We established a new laboratory for noble gas mass spectrometry that is
dedicated to the development and application to cosmogenic nuclides at the
University of Cologne (Germany). At the core of the laboratory are a
state-of-the-art high-mass-resolution multicollector Helix MC Plus
(Thermo Fisher Scientific) noble gas mass spectrometer and a novel custom-designed
automated extraction line. The mass spectrometer is equipped with five
combined Faraday multiplier collectors, with 1012 and
1013Ω pre-amplifiers for faraday collectors. We describe the
extraction line and the automated procedure for cosmogenic neon and the
current performance of the experimental set-up. Performance tests were
conducted using gas of atmospheric isotopic composition (our primary
standard gas), as well as CREU-1 intercomparison material, containing a
mixture of neon of atmospheric and cosmogenic composition. We use the
results from repeated analysis of CREU-1 to assess the performance of the
current experimental set-up at Cologne. The precision in determining the
abundance of cosmogenic 21Ne is equal to or better than those reported
for other laboratories. The absolute value we obtain for the concentration
of cosmogenic 21Ne in CREU is indistinguishable from the published
value.
Introduction
Cosmogenic Ne isotopes are stable and compared to other cosmogenic
radionuclides (e.g. 10Be, 26Al) exhibit the potential to date
beyond the physical limit of radionuclides. The particular strength of
cosmogenic neon is its application to date quartz clasts of very old
surfaces (> 4 Ma) or very slowly eroding landscapes
(< 10 cm Ma-1), which is unattainable with most other radionuclides (Dunai,
2010). Cosmogenic Ne analysis can be applied to a range of neon-retentive
minerals (e.g. quartz, olivine and pyroxene), amongst which quartz is the
most commonly used. Ne can be measured on conventional sector field noble
gas mass spectrometers, is less time consuming and requires less
sample-preparation compared to AMS measurements required for the cosmogenic
radionuclides. Recent studies used cosmogenic Ne for dating old surfaces
(e.g. Ritter et al., 2018; Dunai et al., 2005; Binnie et al., 2020),
reconstructing erosion rates (e.g. Ma et al., 2016) or
10Be /21Ne burial dating (e.g. McPhillips et al., 2016).
The advantage of also using other minerals than quartz led to several studies
using 21Ne to date for example basalt flows (e.g.
Espanon et al., 2014; Gillen et al., 2010). Neon has three stable isotopes
20Ne, 21Ne and 22Ne, of which 20Ne is the most
abundant; the atmospheric 21Ne /20Ne and 22Ne /20Ne ratios
are 0.002959 ± 0.000022 and 0.1020 ± 0.0008, respectively
(Eberhardt et al., 1965). There are several recent
re-determinations of the atmospheric 21Ne /20Ne ratio (e.g.
Honda et al., 2015; Wielandt and Storey, 2019; Saxton, 2020; Györe et
al., 2019) one of which yields a ∼ 2 % lower value
(Honda et al., 2015). For the evaluation of our data, we utilise
the 21Ne /20Ne value of Wielandt and Storey (2019) of
0.0029577 ± 0.0000014 and for 22Ne /20Ne that of
Eberhardt et al. (1965). Note that in the context of the
determination of the abundance of cosmogenic nuclides in a sample, eventual differences between the used and the actual value of the atmospheric 21Ne /20Ne
ratio are unimportant, if (i) atmospheric neon is used as calibration gas,
(ii) the same value for the composition of atmospheric neon is used
consistently throughout the evaluation of the isotope data (mass
discrimination etc.) and calculation of abundances, and (iii) the atmospheric
value used is reported along with the data.
All three neon isotopes are produced in about equal proportions by neutron
spallation in quartz (Niedermann et al., 1994). Due to the lower
abundances of 21Ne and 22Ne as compared to 20Ne in air, and
the ubiquitous presence of atmospheric neon in samples, any contribution
from cosmogenic production in samples is most easily picked up with the
former two isotopes. Consequently, the neon three-isotope diagram with
20Ne as common denominator (Niedermann et al., 1994;
Niedermann, 2002) is customarily used to assess 21Ne data for the
presence of terrestrial cosmogenic Ne and its discrimination from other
non-atmospheric Ne components (Dunai, 2010). The latter may be
nucleonic Ne and/or mantle-derived Ne. Hence, the accurate determination of
cosmogenic Ne and its discrimination from other components requires the
accurate discrimination from any other component.
Common isobaric interferences for neon measurements are at m/e= 20
(40Ar2+, H19F+ and H218O+ interfering
with 20Ne+), at m/e= 21 (20NeH+, interfering with
21Ne+) and at m/e= 22 (44CO22+ interfering
with 22Ne+). 40Ar2+ and 12C16O22+
interferences are considered to be the main challenges for neon analysis.
