Abstract

Noble gases are frequently probed for investigating fluid inclusions in minerals to unravel rock-forming processes through time. Over the last decades, heating and crushing have been the two main methods applied for noble gas extraction from fluid inclusions in ultrahigh vacuum (about 10-9 mBar). The heating of minerals or pieces of bulk rock causes the release of noble gases from both fluid inclusions and the mineral or rock matrix, the latter due to temperature-dependent mineral dehydration. Crushing of minerals only affects fluid inclusions and allows a release of noble gases at room temperature with minor contributions from the mineral matrix. This review describes different ultravacuum crushing techniques for noble gas analysis from fluid inclusions. It examines the technical details and operational conditions of each crushing system as well as methods to prepare samples prior to crushing. Crushing systems were found to have unique designs across the different laboratories reviewed; they include single or multiple sample loadings and manual, magnetic, or hydraulic operation of the crushing pistons. Due to the small amounts of noble gases released, the technology requires several mg to a few grams of rock material to achieve a measurement of all stable noble gas isotopes in a single fluid inclusion. While theoretically all stable noble gas isotopes are of interest, the elements and isotopes reported in different studies vary widely and reference materials as well as laboratory intercomparisons are lacking. The review includes applications on the origins of magmatic rocks and geochemical processes in the Earth’s mantle, the origin and chemical composition of deep crustal fluids and how these contribute to the formation of minerals of economic interest, and paleoclimate studies based on speleothems.

1. Introduction

Fluids trapped as inclusions within minerals can be millions to billions of years old and maintain a record of the fluid composition (e.g., halogens, noble gases, CO2, CH4, H2O, H2S, and H2) and environment at the time of mineral genesis (e.g., [1]). Such mineral fluid inclusions contain information about rock-forming processes through time and have been applied to study the genesis of ore deposits [26], mantle dynamics [7, 8], origin/features of extraterrestrial rocks [9, 10], surface processes [11], paleoclimate reconstruction [1218], and deep groundwater circulation research such as in the Paris Basin [19, 20] or in the Permian crystalline basement of Canada [21]. Due to the growing need to also understand deep groundwater systems, their provenance and residence times, and how groundwaters have been interacting with mineral phases over time, an increasing number of studies aimed at characterizing the degree of isolation of aquifers from other reservoirs (e.g., for hydrocarbon production), for applications such as CO2 geosequestration [22], waste disposal [23], or for tracing deep hydrogen and carbon cycles [24].

Noble gases enclosed in rock minerals provide important insights about different evolutionary stages of minerals or rock formations and of geological processes, such as pressure-temperature conditions during ore/rock formation, the nature, provenance, and evolutionary history as well as ages of geofluids [25]. 40Ar-39Ar chronology acts widely as a main application field (e.g., [26, 27]). The main noble gas isotope ratios used for the identification of fluid inclusion origins in rock material representative of different geoscientific study areas (e.g., “crustal rock studies,” “mantle rock studies,” “ore rock studies,” extraterrestrial rock studies,” or “deepwater studies”) are shown in Figure 1. Particularly He isotope ratios have been successfully used in many studies, due to He isotopes being preferentially measured in the earlier studies with the available analytical set-up. As a result, He isotope systematics are relatively well understood and have been described in many studies [2830]. The combination of noble gas analysis and halogens of irradiated samples has been successful in studying 40Ar/39Ar age systematics [31]. Analysis of halogens (Cl, Br, I) in fluid inclusions can be also accomplished by measuring irradiation-produced stable noble gas isotopes. The method is based on sample irradiation and neutron activation of halogens that produce noble gas isotopes such as 38ArCl, 80KrBr, 128XeI, or 39ArK as proxy isotopes for 40Ar-39Ar dating [32]. The main ultravacuum crushing method is made up of a Nupro valve type crusher (screw-type crushing systems) for noble gas analysis for 40Ar-39Ar dating and is often combined with heating experiments. The different crushing systems used in relation to the above-mentioned applications are summarized in Figure 2.

Fundamental for all applications of noble gas investigations in minerals and rocks is the knowledge about the characteristics of mineral fluid inclusions, and their geological context and how diffusion of noble gases within and between mineral phases occurs. Diffusion in minerals may occur within large mineral grains, along fast pathways (fractures), along the crystal lattice by Fickian diffusion [33, 34], and along intracrystalline defects such as dislocations [35]. Different noble gases exhibited different time scales of migration via these different pathways owing to the difference in noble gas atomic size. For instance, 40Ar has a lower diffusion coefficient in olivine (ca. m2/s at 673-1673°C [36]) compared to 3He (ca. m2/s at 553-893°C, [37]) due to the higher atomic size of argon (Ar: -10 m atomic size [38]) compared to helium (He:  m atomic size [38]).

