Multi-extreme conditions at the Second Target Station

Materials are now routinely subjected to extreme conditions of pressure, magnetic field, and/or temperature. Their properties under extremes can be probed on an atomic level using in situ neutron scattering to provide unique insights into structural properties and phase diagrams, electron and phonon behaviors, magnetic phase diagrams, and the formation of exotic matters of state even far from equilibrium in a material’s energy landscape. To accommodate these studies, extensive efforts are made at neutron sources to develop extreme environments coupled with the appropriate instrumentation. While this has often focused on adapting one individual extreme to in situ neutron scattering, there have also been significant efforts devoted to the development of multi-extremes, that is, the simultaneous application of two or more extreme conditions.

Complementary development to those efforts at neutron sources has also been made at synchrotron sources. While the same extremes are addressed at synchrotrons, the inherent differences between neutrons and synchrotron x rays are also often reflected in the extreme environment. This is particularly evident for applications where the different scattering cross-sections are relevant such as those that require small sample volumes. With the advent of higher-flux neutron sources, some gaps are, however, closing. It is thus useful to consider both synchrotron and neutron techniques as many environments may become transferrable. While this present work focuses on novel multi-extreme environments for neutron scattering, a short overview of (multi-)extremes at both neutron and synchrotron sources is given first.

Maximum high temperatures at both neutron and synchrotron sources are typically achieved using levitators with temperatures up to ∼3000 K possible.^1–3 Temperature limits are often imposed by sample evaporation, and surface laser heating often results in large thermal gradients, favoring the rapid acquisition times of synchrotron sources. In contrast, (ultra-)low temperatures are usually achieved in commercial cryostats down to 1.4 K and to sub-Kelvin levels in dilution refrigerators, attaining ∼300 mK at synchrotrons and reaching just a few mK at neutron sources. Achieving these ultra-low temperatures is typically easier at neutron sources since beam heating is an issue for synchrotron x rays, particularly for high energy x rays.

Furthermore, at synchrotrons, high magnetic fields are typically applied via portable, miniature pulsed magnetic field systems, up to 30–40 T.^4–7 Similar pulsed high magnetic fields are also possible at neutron sources, 30 T at the Japan Proton Accelerator Research Complex (J-PARC) (e.g., Ref. 8) and the Spallation Neutron Source (SNS)^9,10 and 40 T at the Institut Laue-Langevin (ILL) for single crystal diffraction.¹¹ A facility scale, series connected hybrid magnet provided DC high magnetic fields up to 26 T at the EXED beamline of Berlin’s Helmholtz Zentrum¹² until that facility closed in 2019.

For high pressure, large differences exist since high pressures require excessively small samples. At synchrotrons, ultra-high pressure is routinely applied in the megabar regime when using a diamond-anvil cell (DAC)¹³ with record ultra-high pressures of 600 GPa achieved.^14–16 Due to the large sample volume requirements, neutron scattering was limited to just a few GPa until the development of Paris-Edinburgh presses allowed routine pressures of above 20 GPa.¹⁷ More recently, neutron DACs have enabled pressures close to 100 GPa at the Spallation Neutron Source (SNS) and J-PARC^18,19 with development tackling the megabar (=100 GPa) barrier.

Differences in the application of multi-extremes are equally pronounced. For example, laser-heating of the DAC is routine at many synchrotrons (e.g., Refs. 20–22) and allows simultaneous high pressure, high temperature of over 4000 K, above 1 Mbar. In contrast, laser-heating has not yet been adapted for neutron DACs and simultaneous high pressure, high temperature above 10 GPa is typically achieved in multi-anvil assemblies.^23,24 Simultaneous high pressure, low temperature is more similar, with DAC studies to ∼5 K routinely possible at both type sources. However, ongoing developments at the SNS aim to adapt neutron DACs to dilution refrigerators for pressures of ∼10 GPa at sub-Kelvin temperatures. As DAC studies require hard x rays, synchrotron studies at high pressure, low temperature are limited to ∼1.5 K due to beam heating.