Recent studies demonstrated the ability of the Helix MC Plus to fully resolve
the 40Ar2+, H19F+ and H218O+ peaks from
the 20Ne+ peak (e.g. Honda et al., 2015;
Wielandt and Storey, 2019) and its ability to reliably measure 21Ne at
an off-centre peak position that is free of interference from
20NeH+ (Honda et al., 2015; Wielandt and Storey,
2019). The remaining interference of 12C16O22+ at
m/e= 22 can be corrected via monitoring of the ratio of double- to single-charged CO2 in between samples (Honda et al., 2015) or the
measurement of 13C16O22+ at m/e= 22.5 during sample
analysis (Wielandt and Storey, 2019). Recently mass spectrometers with
higher resolution have become available, which permit almost full separation
of 12C16O22+ and 22Ne (Farley et al.,
2020).
Besides the resolution and characteristics of a noble gas mass spectrometer
to resolve and quantitatively determine neon compositions of an unknown
sample, the calibration, sample extraction and purification are crucial to
achieving accurate and reproducible results. Automation of extraction
protocols and workflows may assist in achieving a high degree of
reproducibility by eliminating inaccuracies or errors by operators having a
variable degree of expertise. In this paper, we describe the current set-up
of the noble gas mass spectrometer and its automated extraction line that is
located in the Institute of Geology and Mineralogy at the University of
Cologne (Germany), and we review its performance for neon analysis.
Experimental set-upNoble gas mass spectrometer
The Cologne noble gas laboratory is equipped with a Helix MC Plus from Thermo
Fisher Scientific with five CFM (combined Faraday multiplier) modules,
called “Aura”. The central, axial module (Ax) is fixed in position, and the four remaining modules (L1, L2 on the low mass side, and H1, H2 on the high mass side of Ax) can be moved. The mass spectrometer configuration and
performance is mostly equivalent to those described elsewhere
(Honda et al., 2015; Wielandt and Storey, 2019); here we
describe potential differences in configuration and performance parameters
that may be unique to a given instrument (Fig. 1).
(a) Schematic plan and (b) picture of the noble gas extraction and purification line at the University of Cologne. From left to right: Rofin Starfiber 600, full-protection laser cage (laser protection windows P1P10, Laservision) housing the laser extraction, clean-up unit, cryogenic separation unit and the Helix Plus NG-MCMS “Aura”. The laboratory is temperature-stabilised to ±0.5 ∘C. Further description is provided in the text.
In the instrument at Cologne University, all but one of the Faraday amplifiers are
equipped with 1013Ω resistors, one with 1012Ω
(H2). The L1 module has 0.3 mm wide collector slits, and all other modules have
0.6 mm wide slits. The CFM at L1 configuration is flipped (i.e. the
relative positions of the Faraday and multiplier are swapped) as compared to
the standard configuration, which is the only difference from the standard
configuration. The two SAES NP10 getters, at the source and the multiplier
block, are kept at room temperature during analysis.
For neon isotope analysis of calibrations and samples, we utilise the H1, Ax
and L1 CFMs (20Ne+ L1 Faraday; 22Ne+ H1 Faraday;
21Ne+ L1 multiplier; CO2+ H1 Faraday; for blanks we
utilise the L1 multiplier also for 20Ne+ and 22Ne++).
With the widest source slit (0.25 mm), mass resolution (at 5 % peak valley)
and mass resolving power (between 10 % and 90 % of peak) on the L1
detector with 0.3 mm collector slit width are approximately 1700 and 6500,
respectively. For the Ax and H1 detectors with 0.6 mm collector slit, the
corresponding values are approximately 1000 and 6000, respectively. As such
the system allows the interference-free determination of 20Ne and
21Ne; for 21Ne this entails measuring at an off-centre peak
position (Honda et al., 2015; Wielandt and Storey, 2019).
Extraction line
The Cologne noble gas extraction and purification line has a modular design
(Fig. 1). Modules are (i) extraction (currently only laser extraction; to be
joined by a crushing device), (ii) calibration gas pipettes and volumes,
(iii) clean-up, and (iv) cryogenic separation. The calibration module is
physically linked to the clean-up module, and the other modules can be
separated, if required. Among the common features of all modules is that all
valves and tubing in contact with the sample gas are made of metal; tubing
is of stainless steel or vacuum-annealed copper. Furthermore, all valves
used for handling of sample and calibration gas are pneumatically actuated
all-metal diaphragm valves (Fujikin MEGA-M LA; FWB(R)-71-6.35) that can be
operated at high temperature (up to 350 ∘C). Tubing and
valves in contact with sample gas are continuously kept at constant
temperature between 160 and 200 ∘C;
exceptions are the functional traps and portions of the tubing in the
cryogenic separation. Temperature is maintained with heating tapes (Horst HS
450 ∘C) and is controlled section-wise (Horst HT30). The
temperature of the heated sections is controlled to ±1 ∘C. Thermal insulation is achieved with high-temperature
resistant silicone foam (HOKOSIL®; resists ≤ 280 ∘C; permitting bake-out at higher than operation
temperatures). Vacuum connections used are VCR (for Fujikin valves), CF (for
adapters, getters and manifold in clean-up) and Swagelok (for flexible
tubing between modules and between ports of the cryogenic separator
(Swagelok 321 Stainless Steel Flexible Tubing with XBA adapter; copper
tubing). Tubing and valves are 1/4 in. outer diameter (Swagelok) or equivalent
(VCR, Fujikin)). The overall internal volume of the extraction line
(laser extraction, clean-up and cryogenic separation) is 530 cm3.