The different diffusion rates of the noble gases can be used for characterizing different features of rocks and minerals. For instance, the modelling of 4He diffusion between pore-water and quartz has been used to infer the permeability of sedimentary rock formations with excellent sealing capacity [39]. Heavy noble gases like Ar and Xe have been studied to understand the original composition of the fluid inclusions [31]. Ratios of radiogenic (e.g., 4He) and primordial (e.g., 3He) isotopes of noble gases can further identify the mixing of mantle, crustal, atmospheric, or solar system endmembers in fluid inclusions. In this regard, Ne isotopes have been applied for clarifying the origin and evolution of minerals and rocks [40, 41]. Argon has been used for differentiating between atmospheric and mantle fluid origins as well as obtaining age information or halogen content [30], while Xe isotope patterns have been applied to identify very old fluids from mines several km deep [42].

Heating and mechanical crushing are the two most common approaches to extracting fluid inclusions from mineral grains for noble gas analysis. Less common is the extraction of noble gases by laser ablation. In comparison to heating, crushing is less efficient in breaking all fluid inclusions [43]. Heating of whole rock or mineral samples is generally applied until melting of the sample results in the release of gases from the rock/mineral matrix. A stepwise increase in the temperature may be used to release different generations of fluid inclusions. A detailed discussion on the performance of stepwise crushing is provided in Section 4.3. However, advantages of crushing measurements include lower blanks (i.e., lower background concentration) of the crushing apparatus, e.g., 10-100 times lower, and there are also less issues with in situ produced isotopes and adsorbed atmospheric gases [43, 44]. Heat may be applied to release fluid inclusions from shungite to study argon isotopes for age estimations [45]. The limitation of the heating technique is that it may induce noble gas fractionation and/or promote the diffusion of noble gases from within the mineral matrix, therefore complicating the analysis.

Several studies have combined both crushing and heating to compare the noble gas composition of the matrix and the fluid inclusions [7, 32, 4653] or to minimize diffusive gas loss from the small grains to ensure quantitative water extraction [15]. Further on, for He, crushing is supposed to be the prime method to extract mainly the magmatic He and less the in situ produced and cosmogenic He located in the crystal matrix [54]. Some minerals are unstable above ca. 800°C as, e.g., dolomite. Therefore, vacuum crushing is preferred over heating for noble gas extraction from fluid inclusions for most applications [55].

This paper reviews mineral crushing techniques for subsequent noble gas isotope analysis and their specific applications. The review focuses on mechanical crushing in ultrahigh vacuum conditions only. The crushing methods examined include magnetic crushing, piston crushing, and milling/shaking crushing. This review covers crushing methods of whole rocks and separate minerals for noble gas analysis of fluid inclusions that have been applied for the last fifty years. Alongside the development of noble gas release methods, the decades of application in the geosciences have been enabled by advancements in the noble gas measurement systems, such as high resolution and high sensitivity multicollector systems with automated detector movements. We summarize sample description and provide measurement specifications, which have been used across different applications.

2. Fluid Inclusion Genesis

Fluid inclusions are trapped fluids in a mineral’s crystal lattice, typically of heterogenous composition, often including gaseous, liquid, and solid phases [55]. The size of fluid inclusions varies from 1 nm [56] to tens to hundreds of micrometres but typically is no larger than a few millimetres [55]. According to the stage of crystal evolution, three different types of fluid inclusions are distinguishable (Figure 3): (i) fluid inclusions evolving during crystallization from melts (melt inclusions), (ii) fluids that are trapped in crystal overgrowths, and (iii) fluids that are trapped along a healing fracture zone within a crystal (secondary fluid inclusions) ([55, 57]). Primary inclusions is a term used for inclusions entrapped during and as a direct result of crystal growth. A fourth type are the so-called pseudo-secondary fluid inclusions: these can occur in mineral grains where mineral fractures originating from deformation processes have healed and included fluids present around the grains during the healing process [55]. Inclusions trapped during the growth of a mineral are most likely to be representative of the conditions that existed during mineral growth, while inclusions trapped after mineral growth are likely to record later, potentially different conditions. Unlike primary inclusions, the secondary ones contain fluids that existed at some time after the growth of the host crystal, possibly millions of years later.

Noble gases are investigated in fluid inclusions occurring in a variety of minerals, e.g., in olivine, pyroxene, amphibole, sulphides, or basaltic glasses; these are of interest when questions about magma origins and mantle contributions are investigated. Studies of crustal composition and origin have analysed fluid inclusions found in olivine, calcite, fluorite, eclogite, corundum, quartz, or whole rock pieces (e.g., [5861]). Common minerals relevant for noble gas analyses in fluid inclusions for ore studies are ore-associated minerals like quartz, pyrite, calcite, or fluorite (e.g., [3, 40, 53, 62]). Palaeoclimate focussed studies have relied on noble gases in cave rocks like stalagmites (e.g., [12, 14, 17]). Meteorites or lunar rocks have also been investigated using noble gases in fluid inclusions for, e.g., investigation of ancient atmospheric composition (e.g., [9, 10]).