Furthermore, the combination of low temperatures and high magnetic fields allows for “cryomagnets” that achieve, for example, 26 T at 0.65 K¹² while synchrotron magnets are currently limited to ∼1.5 K.⁷ Interestingly, combining pressure, magnetic field, and low temperature, a combination that is critically important to many quantum materials studies, proves very difficult. At neutron sources, pressure and high field studies have been limited to 2–4 GPa^25–29 while synchrotron studies in DACs have been limited to a few T. Further work is required to achieve the same maximum in conditions as possible for individual extremes. Similarly, the combination of high temperature and high magnetic fields has not yet been explored in-depth, let alone the combination of the three extremes of high pressure, high temperature, and high fields.

While the scientific motivation for certain combined extreme conditions is described in Sec. II, for neutron scattering, a major factor inhibiting development is the limited brightness of the sample. The design and construction of a new and innovative large-scale facility with improved beam characteristics removes this limiting factor and motivates development. The Second Target Station (STS) to-be-constructed next to the First Target Station (FTS) of the Spallation Neutron Source at Oak Ridge National Laboratory (ORNL) will use high brightness cold neutrons and new focusing technology to enable science on very small samples, including samples smaller than currently measurable at the FTS. This focus on small samples is very useful for the development of multi-extremes and for the necessary co-optimization of various extreme environments.

While neutron source improvements³⁰ and new instrument concepts that leverage improved optics^31–35 are described in more detail elsewhere, here we summarize the improved beam characteristics available at the STS. Comparing the STS and FTS cold sources at 3 Å, the time-averaged brightness is ∼4 to 5 times higher and the peak brightness is ∼25 times higher.³⁶ The in-progress “Proton Power Upgrade (PPU) Project”³⁷ will enable the delivery of 700 kW at 15 Hz to the STS’s water-cooled, rotating (40 rpm), solid tungsten target.^36,38 Such a solid tungsten target can accommodate a higher proton intensity than the FTS target. Thus, the same proton power per pulse (∼46 kJ) and time profile (<1 µs long), as used for the FTS, will be delivered to a smaller area of the target (∼65 cm²) at the STS. Overall, this will provide a more localized and more intense fast neutron source per pulse than the FTS. Furthermore, liquid hydrogen moderators in a parahydrogen state will be positioned closer to the target to accept more spalled high-energy neutrons. These will also employ a more compact geometry to provide a brighter source in the cold neutron wavelength range for instrument guide optics.

New technologies available to the STS instrument suite itself also enhance the brightness of the sample. The availability of higher-m neutron guide mirrors, along with higher brightness sources, enables a new emphasis on transferring brightness instead of intensity in designing STS instrument optics. This allows for high-intensity in a small area together with low intensity surrounding that area, which enables significantly enhanced signal and signal-to-noise for small samples. The lower spallation repetition rate of 15 Hz compared to (45 pulses of the) 60 Hz FTS also provides a ∼4× greater bandwidth for a given instrument geometry and is well matched to the cold neutron wavelength range of the STS, providing simultaneous measurement across the larger length and/or energy scales. These characteristics and capabilities also open exciting new possibilities for multi-extreme conditions and new extreme environments.

As part of this Topic Issue on “New Science Opportunities at the Spallation Neutron Source Second Target Station,” we present targeted development directions for multi-extremes at the STS. Specifically, Sec. II will introduce a concept for a new magnet friendly neutron diamond cell, a concept for laser-heating of a neutron diamond cell, and a concept for levitation at moderate gas pressures. Section III of this paper will address how these multi-extremes could be accommodated by the current first instrument suite of the STS, by some FTS instruments, and by possible future STS instrument concepts specifically focused on co-optimization for multi-extreme conditions.

Enabling in situ neutron scattering under multi-extremes requires not only adaption/optimization to neutron instruments but also co-optimization of the extreme environments themselves. The needs for such co-optimization vary depending on the exact combination of extremes, ranging from material and geometry optimization to adapting instrumentation methods, such as the cinematic/stroboscopic mode. To address future scientific questions in a diverse range of fields, such as quantum materials, materials physics and engineering, high pressure chemistry, and geosciences, a range of multi-extreme environments will be required. Three such concepts optimized for use with neutron instruments at the STS are described here.