Outside the volume used for sample preparation, CF connections are used
throughout. A schematic overview and picture of the extraction line is
provided in Fig. 1.
More specifical information about the individual modules is given in the following:
Laser extraction module (Fig. 1). Up to 18 tungsten cups are loaded in
a sample revolver housed in a DN 200 CF flange sandwich. The sample
revolver is machined from molybdenum, which permits the heating of the
tungsten cups while being situated in the revolver. To minimise heat loss
through conduction, the cups sit on shards of zirconia (synthetic,
cubic-stabilised ZrO2). The tungsten cups can hold up to
∼ 600 mg quartz. The tungsten cups are reused. When analysing
quartz, tungsten cups are emptied with a suction micropicker (Micropicker
MPC100; VU Amsterdam), while remaining in the sample revolver. In cases
where samples are melted during extraction, tungsten cups could be cleaned
in HF (then of course outside the revolver). For sample loading the volume
containing the revolver is vented and continuously flushed with high-purity
nitrogen. During laser extraction the pressure is monitored (MEAS EPB-C1
sensor, welded into a male VCR connector; Disynet), and in case of an eventual
failure of the viewport, the extraction volume is automatically purged with
Argon. Energy for the heat extraction is provided by an output-tuneable 600 W fibre laser (Rofin StarFiber600) at 1064 nm wavelength through galvanometer
scanner optics (Rofin RS S 14 163/67 0∘) and a sapphire viewport
(Kurt Lesker, VPZL-275DUS). For neon extraction of quartz, the cups are
covered with tungsten lids; the heating occurs via scanning of the lids
(scanning speed 20 cm s-1; rastering a circular area of 10 mm diameter) with a
defocussed (∼ 0.5 mm diameter) continuous wave beam with 100 W
power for 15 min. Copper (melting point 1085 ∘C), placed in
the cup assemblies, melts at 80 W laser power (15 min extraction time); we assume that at 100 W laser power the internal temperature is ≥ 1200 ∘C. The temperature of the top of the tungsten lids
is monitored with a pyrometer (CellaTemp PA 29 AF 2/L; Keller HCW). The
laser extraction has a dedicated pumping unit (Pfeiffer HiCube80). Pressures
attained after sample loading and heating of the revolver (via short-term
laser heating – stepwise increased to 200 W – of an empty tungsten cup; the external housing flanges reach ∼ 50 ∘C during
this treatment; temperatures in adjacent cups in the revolver stay below
156.6 ∘C, which was verified with Indium wire) are usually
< 5 × 10-9 mbar (the lower limit of the pressure gauge used) after one night of pumping. Typical blanks, obtained via heating of an empty tungsten cup assembly, are ∼ 0.3 fmol neon. A detailed
description of this novel laser furnace will be provided elsewhere.