Mineral or rock selection is therefore based on the availability of types of minerals in the geological environment being studied, but the selection also depends on how the noble gases are trapped in the minerals. Noble gases can be entrapped in the fluid inclusions during crystal growth (melt inclusion or primary inclusions) or in healed fractures (secondary fluid inclusions). Three processes for modifications of the original composition of noble gases in the fluid inclusions postentrapment have been described [31]: (i) by cosmogenic production in rocks exposed to the Earth surface, (ii) from the production of radiogenic noble gas isotopes from the decay of naturally occurring U, Th, or K, or (iii) from postentrapment leakage of helium via diffusion. Therefore, inferences on when and how the noble gases are trapped in the fluid can be derived from analysing the mineral fractions containing a single type of fluid inclusion or by sequential analyses (e.g., multiple crushing steps) of complex minerals. The key processes resulting in changes to noble gas composition are further discussed based on grouping from [31].

Cosmogenic production of rocks exposed to the Earth surface is important for 3He and 21Ne in terrestrial samples [11, 6366]. The principle is based on cosmic ray-induced in situ production of nuclides in rocks (and their fluid inclusions) exposed at the Earth’s surface and has been used to study erosion of rocks (e.g., [11]).

The radioactive decay of naturally occurring 238U and 232Th in minerals produces 4He, while 40Ar is produced from the radioactive decay of 40K. U, Th, and K can be found in the mineral matrix or within minor impurities [31]. Minerals with low U, Th, and K contents like quartz, calcite, olivine, or sulphide contain negligible radiogenically produced 4He and 40Ar [31]. 3He can be produced in situ by neutron capture by Li, for example, in Li-rich fluid inclusions in combination with neutron supply of U-rich host rocks [67]. This can be significant for fluid inclusions with low He, because production of 3He is very small, only, e.g., <10 3He atoms g-1a-1 in fluid inclusions where the rocks contain ca. 50 ppm Li and 10 ppm U [31, 67].

Leakage of noble gases is not uncommon in quartz crystals where anomalously low He/Ar ratios indicate a loss of He [31, 68, 69] or a difference in the age between fluid inclusion and rock [70]. Heavier noble gases (Ne, Ar, Kr, and Xe) are less diffusive and are less susceptible to being lost from the fluid inclusion after entrapment ([68, 69]). Many of the minerals which keep helium over geological time scales (107-109 years [31, 33]) are dense minerals like pyrite or arsenopyrite. This group of minerals is referred to as opaque minerals, which makes fluid inclusion studies by standard microscopy difficult. It has been argued that there are no petrographic characteristics for selecting minerals with high helium retention, where helium retention is possibly a measure of individual lattice dislocations intersecting fluid inclusions [31]. However, magnetite, fluorite, scheelite, sphalerite, chalcopyrite, and native gold are considered helium retentive [69, 7175]. Identification of any He loss in fluid inclusions is possible via analysis of all noble gasses and a comparison of the isotopic ratios against crustal production ratios, atmospheric ratios, mantle ratios, or solar system ratios.

3. Mineral Crushing Methods

Dependent on the intended geoscientific or industrial applications of fluid inclusion investigations, different types of crushing systems under vacuum have been used. Broadly three types are distinguished: (i) (electro) magnetic crushing, (ii) crushing by milling/shaking, and (iii) piston crushing (Figure 4). Each type can be further subdivided according to pressure strength, the degree of crushing automation, and whether single or multiple crushing containers are used.

3.1. Magnetic Crushing Systems
3.1.1. Manual Systems

The set-up of a magnetic crushing system consists generally of a stainless-steel tube (bottom closed) which keeps the sample at the bottom of the tube; a rod or ball inside the tube is moved magnetically in a manual way to crush the sample. These crushing systems are equipped with a 10 cm long tube connected to the measurement line and a pestle weight of 500 g to crush the sample with 50 steps [53]. Some systems use a 3 cm deep cylindrical stainless-steel cup, placed into the crusher covered with a 9 mm diameter disc [76].

For soft materials such as carbonate rocks from speleothems, rock fragments are crushed with a stainless-steel ball lifted manually with a magnet along a long slim vertical crushing container [14]. The sample container is directly connected to the noble gas mass spectrometer (MS) and sample preparation line (Figure 5).

Other crushing systems consist of 3 high-vacuum, nonmagnetic tubes connected to a noble gas MS with the capacity to crush 3 samples in a row [77]. A metal ball in each tube is operated manually with the use of a permanent magnet to crush the ≤3 g sample in a stepwise manner. The magnetic crusher is also applied at other institutes (see Table 1) to analyse noble gases in mantle rocks including the minerals olivine and pyroxene [78, 79]. This crushing system allows the application of multiple crushing steps ranging from 50 to a few thousand accumulated crushing steps. Mineral grain sizes range from 0.5 to 1.5 mm for quartz and fluorite, 2 to 3 mm for calcite and fluorite, and <10 mm for olive-containing xenoliths (Table 1).

3.1.2. Electromagnetic Systems

Electromagnetic vacuum crushing systems are in use for both soft and hard minerals and operate while fully connected to the measurement device (Table 2). It involves moving a nickel bolt up and down a cylindric crusher container in a magnetic field. Examples of this system can be found at the University of Oxford where ca. 2 g of sample material are crushed with an electromagnetically lifted nickel bolt with adjustable frequency (personal communication Dr. Rosie Jones, University of Oxford) (Figure 6).