A particular focus is the combination of high pressure and high magnetic fields. This combination gives a unique insight into the fundamental thermodynamics of materials as well as allows for control of the state of materials. This enables unique studies in a wide range of condensed matter physics and materials science topics ranging from those focusing on spin dynamics in frustrated systems, the nature and formation of magnetic skyrmions, and superconductivity, to controlling phonons for phonon engineering and understanding spin–phonon interactions.

Specifically, understanding and controlling the dynamics of strongly correlated electrons in quantum materials is essential for their energy and informatics applications.⁴⁰ The characteristic fields for quantum phase transition in such materials are often very strong and beyond the capability of current neutron scattering extreme environments.⁴¹ Similarly, the pressure scale required for such quantum phase transition is also significant. The combination of high pressure and magnetic field will provide unique tools to enable neutron scattering to probe the non-thermal transition in the structure and dynamics in these exciting materials systems. For example, a combined magnetic field and pressure may potentially provide the pathway to quantum spin liquid states. The modulation by pressure and temperature allows fine controls, such as for the exchange interactions, while in situ neutron scattering will be able to probe the spin Hamiltonian and underlying quasi-particles, as shown in a study on SrCu₂(BO₃)₂.⁴² High pressure neutron scattering under a high field will also help to answer some fundamental questions related to unconventional superconductivity.

Furthermore, combining a high magnetic field and high pressure will resolve many long-standing fundamental questions on phonon dynamics and the coupling between the lattice and spin degrees of freedom. The new extreme environment will have to answer questions such as: (1) What is the role of phonons in the damping of coherent spin excitations? (2) How does angular momentum transfer between spin and lattice degrees of freedom? (3) How to control the spin order and excitation through lattice dynamics? Our recent results on antiferromagnetic NiO suggest that such interactions are quite complicated in materials with strong spin–lattice coupling.⁴³ High pressure provides an ideal tool to control the lattice degree of freedom while a simultaneous control of the spin degree of freedom will be achieved by a high magnetic field. This combination will enable neutron scattering studies of many spin–phonon interactions for the first time, providing insights into the applications of such materials in next generation spintronic logic and memory devices.

There are two fundamental ways to technically achieve the combination of magnetic fields and pressure. They are to design a pressure cell to work in a standard cryomagnet or to wind a magnet integral with the pressure cell. The former can be operated as a steady-state magnet, and the latter is typically providing pulsed fields. Each of these two cases will be described below.

For the case of cryomagnets, pressure cells have been largely limited to piston-cylinder-type pressure cells (P_max = 2–3 GPa)^25–28 and hybrid anvil cells (P_max = ∼4 GPa).²⁹ They are made from non-magnetic CuBe or “Russian alloy,” NiCrAl. The advantage of these cells lies in their geometry: A long, cylinder accommodates a large sample volume yet easily fits within the cylindrical bore of a magnet. More recently, there has been interest in using neutron DACs since pressures above 4 GPa require opposed anvil cells yet also require small, non-magnetic cell bodies. DACs can address this need as they can be shrunk in size and can be made from non-magnetic CuBe or NiCrAl. Thus, a clamped neutron DAC developed at ORNL was made from non-magnetic CuBe and limited in outer diameter to 50 mm to fit into existing magnets.⁴⁴ Furthermore, a dedicated development project at the Jülich Center for Neutron Science specifically targets simultaneous high pressure, high magnetic field conditions at the Heinz Maier-Leibnitz Zentrum.⁴⁵ 

Here, we show a DAC concept made entirely from Pascalloy^TM (Tevonic, USA), which is a version of NiCrAl commercially available in the US. NiCrAl is chosen here due to its superior performance in high fields, its advantageous mechanical properties, and, compared to CuBe handling, the ease in its machining.^17,46–49 The cell concept is shown in Fig. 1 and is limited in outer diameter to 30 mm, which is consistent with current magnet bores of 32–50 mm and will fit new magnet concepts that will enable the steady-state field of up to 25 T.³¹ 