Calibration gas pipette module (Fig. 1). The gas pipettes are assemblies of
male and female versions of pneumatically actuated Fujikin diaphragm valves
(MEGA-M LA; FWB(R)-71-6.35); the reservoirs were manufactured by Caburn-MDC, and the insides of the reservoirs are electropolished. We currently have three
different gases available for noble gas calibration (“Linde”, “Air”,
“RedAir”). “Linde” is a noble gas mixture in nitrogen (9.889 ± 0.009 % He, 10.00 ± 0.01 % Ne, 10.01 ± 0.01 % Ar,
0.00987 ± 0.0003 % Kr; 0.01023 ± 0.00002 % Xe; all
uncertainties are ±2σ; remainder N2; prepared
gravimetrically by Linde; values as certified by Linde according to DIN ISO
6141) the He is enriched in 3He (12.3 ± 0.3 Ra; ±2σ; value as certified by Linde according to DIN ISO 6141), and the
remaining noble gases have atmospheric composition. We assume that the
cryogenically purified atmospheric gases used by Linde were not fractionated
during this process; we have verified this for Ne within the limits of
uncertainties reported in this paper. “Air” is a reservoir of air at
atmospheric pressure and “RedAir” a reservoir of air at reduced pressure
(lab name RedAir is the abbreviation of that fact). For the neon
determinations we utilise RedAir. The volumes of all reservoirs and the
pipettes have been determined relative to a gravimetrically calibrated gas
volume (an assembly of a Swagelok SS-4H valve and a Swagelok SS-4CS-TW-50
miniature cylinder; repeatedly weighed (Sartorius MSA524P-1000-DI; the
balance was calibrated prior to calibration of the reference volume) under
vacuum and filled with air at a temperature (n= 16), pressure and relative
humidity measured with traceable and/or certified sensors (thermometer:
testo 110; manometer: Greisinger GMH 3181-12, DKD certificate D19853,
D-K-15070-01-01; hygrometer: VWR traceable 628-0031); reference volume is
51.37 ± 0.18 cm3 (±1σ)). All other volumes (piping
of the calibration gas filling line; pipettes and reservoirs) were
determined by taking pressure readings (MKS Baratron, Type 628FU5TCF1B) from
repeated step-wise expansion of gases. The temperature in the room where
these calibrations were conducted was stable to ±0.5 ∘C over the course of the calibrations. The thus
determined volumes of the reservoir and pipette of RedAir are 8740 ± 35 and 1.457 ± 0.006 cm3 (±1σ),
respectively. For filling of the RedAir reservoir one pipette volume of
air was expanded into the reservoir; the temperature, pressure and humidity
at the time of filling of the pipette were measured with a traceable and
certified sensor (same as above). The first pipette volume extracted from
the RedAir reservoir contained 4.020 ± 0.027 × 10-9 cm3
(±1σ) atmospheric neon at standard temperature and pressure
(179 ± 1 fmol atmospheric neon; ±1σ).
Clean-up module (lab name “Sputnik” referring to the shape and protrusions of the central manifold and its faint resemblance to the first satellite; Fig. 1). Arranged around a central hexagonal 8-port manifold (Kimball
Physics, 2.75′′ spherical hexagon) are the sample/calibration inlet, the
pumping outlet, a pipette leading to a residual gas analyser (Hiden HAL/3F
PIC), two SAES NP50 getters (one operated hot; heating current 1.6 A;
∼ 300 ∘C, the other at room temperature;
getters are housed in SAES GP 50 W2F bodies; water cooling is optional, not
used during sample analysis), an optional expansion volume, an internally
heated capacitance manometer (MKS Baratron, Type 628FU5TCF1B; @
100 ∘C) and the outlet to the cryogenic separation unit
(Fig. 1). The sample/calibration inlet tubing has an auxiliary port, which
for example is used for the crushing extraction module (build around a T4S
crushing unit, VU Amsterdam). The clean-up module is pumped via a manifold
connected through gate valves (MDC E-GV-1500M-P) to a turbopump (Pfeiffer
HiPace 300; backed by a membrane pump, Pfeiffer MVP 030-3) and an ion pump
(Agilent, Vacion 40 plus Starcell).
Cryogenic separation module (Fig. 1). The centre of this module is a double-cold
trap unit (Janis, twin coldhead model 204) that has inlet and outlet lines
to three traps: a water trap (operated at 205 K), a bare steel trap (≥ 24 K) and a charcoal trap (≥ 10 K). The cold trap unit is controlled by a Lakeshore 336 Controller (Cryotronics). This module is pumped by an ion
pump (Agilent, Vacion 40 plus Starcell).
The performance of the bare cold trap unit for He, Ne and Ar separation was
calibrated using the Residual Gas Analyser (Hiden HAL/3F PIC). Neon is
quantitatively adsorbed on the bare trap at 24 K (> 99.9985 %
is adsorbed at 24 K; Fig. 2); in equilibrium about 60 % of the helium is
adsorbed at 24 K. We use this to separate helium from neon. Helium is
removed (distilled off in disequilibrium) either to the ion pump or the 10 K
charcoal head, the latter if the He is to be retained for analysis. Neon is
fully released from the bare trap at 80 K; at this temperature argon is
quantitatively retained on the bare trap, permitting quantitative separation
of the two gases (Fig. 2).
Desorption curves of Ne and Ar on the stainless-steel cold trap measured with the Hiden quadrupole. The uncertainties of the argon determinations at low fractions released are due to a significant Ar background of the quadrupole (e.g. measurement at 100 K was just 5 % higher than the background).
Besides its functionality to separate noble gases from each other the bare
trap serves as cold trap during Ne analysis (held at 80 K) and replaces a
liquid-nitrogen-cooled trap, which would otherwise customarily be used for
this purpose. The latter may introduce intensity fluctuations during
analysis due to changing coolant level, which we avoid with our set-up. The
last pneumatically actuated valve before the Helix-Plus MCMS serves as inlet
valve, and the manual valve of the Helix-Plus MCMS is permanently open.