Modifications of this set up include the use of a tungsten carbide piston with a nickel guide [84]. The magnetic field is induced in three water-cooled solenoid coils attached to the crusher container creating an operating current of 30 amperes with ca. 300 strokes/min of the piston [84]. Before starting the experiments, this crusher is operated sample-free for ca. 30 min [84]. A modification of this device was used in [85], which includes the possibility to separate the solenoid coils from the crushing device thus allowing to connect numerous crushers for running multiple crush experiments.

Another modification was introduced by [47], who used a stainless-steel electromagnetic crusher with a crushing rod weight of 80 g and a crushing tube of 20 cm height.

Further details about the operating parameters of an electromagnetic crusher can be found in [86]. Their five computer-controlled solenoids operate in sequence to lift the magnetic crushing rod with a frequency of 1 stroke/2 sec and an energy impact release of ca. 0.4 Joule/stroke. The duration of each crushing cycle and the number of crushing steps can be adjusted to the requirements of the experiment. Crushing steps up to 100-300 strokes were used on 0.5 g of basalt samples ([7]). Multiple crushing steps were also used by [87] to maximise the release of noble gases.

The number of steps may be adjusted depending on the mineral type, e.g., for carbonates a sequence of three steps with 15, 300, and +300 strokes were used while for apatite 15, 500 and +500 strokes were used [88]. The first step is typically applied to release the gases from the larger inclusions with possibly a larger fraction of atmospheric gases; the second step delivers the main extraction while completion is achieved in the third step.

Crushing experiments have been performed by activating the bolt for 3 to 8 min (ca. 60 drops/min) with sample amounts not more than 700 mg [54]. Higher sample amounts were shown to become compacted while crushing, rendering the experiment less reliable. This is due to the dimension of the crusher; a larger sample amount would compact easier with the less efficient crushing of the grains as only part of the grains is able to get crushed in the confined space of the crusher. Using a smaller amount of sample provides more space to crush grains owing to the larger space between the grains.

3.2. Milling and Shaking Crushing Systems

Several off-line high-vacuum crushing systems have been developed over the years, often based on the principle of crushing rock or mineral samples in sealed glass (Table 3). The design used by [70, 95] is based on a steel roller mill with 3 mm diameter iron balls or a disc mill. Generally, the glass ampoules are sealed with the sample inside and steal rollers (6 mm length, 4 mm in diameter) or a disc mill to crush the sample by vibration and rotation [70, 96] for between 20 to 30 minutes [70, 97].

The shaking-based systems require a stainless-steel container, a crush rod, and two high-vacuum, all-metal valves. The system is operated by shaking the unit for less than 5 minutes for quartz mineral grains [60]. Mineral fragments of 0.7 to 3.1 g can produce a grain size of 0.5 to 1 mm after crushing [60].

3.3. Piston-Style Crushing Systems
3.3.1. Hydraulic Press Systems

Piston-style crushers are connected to a hydraulic press to crush in manual or automated mode (for a summary of studies, see Table 4). They have been applied to various minerals, including olivine, pyrite, quartz, or sedimentary rocks. At the University of Oxford (Figure 7 and Table 4), the system consists of one stainless steel sample container and 2 sets of tungsten carbide sample plates for processing 1 g samples in between each plate pair (personal communication Dr. Rosie Jones, University of Oxford). The crusher is driven by a manual 10 ton press (ca. 1544 bar), and it is noted that 8 tons (ca. 1235 bar) is generally sufficient to crush most minerals. While crushing, the system is permanently connected to the measurement unit. A similar experimental setup was reported by [85], who used pressures up to 700 bar delivered by a hand pump.

The crusher systems reported by [95, 98, 99] are all variations of the same basic design, with differences in maximum applicable pressure (Table 4). For example, one crusher uses a stainless-steel system connected to an automated hydraulic press with adjustable pressures up to 200 bar [95]. The sample container can be filled with ca. 1 g sample material of similar grain size between 0.5 and 2 mm, with pressures applied vertically via a piston. Another crusher uses single-step systems with a stainless-steel bowl in ultrahigh vacuum conditions which can apply up to 2000 bar pressure [98]. This crushing system is capable to process six samples simultaneously [61, 100, 101]. [99, 102] used a combination of a 1 cm diameter and a 6 to 10 cm long stainless-steel sampler, a hydraulic piston with applied pressure up to 200 bar, and a stainless-steel plate (AISI 316 LRN) placed below the sampler. Flexible stainless-steel tubes, heated to about 200°C with heater tape to prevent gas adsorption and improve gas flow in the system, connected the sample holder with the purification line. Purification steps were build-in to separate the CO2 from the He fraction in a glass sampler; the He trapped in the glass sampler was transferred to a mass spectrometer for measurement. A single-step crushing device that uses a hydraulic press to crush samples placed in an aluminium holder separated by an aluminium plate at about 138 bar pressure was reported by [103, 104].