Its design is optimized for maximum pressure, ease of handling, and temperature stability and is based on experiences with past ORNL neutron DAC designs.^44,50–52 Specifically, like previous clamped DACs, the cell boasts a spring to stabilize pressure upon cooling.^44,50 The spring is designed that, based on Pascalloy’s strength,⁵³ a maximum load of eight tonnes can be applied. Ultimately, experimental evaluation of springs will determine the final thickness for the best compromise of elasticity and strength. Furthermore, here the external load is directly applied onto the spring rather than the opposing side. This allows for the addition of set screws that align the opposing piston/anvil assembly. These improved alignment capabilities aid in ease of cell preparation, overall pressure stability, and maximum pressures possible.

The cell can be used in beam-through gasket geometry. In the past, this proved feasible for single crystal diffraction and neutron spectroscopy^44,50 but has not yet been tested for powder diffraction. It could also be adapted for use in a beam-through anvil geometry. This is useful for powder diffraction and would additionally enable its use within a magnet for small-angle neutron scattering. This adaption would occur via the use of a separate seat for the anvil that includes final collimation just behind the anvil.^51,52

The cell can be equipped with polycrystalline diamond anvils, such as Versimax^®, or with large single crystal diamond anvils depending on maximum pressures desired and minimum sample volumes required. Based on previous experiments in such clamped DACs, eight tonnes on a single crystal anvil with 2 mm culets yield at least 20 GPa⁴⁴ on sample volumes of 0.15 mm³.⁵² In contrast, at the same load, Versimax anvils with a 2.5 mm culet yield a pressure of ∼15 GPa⁵⁰ yet on a sample volume of 0.5 mm³. Note that the properties of Versimax in a high magnetic field have not yet been assessed in detail while single crystal diamond is non-magnetic.

Furthermore, previous designs used steel gaskets with steels high in Cr content, low in Ni contents, and without Co, namely, 301 stainless (6–8 wt. % Ni, 16–18 wt. % Cr) and 15-5 PH stainless (3.5–5.5 wt. % Ni, 14–15.5 wt. % Cr). When combining high pressure and high magnetic field these steel gaskets have to be replaced with non-magnetic materials, such as NiCrAl. NiCrAl has indeed previously been used as gasket material in non-neutron DACs (e.g., Refs. 49 and 54), and it seems reasonable to expect that it can also be adapted to neutron DACs in terms of its mechanical properties (e.g., ductility, brittleness, toughness).

Additionally, several applications used a beam-through-gasket geometry, and thus the neutronic properties of Pascalloy need to be considered. Russian alloy (the 40HNU-VI alloy of NiCrAl) has been assessed in detail for inelastic neutron scattering (INS) by Kibble et al.,⁴⁶ while Cheng et al.⁴⁷ probed a self-manufactured NiCrAl alloy for high pressure neutron scattering. To confirm similar properties of the Pascalloy available to us, it was briefly assessed for its structural characteristics and its transmission. Our neutron diffraction data appear overall very similar to the previously reported synchrotron x-ray diffraction data,⁴⁷ including the detection of the long d-spacing 101 and 001 Ni₃Al peaks (see the  Appendix for details). Furthermore, simultaneous collection of transmission data revealed relatively small differences between Pascalloy and a 301 stainless steel gasket in the high-energy waveband collected (∼0.9 to 3.6 Å). The only difference detected between Pascalloy and 301 stainless was an asymmetric edge with the lowest transmission at 2.03 Å seen only for Pascalloy. This appears consistent with such a Bragg edge measured for NiCrAl by Kibble et al.⁴⁶ However, our overall data quality taken under typical “DAC measurement conditions” was too poor for detailed further analysis. Nonetheless, overall there is no reason to suspect that Pascalloy would perform worse than the previously used steel gaskets.

Overall, this concept presented here strongly suggests that a NiCrAl neutron DAC capable of pressures above 10 GPa and equipped with non-magnetic NiCrAl gaskets can be readily combined with high-field, steady-state neutron cryomagnets. Additionally, such a DAC may also open options for use with pulsed-field magnets, i.e., magnetic mini-coils.