Automation
The extraction and purification line can either be operated manually, via a
switchboard for the pneumatic valves and the components' original
controllers, or automatically via LabView. Manual operation is mainly used
for development of analytical routines, automatic operation generally for
sample and calibration-gas analysis. Automatic operation liberates the
operator from conducting necessarily repetitive tasks and thus helps to prevent
mistakes and inconsistencies from oversight or negligence; it allows gas purification and separation to be conducted under precisely identical
conditions. The latter is also assisted by avoiding liquid coolants, which
commonly are affected by variable coolant levels (unless automatically
filled with a suitably precise system or an experienced and conscientious
operator). Currently the laser system is operated manually (due to safety
regulations); all subsequent steps – until admission of the gas to the mass
spectrometer – are automated utilising LabView (Version 2018) in a Windows
10 environment. The mass spectrometry analysis of the purified gas is
conducted with Qtegra (Thermo Fisher Scientific).
Screenshots of the operating VI program interface of the Cologne (CGN) Noble Gas Helix MC Plus. (a) The neon VI informs the user in real time about current data, such as pressure and temperature, as well as about the current status of the preparation. (b) Valve circuit overview. M1–M5 indicate the different modules of the extraction line. Valve numbers (1–10, 20–31, M, T, I) are coloured depending on the current state (green = open, red = closed).
Reproducibility of standard gas RedAir measurements for sample runs at CGN noble gas lab, during the period between March 2020 and December 2020. Isotopic ratios are normalised to air for each run (mean of isotope ratios obtained in the run). The larger errors of the 21Ne /20Ne ratios of the second run may be due to the fact that prior to that run a longer development period of other noble gas species, and other sample materials, was conducted. During developmental work on a noble gas line, particularly when other gas species are analysed, the residual gas composition in the extraction line and in the mass spectrometer may change. The latter may affect the response and stability of multipliers (21Ne is the only isotope we measure on the multipliers; thus it is the 21Ne /20Ne that shows the higher variability). Stippled black lines delineate individual runs. Error bars on individual data points are ±1σ. Symbol size is commonly larger than the corresponding error bars, which may therefore be hidden.
Valve control electronics were developed and implemented in-house, including
digital input/output modules (I/O modules from National Instruments) and
RS-232 communication. The main devices such as SAES getter control, Lakeshore
Cryo-Controller, turbopumps and ion pumps already offered LabView-compatible
Sub-VIs (Virtual Instrument, program codes), which were implemented into
the operation VI (Fig. 3). The Agilent Ion Pump Control connection via the
computer interfaces was written and developed in-house.
The gauges and controllers of the turbo pumps (Pfeiffer) and ion pumps
(Agilent) are monitored via the operation VI (Fig. 3). Automatic safety
protocols are implemented to protect the extraction line and equipment
against sudden pressure increases. Temperature setting and monitoring of the
three cold traps (Janis Cryostat) is performed by the Lakeshore 336
controller, which in turn is controlled via the operation VI (Fig. 3).
LabView computing of the extraction sequence/protocol was programmed in
single commands and steps, joined into command sequences connected in series
as sub-VIs for each extraction protocol (various noble gases and sources of
samples or calibration gas). Pressure and temperature control sequences are
programmed in a continuous loop to ensure stability and safety during
operation. For handling, a structured user–program interface was designed
(Fig. 3), which provides the user with information about all parameters and
total duration and additionally logs every extraction step.
Neon three-isotope plot for CREU-1 intercomparison material measured in Cologne. Error bars are ±1σ. The cloud of green symbols displays single-step CREU extractions (100 W–15 min), the green dots to the right of the cluster are the initial heating (first extraction of a sample) steps of stepwise extractions (at varying laser output), grey rectangles are the subsequent steps that invariably had low abundance; for details see Table 1. Data of samples depicted in green are included in the regression calculation; data of the grey rectangles are excluded. The slope of the regression of the data (forced through air) is 1.078 ± 0.022 (±2σ), which is indistinguishable from the published value of 1.108 ± 0.014 (±2σ; Vermeesch et al., 2015). The dotted line denotes the 95 % confidence interval.
Compilation of CREU-1 21Ne concentrations (±2σ uncertainties) measured at Cologne (CGN), compared to reported 21Ne concentrations from interlaboratory comparison from Vermeesch et al. (2015) and data from the Peking noble gas lab from Ma et al. (2015). Black bars were considered outliers by the original authors and not used for calculation of averages (Vermeesch et al., 2015). The data are divided into three sections, each for a different CREU-1 grain size analysed. The average 21Ne concentration for CREU-1 of 3.48 ± 10 × 108 atoms g-1 reported by Vermeesch et al. (2015) is marked as a light-grey band and a red line for the mean. Lab-individual error-weighted means are displayed as black lines with their respective uncertainty in dark grey. The average obtained for CREU-1 at Cologne (all grain sizes, n= 22) is 3.48 ± 0.02 × 108 atoms g-1 (±2σ; error-weighted standard deviation). The MSWD values (mean square of the weighted deviates (“reduced Chi-square”, Mcintyre et al., 1966)) are reported for all individual laboratory means (Vermeesch et al., 2015; this study). CGN = University of Cologne, ETH = Eidgenössische Technische Hochschule Zürich, BGC = Berkeley Geochronology Center, SUERC = Scottish Universities Environmental Research Centre Glasgow, CRPG = Centre de Recherches Pétrographiques et Géochimiques Nancy, GFZ = Deutsches GeoForschungsZentrum Potsdam.