One of only a few automated hydraulic press systems was reported by [115]. The automated press operates with a sample chamber which includes 6 stainless steel sample containers, each with their own little hammer (Figure 8). In any given measurement sequence, one sample container is used as a blank sample. The automated system is connected to a piston, which can be moved via a turn table to the sample container with the hammer to crush a sample amount of 0.4-0.7 g. Depending on the mineral hardness, crushing steps will be adapted, e.g., using 300, 500, or 800 steps.

3.3.2. Screw-Type Crusher

Manually driven screw-type stainless-steel crushers with a single sample chamber have been used for both hard and soft rocks [9, 11] (Figure 9 and Table 5). These systems use a single piston which can be rotated ([9]) or pressed ([11]) to crush the samples. In [9], the crusher has a set of modified high-vacuum angle valves (CF16 flanges) and a rotation system inside for manually adjusting the piston up and down. This allows for controlling the crushing intensities, particularly for softer samples. Multiple crushing steps consist of slowly pressing first until the first cracks appear, then repeating the crushing until the cracking sound of the sample has stopped [9]. The first crushing step delivers most of the gases, i.e., primarily those residing within vesicles. The contribution of gases from the matrix is only obtained by applying another method, i.e. using CO2 laser heating extraction.

With screw-type crushers, the pressure to crush can be increased in a stepwise fashion to crack the fluid inclusions. Some screw-type crushers have the capability to remove samples from the base plate without venting the system (Figure 9) and with the attachment of a PYREX™ sample carrier that can load up to 6 samples [116]. Special alloys are used for the front part of the piston like W-T-Co alloy to avoid cracking of the piston by the sample [117].

Sample amounts used vary from 0.4 to 0.7 g [9] to 1 to 1.5 g [11]. The crushing efficiency of an ultravacuum piston crusher is managed by a cylinder container instead of using bellows; this way samples of up to 1 g (2-3 mm grains) can be placed on a stainless-steel disc [118120].

Modified Nupro valves® (Figure 10 and Table 5) are applied in several studies for crushing minerals. [122], for instance, used a manually driven crushing system consisting of sample chambers as part of the bottom part of the Nupro valves®, whereas the upper part includes the manually driven piston. Advantages of this type of crusher include the ability to crush up to 5 samples consecutively without opening the system (independent of mineral hardness), and more crucibles can be added for individual experiments. Rocks or minerals are crushed that vary in size from a single grain (e.g., [28]) up to a sample size of 1 g (e.g., [3, 4]).

4. Crusher Performance

The performance of crushers can be expressed in several ways. Here, we distinguish (1) crushing efficiency, (2) blank levels and noble gas adsorption to experimental equipment, (3) the number of crushing steps, and (4) sample preparation steps.

A comparison of features of the different crusher types discussed in previous sections is available in Table 6. Most of the crusher types are directly connected to the gas separation and measurement system (“online” crusher), whereas “offline” crushing systems are operated disconnected from the measurement system. The “offline” systems first collect their gases into containers like the stainless tubes of a magnetic crusher, stainless steel vessels of a shaking crusher, or milling crushers with glass ampoule crushers; subsequently, gases from the containers are released into the measurement system.

4.1. Crushing Efficiency

The noble gas concentrations obtained from crushing experiments may be influenced by the different sample manipulations such as further grain-size reduction due to chiselling out crushed samples from sample holders, aggregates lumping together and not passing a sieve, or loss of noble gases from rupturing of large (e.g., >75 μm) fluid inclusions [133, 137].

To check the efficiency of crushing, some studies have analysed grain size distributions after crushing (Table 6) resulting in values between <630 and <200 μm for speleothems [14], <1500 μm for xenoliths [87], <150 μm for olivines [85], and<149 μm for pyrite [6]. Others have continued the crushing until no audible crushing sound can be heard in an attempt to reach a maximum extent of crushing [104]; alternatively, crushing experiments can be applied multiple times for improved efficiency [135].

Efficiency testing of crushing by means of scanning microscope images and computer analysis to calculate the crushed size fractions was done by [70] for fractions between 0.01 and 1 mm applied by a glass ampoule crusher. Others used a Beckman Coulter laser diffraction size analyzer for analysing the grain size distribution after crushing [90].

Using different amounts of quartz in the crushing experiments of [125] did not influence efficiency as all tests showed the same fraction (i.e., 60%) of grains <90 μm. Contrasting results were reported by [54]: a larger amount of sample resulted in a lower efficiency caused by crushed fine grains compacting the sample in the crusher. The use of different crushers can influence efficiency, as was demonstrated by [90]: their piston-style crusher (with a modified ¾ Varian vacuum valve) produced coarser powders compared to a magnetic crusher able to exert a greater force.

Primary fluid inclusions smaller than 5 μm may remain in crushed minerals when their size exceeds that of the inclusion; as a result, noble gases in such small inclusions may be underrepresented in the analysis. Note that inclusions may come in all sizes and shapes, but commonly, many inclusions in a sample may be smaller than a micrometer [138]. Furthermore, in most samples, the secondary inclusions are much more abundant than the primary inclusions [139], and, importantly, fluids from the secondary inclusions distributed along microcracks are readily extracted by crushing [140]. This points to one of the limitations of the mineral crushing technique compared to other techniques such as heating. Especially those crushing techniques that generate very large grain sizes (Table 6) result in low efficiency in liberating fluids from primary inclusions.

