliquots of both standards were analyzed during the same analytical sessions (Fig. 1). Here, “analytical session” means a continuous period of measurements during which the mass spectrometer tuning and extraction and analysis procedure were not changed. CRONUS-A and CoQtz-N were originally prepared by repeated HF etching of quartz samples sieved to the 0.25–0.8 and 0.25–0.71 mm grain-size ranges, respectively. For this work, both standards were analyzed as they were supplied, without further preparation. Although there is no reason to expect that cosmogenic ²¹Ne concentrations vary with grain size, each bulk container was thoroughly mixed to homogenize the grain size distribution to the extent possible before subsampling of aliquots for analysis.

²¹Ne measurements at BGC were carried out on the BGC “Ohio” mass spectrometer and associated extraction line as described in a number of studies (e.g., Balco et al., 2019; Balter-Kennedy et al., 2020; Bergelin et al., 2022). 75–150 mg of quartz is encapsulated in a crimped segment of Ta tubing, which is placed under vacuum and heated with an 810 nm, 150 W diode laser coupled to a coaxial optical pyrometer in a control loop that allows for continuous adjustment of the laser power to achieve a fixed pyrometer temperature. At BGC, all samples described here were analyzed in two heating steps at 850 and 1100 °C (the complete heating schedules are listed in Supplement Table S1). Gases released during heating are reacted with cold SAES getters, and remaining noble gases are frozen to an activated charcoal cold trap at 33 K. Helium is pumped away, and the cold trap is warmed to 75 K to release Ne into a MAP-215-50 mass spectrometer originally constructed in 1992 but modernized with updated detectors and counting electronics. All Ne isotopes, along with other masses necessary for isobar corrections on masses 20 and 22, are measured on a single Channeltron-type electron multiplier operated in ion-counting mode. As this mass spectrometer cannot resolve ²⁰Ne⁺ and ⁴⁰Ar${}^{++}$ on mass 20, correction for ⁴⁰Ar${}^{++}$ is accomplished using a ³⁹Ar spike that allows monitoring of the ${\mathrm{Ar}}^{++}/{\mathrm{Ar}}^{+}$ ratio during analysis, as described in Balco and Shuster (2009). Correction for CO${}_{\mathrm{2}}^{++}$ on mass 22, as also described in that reference, is accomplished by obtaining a relationship between Ar and CO₂ charge ratios in offline experiments. Although Ar and CO₂ signals and therefore corrections are variable among analyses, typical corrections are equivalent to 0.15–0.22 × 10⁹ atoms ²⁰Ne and 2.5–2.7 × 10⁶ atoms ²²Ne. No ²⁰NeH contribution on mass 21 can be detected on this instrument under typical operating conditions. Ne isotope signals after isobar correction are further corrected for “cold” blanks measured without any introduction of sample to the extraction line; cold blanks are <1 % of observed signals. The $\mathrm{HF}/{\mathrm{H}}_{\mathrm{2}}^{\mathrm{18}}\mathrm{O}$ interferences are not resolvable on this instrument and are therefore included in the cold blank correction; they are most likely negligible under normal conditions. “Hot” blanks in which empty Ta packets are heated as if they were samples were measured separately from this series of analyses whenever new batches of Ta were received. Hot blanks released small amounts of Ne with ²¹Ne abundances in the range of 0–50,000 atoms and Ne isotope composition indistinguishable from atmosphere. As the aim is eventually to calculate excess ²¹Ne in relation to atmosphere, a separate hot blank correction is therefore not made. Calibration of Ne isotope signals to absolute amounts is by peak height comparison with an air standard, originally calibrated by pressure/volume measurements using an MKS Baratron pressure transducer, which is analyzed several times daily. By varying the amount of the air standard delivered, the machine response was found to be linear in the pressure range encompassing the samples analyzed here. Corrections for mass discrimination also utilize the air standard.