Analytical procedure
Quartz samples are cleaned using standard procedures using dilute HF as
etchant (Kohl and Nishiizumi, 1992). Up to 600 mg of quartz are
loaded into tungsten cups and covered with a tungsten lid; the latter has a
small hole to facilitate gas release. When opening the laser furnace for
re-loading, the furnace is vented and purged with a continuous flow of pure
nitrogen. In normal operation, after the initial installation and bake-out,
the internal parts of the furnace are never again exposed to air. The
tungsten cups and lids remain in the nitrogen atmosphere during sample
(re-)loading. Cups are emptied with a suction micropicker (Micropicker
MPC100, VU Amsterdam) while seated in the revolver, and weighed samples are
transferred from the glass vials into the cup through a miniature metal
funnel (glass funnels produced undesirable static effects). After reloading,
the sample revolver is heated by firing the laser on an empty cup; pressure
< 5 × 10-9 mbar is usually achieved after pumping overnight.
During this clean-up, and during subsequent analyses, the temperature of
adjacent cups does not exceed 156.6 ∘C (verified with
Indium wire). Cosmogenic Ne is extracted from quartz by heating the sample
with a defocussed laser beam at 100 W for 15 min; at these settings, the
cup insides reach ∼ 1200 ∘C. This temperature
allows reliable extraction of cosmogenic neon
(Vermeesch et al., 2015). After heating the
furnace, it is allowed to cool for 5 min before the sample is
expanded to the clean-up module.
For calibrations, the calibration gas is expanded for 30 s into the
pipette, and the pipette volume is then expanded into the clean-up volume (Fig. 1). After this step, purification is identical for sample and calibration
gases. The pipetting of calibration gas and the purification of sample and
calibration gases are fully automatised.
Reactive gases are removed by sequential exposure to two metal getters (SAES
NP50, Fig. 1); the first is operated hot, the other at room temperature. The
gas is exposed to each for 15 min. Subsequently the gas is exposed to the
water trap at 205 K for 10 min. The remaining inert gases are exposed to the bare-metal trap at 24 K for 20 min, which is then pumped for 5 min to remove
helium from the sample gas (Fig. 1). The trap is then isolated and heated to
80 K, followed by 5 min holding time for re-equilibration. Neon
is quantitatively released, and argon is quantitatively retained on the
trap. Subsequently, Ne gas is expanded into the Helix MCMS for analysis. The
bare trap at 80 K remains connected to the mass spectrometer during
analysis, for pumping of CO2 and Ar evolving from the mass
spectrometer.
The configuration of the Helix is described above. For maximum sensitivity
and precision for abundance determination (Wielandt and Storey, 2019),
we use the widest (0.25 mm) source slit for neon analysis. We run the source
at an electron energy of 115 eV, trap current of 200 µA and an
acceleration voltage of 9.9 kV.
20Ne is measured on the high-resolution L1 Faraday cup (fitted with
1013Ω pre-amplifier), fully resolved from 40Ar2+ and
from molecular interferences such as HF+, H218O+.
21Ne is measured off-centre on the high-resolution L1 multiplier, at
a position that is free from interference from 20NeH+. 22Ne
is measured at peak centre on the H1 Faraday cup (fitted with 1013Ω pre-amplifier); interference from CO22+ is corrected via
monitoring of the ratio of double- to single-charged CO2 in between
samples and measurement of CO2 during sample analysis, which we found
to be stable at 0.0437 ± 0.001 for our system throughout the period
for which the data we report here were obtained. The corresponding
corrections of 22Ne intensities are < 0.3 % for one shot of
RedAir calibration gas (∼ 17 fmol 22Ne). The
uncertainties of the correction are ∼ 2 %, which add
< 0.006 % uncertainty to the intensity determinations for
RedAir. These values scale linearly for smaller or larger amounts of
22Ne as found in samples. CO2+ is measured on the Faraday cup
of the Axial collector (fitted with 1013Ω pre-amplifier). We
refrain from analysing the larger neon beams (22Ne, 20Ne) on the
multipliers, since we found that they are a significant source of CO2
upon being hit by beams larger than those typical for 21Ne signals (for
analysing blanks, however, we use a multiplier for 20Ne and 22Ne).