The choice between heating and crushing also depends on whether primary or secondary inclusions are the target. As primary and secondary fluid inclusions have different size distributions, crushing and/or heating methods that have a different size cut-off will result in different fractions of primary vs. secondary fluid inclusions. Unlike crushing, heating would provide results that are more easily comparable between different studies because it does less strongly discriminate between different sizes, i.e., one always gets the same mix between primary and secondary inclusions. To improve the intercomparison between fluid inclusion studies, we recommend that any crushing study reports at least a best estimate of the grain size after crushing (and ideally a distribution rather than a simple cut-off). An example of postcrushing particle size data is available from [141] who showed that crushing quartz grains by a total of 10,245 pestle drops resulted in 6% of grains <1 μm, 82% between 4 and 1 μm, and 12% of the grains were >4 μm. The smallest grains were approximately 50 nm in size. As most of the primary fluid inclusions were 5–20 μm, their crusher effectively extracted the majority of the fluid inclusions visible with an optical microscope.

4.2. Blank Levels and Adsorption

Blank levels refer not only to the crushing devices but also refer to the blank level of the noble gas mass spectrometer (Tables 7 and 8). Indeed, the adsorption of gases released from the crushing device show temperature and pressure-dependent adsorption onto components of the measurement device, which impacts the blank level and can influence the analysis of mobile noble gases or reactive gases like N2, CO2, or H2O. The use of glass equipment in the extraction line of a mass spectrometer is less reliable for He analysis; Pyrex ® glass had higher He blanks in comparison to stainless steel, while Kovar ® glass had a 7 times smaller He blank caused by He diffusion through glass in comparison to Pyrex ® glass [79]. The ability for crushed minerals to adsorb gases is known to occur on basalt glass and quartz surfaces with the order of adsorption (strong to weak) N2<CH4<CO<CO2< H2O [142], and on meteorites with stronger adsorption towards heavier noble gases leading to fractionation at low temperatures [143].

To avoid unwanted background levels of reactive gases, gentle crushing was shown to produce particularly low methane background levels in a steel crusher, while crushing laboratory glass allowed the detection of the background of adsorbed gases [121]. The background of adsorbed gases can also be minimized in stainless steel crushers and its tubing by baking the crusher assembly up to 80°C [116] because gas desorption is promoted at higher temperature [118]. Using an electromagnetic crusher, no adsorption of CO2 or CH4 was detected when operating an empty device [142, 143].

Fluid inclusion-free samples have been used to determine active blanks [43, 69]. Using inclusion-free quartz resulted in ±50% error after repeated crushing experiments, not significantly higher than passive blanks from gas buildup during extraction and clean-up measurements with a screw-type crusher [43, 116]. Some types of crushers have been reported to result in low blanks, such as manual piston-style crushers in comparison with manual magnetic crushers, owing to the improved control of crushing and therefore softer crushing without wall grinding [120].

A limited comparison between electromagnetic and piston-style crushers has delivered some insights regarding their performance. [85] describes both crusher types: their hydraulic-type crusher uses up to 3 g of sample with an average grain size larger than 250 μm which is mechanically crushed using a hydraulic press of up to 70 MPa. This results in approximately 50% of the sample being crushed into powder with an average grain size of less than 150 μm. Their electromagnetic solenoid-type crusher can load up to 1 g of sample material into the crushing chamber; more than 70% of the sample is crushed into less than 150 μm after 1000-2000 strokes. Other differences between both crushing methods concern the blank levels of helium: these are  cm3 STP for the hydraulic crusher and  cm3 STP for the solenoid crusher. Regarding the fraction of noble gases released, both systems seem to perform in a similar way. The electromagnetic crusher primarily releases noble gases from the microinclusions and minimizes release from matrix-hosted noble gases retained within solid phases [144]. Likewise, the piston-style system with single-step crushing also minimizes the release of matrix-sited noble gases [100]. Some vacuum crushers can produce very fine grains; e.g., for diamond crushing, the electromagnetic crusher was shown to produce very fine grains <100 nm, while some larger grains (up to 5 μm) [145].

4.3. Number of Crushing Steps

The application of multiple crushing steps to increase the efficiency of gas release has been applied in several studies for most crusher types. Initial crushes generally resulted in the highest volume of noble gases, with typically >80% of the helium being released by the application of an electromagnetic crushing device [54]. Especially stepwise release enables the stepwise extraction of noble gases and avoids the release of gases stored in the mineral matrix [78, 80]. It has been argued that extraction progresses from the largest to the smallest fluid inclusions and from dense fluid inclusion assemblages to less dense fluid inclusions [80]. Similar conclusions about the ability of stepwise crushing of magmatic minerals to release noble gases from two different gas reservoirs, i.e., primary (early release after ~0.5-~3 minutes) and secondary inclusions together with noble gases associated with the matrix (released after ~7–~13 minutes), can be found in [86].