²¹Ne measurements at LLNL utilized an effectively identical laser heating system coupled to the LLNL “NG2” extraction line and mass spectrometer system. The sample chamber used for laser heating has a slightly different internal configuration, which is less effective at preventing cross-heating of adjacent samples than the BGC chamber. Thus, in the BGC system, samples could be analyzed in two heating steps at 850 and 1100 °C, in which lower-temperature heating steps were completed for all samples in the vacuum chamber before proceeding to high-temperature steps, without risk of gas loss from adjacent samples by cross-heating, but the LLNL system required analysis in three heating steps at 450, 850, and 1100 °C. Gases released by heating are reacted with SAES getters, noble gases are trapped on activated charcoal at 33 K, and neon is released into the mass spectrometer at 70 K. The difference in release temperatures between BGC and LLNL procedures is because the two charcoal traps were found to have slightly different noble gas adsorption characteristics. The LLNL-NG2 mass spectrometer is a Nu Instruments Noblesse HR with five switchable Faraday/multiplier detectors at a spacing nominally intended for multicollection of five Ar isotopes but also permitting simultaneous measurement of ²⁰Ne, ²¹Ne, and ²²Ne. This instrument is capable of edge resolution of ²⁰Ne⁺ and ⁴⁰Ar${}^{++}$ separately (see Saxton, 2020). Thus, the measurement sequence consists of (i) simultaneous measurement of the combined ${}^{\mathrm{22}}{\mathrm{Ne}}^{+}/{\mathrm{CO}}_{\mathrm{2}}^{++}$ peak on mass 22, the ²¹Ne peak on mass 21, and combined ${}^{\mathrm{20}}{\mathrm{Ne}}^{+}/\mathrm{HF}$ on mass 20; (ii) the edge-resolved HF peak flat on the high side of the ²⁰Ne⁺ peak; and (iii) CO${}_{\mathrm{2}}^{+}$ on mass 44. A corrected ²⁰Ne signal is obtained by subtracting the HF signal from the combined mass 20 signal; the HF contribution is typically <0.1 % and therefore effectively negligible but is monitored in any case because quartz samples are commonly prepared by HF etching, so a possibility exists that the combustion of fluorides could lead to a significant HF interference (however, none was observed for these samples). A corrected ²²Ne signal is then obtained using a ${\mathrm{CO}}_{\mathrm{2}}^{++}/{\mathrm{CO}}_{\mathrm{2}}^{+}$ ratio measured offline between analyses; the charge ratio for this instrument during these measurements was found to be in the range 0.0155–0.0157, and corrections were equivalent to 2–12 × 10⁶ atoms ²²Ne. No ²⁰NeH contribution on mass 21 could be observed under these operating conditions, and discussion in Saxton (2020) indicates that it is likely negligible. Corrections for mass discrimination and detector gains utilize an air standard that contains a similar amount of Ne to typical samples and that is analyzed several times daily. Results of cold and hot blank analyses on the LLNL system were equivalent to the results described above for the BGC system, and blank corrections were carried out in the same way.

Calibration of Ne isotope signals to absolute abundances on the LLNL system also utilizes the air standard. At the beginning of this work, the air standard on the LLNL system was known to contain approximately the same amount of atmosphere as the BGC air standard but had not been precisely calibrated. Therefore, for the purpose of cosmogenic ²¹Ne measurements, a precise calibration was established by analyzing 11 aliquots of the CRONUS-A standard (see Results and discussion below) and assigning the absolute Ne abundance in the standard such that the mean excess ²¹Ne concentration in all aliquots was equal to the accepted value of 320 × 10⁶ atoms g⁻¹ (Vermeesch et al., 2015). Aliquots of air and CRONUS-A of various sizes were then analyzed to verify that the sensitivity was linear over the relevant pressure range. Although the procedure of calibrating a gas standard from a mineral standard that is not a primary or certified standard has not routinely been used in noble gas analysis, it is effectively equivalent to the more common practice of establishing the Ne abundance in the gas standard from first-principles pressure/volume measurements and then normalizing subsequent ²¹Ne measurements to an accepted value for CRONUS-A. In fact, as the results below indicate that excess ²¹Ne in CRONUS-A can be routinely measured to ∼ 1 % precision, the 8 % overdispersion observed in the intercomparison results of Vermeesch et al. (2015), among systems calibrated against first-principles gas standards, suggests that calibrating a gas standard against CRONUS-A measurements may be more precise, if not more accurate, than calibrating based on first principles.

Figure 1 Excess ²¹Ne concentrations in CRONUS-A and CoQtz-N standards, measured in three analytical sessions at BGC (August 2021) and LLNL (April 2024, November 2024). Data are shown in order of measurement, although multiple heating steps alternated for sets of samples run in the same batch. Error bars are 68 % confidence. The solid and dotted black lines indicate means and standard deviations of BGC and LLNL measurements.