Besides, the Faraday cups have a superior linearity and stability over time
(Wielandt and Storey, 2019). The mass spectrometer sensitivity,
mass discrimination and multiplier vs. Faraday gain are calibrated with
RedAir, which is measured at least once a day during sample runs. Each
batch of samples includes at least one measurement of ∼ 100 mg
CREU-1 (Vermeesch et al., 2015) to monitor the
performance of the extraction and purification system. We are in the process
of producing a new intercomparison material to replace CREU-1, whose
supplies are limited and eventually will run too low for regular use.
Performance
The within-run reproducibility of neon isotope ratios as determined for
calibration gas (RedAir, ∼ 17 fmol atmospheric Ne) is
similar for 21Ne /20Ne and 22Ne /20Ne ratios, with
0.46 % and 0.37 % (±1σ, n= 52), respectively. This
dispersion is larger than the uncertainty of individual measurements (Fig. 4); this feature, and the values for dispersion, are similar to those
reported for other Helix Plus instruments (Honda et al., 2015;
Wielandt and Storey, 2019). The second measurement period, with the
increased uncertainties of the 21Ne /20Ne ratios, was performed
after an extended period of development work for other noble gas isotopes.
We use the means and the uncertainty of the means of calibrations within
runs to calibrate the measurement samples, i.e. propagate the observed
dispersion in calculations of the abundance of cosmogenic 21Ne in
samples. Derived 21Ne /20Ne and 22Ne /20Ne ratios of 22
aliquots of CREU-1, including five power step extractions (Table 1), reveal
a spallation line of 1.078 ± 0.022 (±2σ), which is
indistinguishable from the published value of 1.108 ± 0.014
(±2σ; Vermeesch et al., 2015, Fig. 5). The calculated cosmogenic 21Ne abundances from 22 aliquots of
CREU-1 (Table 1) all agree within 2σ with their arithmetic mean (348 ± 10 × 106 atoms g-1; ±2σ); thus, we may calculate
an error-weighted mean: 348 ± 2 × 106 atoms g-1 (±2σ), which is indistinguishable from the published value
(348 ± 10 × 106 atoms g-1; Vermeesch et
al., 2015; see Fig. 6). We conclude that the reproducibility and accuracy of
the current set-up at the University of Cologne for determining cosmogenic
21Ne in quartz is similar to or better than those reported for other
laboratories worldwide (Vermeesch et al., 2015, Fig. 6; Farley et al.,
2020; Ma et al., 2015).
Conclusion
The performance of the set-up for neon isotope measurements in the new noble
gas laboratory at the University Cologne permits state-of-the art analysis
of cosmogenic neon. We now regularly perform analysis of samples for
cosmogenic neon for our running projects and are open to new scientific
collaborations.
Data availability
The authors confirm that the data supporting the findings of this study are available within the article (see Table 1).
Author contributions
TJD, BR and AV developed the Cologne noble gas system. TJD and BR performed the experiments and tests. BR and TJD wrote the paper.
Competing interests
The authors declare that they have no conflict of interest.
Disclaimer
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Acknowledgements
The equipment for the noble gas mass spectrometry laboratory described in
this paper was funded by the Deutsche Forschungsgemeinschaft (DFG) (project
number 259990027) through Tibor J. Dunai. The performance test was conducted and funded in the framework of the Collaborative Research Center 1211, Earth Evolution at the Dry Limit, Deutsche Forschungsgemeinschaft (DFG) (project number 268236062 – SFB 1211). Special thanks go to Dave Wanless for patient training and continuing support in mastering “Aura”. Furthermore, we want to thank Rainer Wieler and one anonymous reviewer for their constructive
feedback on the submitted manuscript.
Financial support
This research has been supported by the Deutsche Forschungsgemeinschaft (project nos. 259990027 and 268236062 – SFB 1211).
Review statement
This paper was edited by Cecile Gautheron and reviewed by Rainer Wieler and one anonymous referee.
ReferencesBinnie, S., Reicherter, K., Victor, P., González, G., Binnie, A.,
Niemann, K., Stuart, F., Lenting, C., Heinze, S., Freeman, S., and Dunai, T.
J.: The origins and implications of paleochannels in hyperarid, tectonically
active regions: The northern Atacama Desert, Chile, Global Planet.