The validity of the 40Ar/39Ar stepwise crushing technique for age dating of cassiterite and wolframite mineralisation was demonstrated by comparison with contemporaneous K-rich muscovite ages determined by 40Ar/39Ar laser stepwise heating [141]. The fluid inclusions in cassiterite and wolframite were extracted by stepwise crushing using a total of 18 steps, taking care of the number of pestle drops for each successive step increased to obtain argon levels conducive to precise measurements. The argon release pattern by progressive crushing revealed three broad types of argon reservoirs, similar to previous findings by [147]: (1) excess 40Ar within secondary fluid inclusions typically distributed along cracks and therefore more easily extracted during the first couple of crushing steps (6 steps for cassiterite and 6-7 steps for wolframite); and (2) a mixture of radiogenic and trapped argon in small primary fluid inclusions and atmospheric argon from the crusher released during the final crushing steps. A third type contains a mixture of primary and secondary reservoirs. Importantly, all the cassiterite and wolframite samples dated by the stepwise crushing method resulted in isochron ages in agreement with those of the muscovite samples derived from laser stepwise heating. [141] suggest that the progressive crushing technique is suitable to directly determine the mineralisation ages of several ore-minerals, including cassiterite, wolframite, sphalerite, chalcopyrite, and pyrite, on the proviso that they contain sufficient primary fluid inclusions with adequate potassium concentrations (or salinity).

In [148], the gas release sequence from stepwise crushing of samples from granite and greisen (a highly altered granitic rock) was further refined by introducing six groups of gas reservoirs, here ordered according to increasing crushing steps: (1) microcracks with excess 40Ar and an extremely low amount of potassium, (2) secondary inclusions with excess 40Ar distributed along healed microcracks suitable for derivation of an inverse isochrone line typical of postmineralisation activities, (3) mixtures from the secondary and primary inclusions, (4) isolated, micrometer-sized primary inclusions suitable to derive an inverse isochron line for age dating of the primary mineralisation, (5) mixtures from secondary and primary inclusions representing a gas mixing line from fluids and solids, and (6) micro- to nanometer-sized solid minerals released by extended crushing. [149] commented that not all the above groups would necessarily be observed for any specific sample. The 40Ar/39Ar progressive crushing methodology resulted in primary inclusion isochron ages that agreed with the mica ages from laser step heating and cassiterite U-Pb ages using laser ablation-inductively coupled plasma-mass spectrometry. This demonstrates that the progressive crushing technique combined with multi-collector noble gas mass spectrometry of 40Ar/39Ar is a valid methodology to derive hydrothermal mineralization ages to date geologic events associated with K-bearing fluids.

4.4. Sample Preparation before Crushing

Sample preparation for noble gas extraction from fluid inclusions depends on the type of geological material that needs investigation: mineral veins, volcanic rocks, or sedimentary formations. The first step involves coarse crushing of the rock specimen followed by handpicking of minerals under a binocular microscope (e.g., [53, 62, 76, 120]). Prior to mineral selection under the microscope, crushed materials are sieved to a grain size fraction of 0.5-2 mm ([53, 76]) or >315 μm [59]. What minerals are selected depends on the type of fluid inclusions of interest; e.g., olivine is selected for their large fluid inclusions that are known to contain higher amounts of He [28].

A variety of chemical treatments exist to remove unwanted compounds from mineral surfaces. For instance, quartz clasts are leached with HCl at 80°C for 4-5 hours, followed by three leaching steps with HF/HNO3 at 80°C for 5-10 hours in an ultrasonic bath for etching several micrometres of the quartz surfaces [11]. Megacrystal separates of volcanic rocks like corundum, titanomagnetite, pyroxenite, lherzolite, and wehrlite are cleaned with 5% HNO3 ultrasonically for 20 min for removal of post eruption radiogenic components and then rinsed with distilled water and dried [103]. Olivine, amphibole, and whole rock separates are leached with 7% HF for 10 min and washed with distilled water [94].

Some of the unwanted compounds at the mineral surfaces include atmospheric gases. Desorption of such compounds may require distilled water and organic solvents like acetone ([3, 7, 62, 90, 120]), ethyl-alcohol [132], ethanol ([58, 94]), or methanol in an ultrasonic bath.

Removal of unwanted compounds from mineral surfaces can also be achieved by heating the samples when loaded in the crusher container, at which time also gases adsorbed onto the inside of the crusher equipment are also removed. Heating temperature and duration vary: 70°C for at least 8 hours for stalagmites [14], 100°C for 24 hours for sphalerite [133], 100°C for 8 hours for fresh volcanic glass chips [120], 150°C for chalcopyrite, bornite, and galena [133], 150°C for 24 hours for basalts [7], and ca. 200°C for 48 hours for serpentine and chlorite [149].