For all analyses, excess ²¹Ne is then calculated as $N_{\mathrm{21},\mathrm{xs}}=N_{\mathrm{21},\mathrm{tot}}-N_{\mathrm{20},\mathrm{tot}}{R}_{\mathrm{2120}\mathrm{a}}$, where N_21,xs is the excess ²¹Ne concentration (atoms g⁻¹), N_20,tot and N_21,tot are the total measured concentrations of ²⁰Ne and ²¹Ne (atoms g⁻¹), and R_2120a is the atmospheric ${}^{\mathrm{21}}\mathrm{Ne}{/}^{\mathrm{20}}\mathrm{Ne}$ ratio (0.002959; see Eberhardt et al., 1965). Excess ²¹Ne concentrations in all heating steps for each sample are then summed (see further discussion below). Reported uncertainties for both BGC and LLNL measurements include (i) counting uncertainties on all masses, fully propagated through all detector gain, mass discrimination, and isobaric interference corrections, and (ii) an uncertainty in machine sensitivity derived from the reproducibility of the air standards. For both systems, the largest component uncertainty is the reproducibility of the ²¹Ne signal in the air standard, which is approximately 1.5 % for the BGC system and 0.8 % at LLNL. For CRONUS-A (see below), this results in total calculated uncertainties on the excess ²¹Ne concentration of approximately 1.8 % for BGC measurements and 1 % for LLNL measurements. As machine sensitivities are similar on both instruments and ²¹Ne signals are relatively high for CRONUS-A, the improved uncertainties on the LLNL system for CRONUS-A are mostly attributable to improved instrument stability and therefore better reproducibility of the ²¹Ne signal in the standards, most likely because the electronics on the Nu Noblesse are 30 years younger than those on the MAP-215. The improvement in counting statistics due to multicollection and the simplified isobar corrections due to higher instrument resolution make minor contributions. For CoQtz-N, signals on mass 21 are lower, so counting statistics make a larger contribution to the uncertainty budget, but differences in calculated uncertainties between systems are likewise primarily due to improved reproducibility of ²⁰Ne and ²¹Ne signals in the air standards.

Figure 2 Neon three-isotope plot for step-heating analyses of CoQtz-N at BGC and LLNL. The black dot is the composition of atmospheric neon, and the black line is the atmospheric–cosmogenic mixing line. Ellipses are 68 % confidence regions.

3 Results and discussion

Three measurements of the excess ²¹Ne concentration in CRONUS-A at BGC had a mean and standard deviation of 318.7 ± 3.9 × 10⁶ atoms g⁻¹, which is indistinguishable from the accepted value of 320, so the BGC measurements were not renormalized (Fig. 1; Supplement Table S1). The 3 aliquots of CoQtz-N measured at BGC had excess ²¹Ne concentrations with mean and standard deviation of 13.7 ± 1.2 × 10⁶ atoms g⁻¹. A total of 11 aliquots of CRONUS-A measured at LLNL, as noted, were used for calibration of the air standard, so they had excess ²¹Ne concentrations with mean and standard deviation of 320.0 ± 3.8 × 10⁶ atoms g⁻¹. A total of 10 aliquots of CoQtz-N measured at LLNL had mean and standard deviation of 13.81 ± 0.80 × 10⁶ atoms g⁻¹. Thus, CoQtz-N measurements at BGC and LLNL agree. The 13 measurements together have an MSWD of 0.80, an error-weighted mean of 13.87 × 10⁶ atoms g⁻¹, and a standard error of 0.24 × 10⁶ atoms g⁻¹. The p value of the χ² statistic is 0.66, so the hypothesis that the measurements belong to a single population and differ only due to measurement uncertainties is not rejected, or, equivalently, no overdispersion is present.