Change, 185, 103083, 10.1016/j.gloplacha.2019.103083, 2020.Dunai, T. J.: Cosmogenic Nuclides: Principles, concepts and applications in
the Earth surface sciences, Cambridge University Press,
10.1017/CBO9780511804519, 2010.Dunai, T. J., Lopez, G. A. G., and Juez-Larre, J.: Oligocene-Miocene age of
aridity in the Atacama Desert revealed by exposure dating of
erosion-sensitive landforms, Geology, 33, 321–324, 10.1130/g21184.1,
2005.Eberhardt, P., Eugster, O., and Marti, K.: A redetermination of the isotopic
composition of atmospheric neon, Z. Naturforsch., 20a,
623–624, 10.1515/zna-1965-0420, 1965.Espanon, V. R., Honda, M., and Chivas, A. R.: Cosmogenic 3He and 21Ne surface exposure dating of young basalts from Southern Mendoza, Argentina, Quat. Geochronol., 19, 76–86, 10.1016/j.quageo.2013.09.002,
2014.Farley, K., Treffkorn, J., and Hamilton, D.: Isobar-free neon isotope
measurements of flux-fused potential reference minerals on a
Helix-MC-Plus10K mass spectrometer, Chem. Geol., 537, 119487,
10.1016/j.chemgeo.2020.119487, 2020.Gillen, D., Honda, M., Chivas, A. R., Yatsevich, I., Patterson, D. B., and
Carr, P. F.: Cosmogenic 21Ne exposure dating of young basaltic lava flows from the Newer Volcanic Province, western Victoria, Australia, Quat. Geochronol., 5, 1–9, 10.1016/j.quageo.2009.08.004, 2010.Györe, D., Tait, A., Hamilton, D., and Stuart, F. M.: The formation of
NeH+ in static vacuum mass spectrometers and re-determination of 21Ne /20Ne of air, Geochim. Cosmochim. Ac., 263, 1–12,
10.1016/j.gca.2019.07.059, 2019.Honda, M., Zhang, X., Phillips, D., Hamilton, D., Deerberg, M., and
Schwieters, J. B.: Redetermination of the 21Ne relative abundance of the
atmosphere, using a high resolution, multi-collector noble gas mass
spectrometer (HELIX-MC Plus), Int. J. Mass Spectrom.,
387, 1–7, 10.1016/j.ijms.2015.05.012, 2015.Kohl, C. and Nishiizumi, K.: Chemical isolation of quartz for measurement of
in-situ-produced cosmogenic nuclides, Geochim. Cosmochim. Ac., 56,
3583–3587, 10.1016/0016-7037(92)90401-4, 1992.Ma, Y., Wu, Y., Li, D., and Zheng, D.: Analytical procedure of neon
measurements on GV 5400 noble gas mass spectrometer and its evaluation by
quartz standard CREU-1, Int. J. Mass Spectrom., 380,
26–33, 10.1016/j.ijms.2015.03.004, 2015.Ma, Y., Wu, Y., Li, D., Zheng, D., Zheng, W., Zhang, H., Pang, J., and Wang,
Y.: Erosion rate in the Shapotou area, northwestern China, constrained by in
situ-produced cosmogenic 21Ne in long-exposed erosional surfaces, Quat. Geochronol., 31, 3–11, 10.1016/j.quageo.2015.10.001, 2016.
McIntyre, G., Brooks, C., Compston, W., and Turek, A.: The statistical
assessment of Rb-Sr isochrons, J. Geophys. Res., 71,
5459–5468, 10.1029/JZ071i022p05459, 1966.McPhillips, D., Hoke, G. D., Liu-Zeng, J., Bierman, P. R., Rood, D. H., and
Niedermann, S.: Dating the incision of the Yangtze River gorge at the First
Bend using three-nuclide burial ages, Geophys. Res. Lett., 43,
101–110, 10.1002/2015GL066780, 2016.
Niedermann, S.: Cosmic-ray-produced noble gases in terrestrial rocks: dating
tools for surface processes, Rev. Mineral. Geochem., 47,
731–784, 2002.Niedermann, S., Graf, T., Kim, J., Kohl, C., Marti, K., and Nishiizumi, K.:
Cosmic-ray-produced 21 Ne in terrestrial quartz: the neon inventory of
Sierra Nevada quartz separates, Earth Planet. Sc. Lett., 125,
341–355, 10.1016/0012-821X(94)90225-9, 1994.Ritter, B., Stuart, F. M., Binnie, S. A., Gerdes, A., Wennrich, V., and
Dunai, T. J.: Neogene fluvial landscape evolution in the hyperarid core of
the Atacama Desert, Sci. Rep., 8, 13952,
10.1038/s41598-018-32339-9, 2018.Saxton, J.: The 21Ne /20Ne ratio of atmospheric neon, J. Anal. Atom. Spectrom., 35, 943–952, 10.1039/D0JA00031K 2020.Vermeesch, P., Balco, G., Blard, P. H., Dunai, T. J., Kober, F., Niedermann,
S., Shuster, D. L., Strasky, S., Stuart, F. M., Wieler, R., and Zimmermann,
L.: Interlaboratory comparison of cosmogenic Ne-21 in quartz, Quat. Geochronol., 26, 20–28, 10.1016/j.quageo.2012.11.009, 2015.Wielandt, D. and Storey, M.: A new high precision determination of the
atmospheric 21Ne abundance, J. Anal. Atom. Spectrom., 34,
535–549, 10.1039/C8JA00336J, 2019.