5. Example Method Studies

Method studies involve a detailed investigation of fluid inclusions in different rocks or minerals, or they compare noble gases released from crushing or heating with the aim of assessing the quality of vacuum crushing for different mineral types. Manual piston-style crushers (partly with manual hydraulic press systems) were applied in some studies, while other studies used manual screw-type crusher, magnetic crusher, or electromagnetic crusher. Examples of method studies are briefly discussed.

Significant improvements in helium analysis may be obtained by applying a double collector system and two connectable crusher systems [85]. The combined system included a hydraulic press-type crusher and a solenoid crusher system connected to a high-vacuum heating system. All systems showed low He blanks and suitable natural samples with low noble gas concentrations or total gas amounts.

The effect of prolonged crushing on the validity and accuracy of helium concentrations and isotope ratios was demonstrated for magmatic crystals subjected to an electromagnetic crusher ([86]). Long crushing times revealed two exponential releases linked to two distinct gas reservoirs. Gases from the first reservoir were released during the first 0.5 to 3 min of crushing, while gases from the subsequent 7-13 min were linked to a secondary He reservoir. Because extended crushing may lead to a possible mix of released He from both primary and secondary inclusions or matrix-stored He, only the initial short-duration crushing steps should be considered to avoid such mixing.

To investigate irradiation-produced noble gas isotopes in scapolite and hydrothermal quartz, two crusher types, a Nupro valve crusher and a hydraulic press, were compared (e.g., [27]). Crushing the scapolite released ca. 4% of the irradiation-produced 38ArCl in comparison to 98% released for quartz. Ar was found to be preferentially released over Kr and Xe which were believed to be stored in the mineral lattice; up to 88% of the lattice-hosted noble gases were released by prolonged crushing. For obtaining Br/Cl, I/Cl, and 40Ar/39Ar age systematics, longer crushing times were found to be an effective method [27].

The relationship between paleocrustal fluids and contemporary groundwaters was investigated based on noble gas ratios of 40Ar/36Ar, 3He/4He, 40Ar/4He in quartz, and fluorite from the North Pennines ore field (England) [69]. Results indicated the fluid inclusions were characterized by air-saturated water patterns with partly unfractionated atmospheric gases in minerals that had precipitated under atmospheric conditions. Minerals of the North Pennines ore field were identified as crustal He based on 3He/4He ratios, while crustal radiogenic production of 40Ar resulted in elevated 40Ar/36Ar ratios above atmospheric. While up to five crushes were applied to olivine, the main release of 4He and 40Ar occurred at the first crush and then decreased with further crush steps. At the same time, an increase in the 40Ar/36Ar ratio with each crush step was interpreted as a contribution of atmospheric argon from secondary inclusions.

Finally, although a summary of noble gas characteristics in different layers of the crust and mantle would be very interesting, this was beyond the scope of the current review of crushing techniques. Note that a summary of noble gas isotopes in different rock types of mantle or crustal origin is given in Figures 1 and 2. Readers may find further information on noble gas isotope patterns in mantle and crust rocks in [150152].

6. Summary and Conclusions

This review of crushing techniques identified three main ways of ultrahigh vacuum crushing systems: magnetic crusher, shaking- or milling-type crusher, and piston-type crusher. Several modifications of these three crusher types were reported and involved automatic handling, the use of multiple crushing containers, and adjustable pressure capabilities. Piston-style crushers were shown to be applicable to most if not all geoscientific studies reviewed, while magnetic crushers were primarily used on crustal rock for the purpose of mantle studies. A wide range of noble gas isotopes has been analysed following mineral crushing, with the selection of isotopes not limited by the crushing technique but rather depending on the research question at hand and the available analytical measurement technique. The crushing techniques reviewed had been used for a wide range of geoscientific applications, including the characterization of ore evolvement, mantle-crustal identification, paleoclimate estimations, characterization of surface and erosion processes, and the identification of extraterrestrial rocks or deep fluids in Earth’s crust.

An important limitation of all crushing techniques reviewed is their inability to separate inclusions embedded in fracture infillings from inclusions within the mineral matrix. For example, crushing of quartz grains typically releases noble gases from both fluid inclusions of different generations (melt inclusions, primary and secondary inclusions) and to a minor degree the matrix. Unless one can separate those, any mineral/rock age determination will be a based on a mixture of those three sources of noble gases.

This review has shown that there is a lack of technical details of crushing conditions before and after the experiments as well as comparisons to reference systems. Intercomparison of results is therefore difficult because every laboratory has an almost unique approach to undertaking crushing experiments which involves particular modifications resulting in custom-made crushing systems. Key recommendations for future studies are (1) undertake laboratory intercomparison studies to evaluate how comparable the data are that have used different crusher types and (2) explore how selectively measuring larger versus smaller fluid inclusions affects the results, which is especially important for crushing as the smallest fluid inclusions are not typically crushed.

Data Availability

Data can be obtained from this link https://data.csiro.au/collection/csiro%3A58349v1.

Conflicts of Interest

The authors declare no conflict of interests for this submission.

Acknowledgments

This study is funded by the Environment Business Unit of CSIRO, Australia’s National Science Agency. Open Access funding enabled and organized by the CSIRO 2023.