Ne isotope ratios for all heating steps of CoQtz-N are indistinguishable from the cosmogenic–atmospheric mixing line (Fig. 2). Low-temperature heating steps are enriched in cosmogenic neon relative to higher-temperature steps, as is typical for most quartz samples. High-temperature steps have an isotope composition indistinguishable from atmosphere when each analysis is considered individually (Fig. 2), but all except one of the high-temperature steps have ${}^{\mathrm{21}}\mathrm{Ne}{/}^{\mathrm{20}}\mathrm{Ne}$ ratios slightly above atmosphere, which would be statistically improbable if in fact only atmospheric Ne were present. Thus, high-temperature steps contain a nonzero amount of excess ²¹Ne and are included in calculated total excess ²¹Ne for each sample. Atmospheric Ne concentrations in CoQtz-N quartz, expressed as ²¹Ne, are 35–50 × 10⁶ atoms g⁻¹ (Supplement Table S1), so excess ²¹Ne is about 25 % of total ²¹Ne. Although the fact that most of the ²¹Ne in CoQtz-N is atmospheric limits the precision with which the excess ²¹Ne concentration can be measured, measurements using a modern mass spectrometer had individual uncertainties in the 5 %–6 % range and an aggregate uncertainty on the full set of measurements near 2 %, which is adequate for monitoring the linearity of laboratory offsets across a range of ²¹Ne concentrations.

Figure 3 Normalized paired nuclide plots for ¹⁰Be, ²¹Ne, and ²⁶Al concentrations in CoQtz-N, assuming an estimated sample location and thickness as discussed in the text. Production rate calculations use the “St” scaling method and default production rate calibration in version 3 of the online exposure age calculator originally described by Balco et al. (2008) and subsequently updated. The ellipses are 68 % confidence regions. If all excess ²¹Ne is assumed to be cosmogenic (darker ellipses), then ²¹Ne concentrations show disequilibrium with ¹⁰Be and ²⁶Al. Assuming that 2 × 10⁶ atoms g⁻¹ of excess ²¹Ne is nucleogenic or otherwise inherited, and subtracting this from the total (lighter ellipses), results in equilibrium concentrations of all three nuclides.

As there are no independent constraints on the abundance or isotope ratio of any nucleogenic Ne that could be present in CoQtz-N, it is not possible to discern from Ne isotope ratios alone whether measured excess ²¹Ne in CoQtz-N is entirely cosmogenic or a mixture of cosmogenic and nucleogenic ²¹Ne. However, comparison with ²⁶Al and ¹⁰Be concentrations from Binnie et al. (2019) provides some insight into this. At the estimated sample location discussed above, ²⁶Al and ¹⁰Be concentrations lie on the steady erosion line at an erosion rate near 2.2 m Myr⁻¹ (Fig. 3). The total excess ²¹Ne concentration, however, implies a lower erosion rate, so ${}^{\mathrm{26}}\mathrm{Al}{/}^{\mathrm{21}}\mathrm{Ne}$ and ${}^{\mathrm{10}}\mathrm{Be}{/}^{\mathrm{21}}\mathrm{Ne}$ pairs lie below the simple exposure region. This indicates that some portion of the total excess ²¹Ne concentration is either nucleogenic or inherited from a prior period of exposure. Assuming that 2 × 10⁶ atoms g⁻¹ of the total excess ²¹Ne concentration is either nucleogenic or otherwise inherited, and subtracting this from the total, reconciles all three nuclide pairs with steady-state erosion (Fig. 3). Thus, up to approximately 2 × 10⁶ atoms g⁻¹ nucleogenic ²¹Ne may be present in CoQtz-N. However, as also noted by Vermeesch et al. (2015), the presence or absence of nucleogenic ²¹Ne does not affect the utility of CoQtz-N as an interlaboratory comparison standard for excess ²¹Ne measurements.

4 Summary

The excess ²¹Ne concentration in the CoQtz-N standard, normalized to the CRONUS-A standard with a nominal excess ²¹Ne concentration of 320 × 10⁶ atoms g⁻¹, is 13.87 ± 0.24 × 10⁶ atoms g⁻¹. Although the precision of excess ²¹Ne measurements in CoQtz-N is limited by the higher concentration of included atmospheric ²¹Ne, analysis of multiple aliquots on a typical modern noble gas mass spectrometer should yield a precise enough measurement to be of value for interlaboratory comparison and evaluation of linearity of interlaboratory offsets. Binnie et al. (2019) reported production of 25 kg of the CoQtz-N standard, which would be sufficient for approximately 200 000 ²¹Ne measurements. Even if the majority of this material is consumed for ¹⁰Be and ²⁶Al analyses, the remainder could, if made available for this purpose, support use as a ²¹Ne intercomparison standard at a number of noble gas laboratories for some time. According to Binnie et al. (2019), the CoQtz-N standard is available from S. Binnie or T. Dunai at the University of Cologne.