The strongest magnetic fields are achieved by placing coils as close to the sample as possible. In magnet design, a compromise is made between maximum field and access to the sample. The advent of the mini-coil technology, developed by Yoshii et al.⁵⁵ and adapted for use at ORNL,^10,11 in combination with a larger version of the DAC shown in Fig. 1 can place the coils much closer to the sample. Here, the coil has a 12 mm bore, a 34 mm outer diameter, is 24 mm long, and consists of ∼140 turns of round CuAg (10% Ag by weight) wire.⁵⁶ It is wound around a G10 former to reduce eddy current heating. An example coil is shown in Fig. 2(a). Previous versions of our neutron DACs were significantly larger (e.g., 50 mm outer diameter, 97 mm tall⁴⁴), and the DAC presented here was specifically reduced in size to fit a cryomagnet. Thus, returning to previous sizes together with using a post thickness of 5 mm each (which provides sufficient mechanical stability for ten tonnes load) an open inner diameter of 40 mm is available. This has sufficient space to accommodate the 34 mm coil.

Thus, if we use the beam-through-anvil scattering configuration, the standard pulsed magnet without modification can be placed around the anvils. For sample loading, the top piston, cap, and spring are removed from the cell to access the gasket chamber. Once the sample is loaded into the gasket chamber, the magnet coil would be slid on the anvil/gasket/sample assembly. Then, the top anvil and piston would be inserted and closed with a spring and cap. With a neutron DAC of previous typical sizes (rather than the reduced-size DAC shown here), such an assembly is straightforward. Once the cell is assembled with the magnet, an outer vacuum chamber that encompasses both the coil and the DAC would be sealed and pumped. Then, the coil can be cooled with LN₂ in isolation from the DAC, which will remain at room temperature or, along with the sample can be cooled independently. The cooling of the coil will require a cryostat that closely wraps only the coil. Thus, a vacuum chamber is required for insulation for the magnet and to prevent condensation on the magnet parts. Such a chamber would have windows made of SiO₂ to allow transmission of the incident and transmitted beams and to allow for optical access to the cell for pressure measurement.

An example of the 30 T field pulse possible with such coils is shown in Fig. 2(b). The maximum field is currently limited to 32 T by the available power supply. However, by adding a larger power supply, these magnets could be driven up to 40 T.⁵⁷ After each full field pulse, the magnet needs 5–7 min to cool. Therefore, this technique will make full use of the bright pulses from the STS. However, this low repetition rate means that it cannot be used for spectroscopy at the FTS and likely also not at the STS. Nevertheless, for white beam diffraction, careful phasing of the magnet pulse with respect to the neutron pulse can recover some of the efficiency loss as described, along with details of the acquisition system and the event-based processing, elsewhere.⁵⁸ 

The above design summarizes possibilities that are within the standard design envelope of the currently available coils. However, the mini-coil is sufficiently cost-effective that one could envision designing a custom magnet per experiment. In such a case, using the anvils themselves as the coil former is a possibility. With the aforementioned larger power supply, split coil designs that would allow a vertical orientation of the DAC, and thus more scattering angle can be considered as well.

For both, materials science and geosciences the simultaneous application of high pressure and high temperature is essential. This combination of extremes gives unique access to materials energy landscapes and can be used to create new structures of elements such as diamond⁵⁹ or cubic nitrogen.⁶⁰ It also enables novel chemistry, such as the synthesis of superhard nitrides, carbides, and borides, e.g., Ref. 61, or the synthesis of metal superhydrides that promise record-breaking temperatures for superconductivity, e.g., Refs. 62–64. The understanding of the temperature–pressure phase behavior is key not only to material science and condensed matter physics but also in geosciences as it is the only method to replicate the conditions within the interior of planetary bodies. These studies are complex, and the results are often controversial. The melting curve of iron alone has been investigated for decades, yet remains controversial, e.g., Refs. 65–67.

The tool of choice for static pressures is the laser-heated DAC and many synchrotrons boast dedicated beamlines, e.g., Refs. 21, 22, and 68–70. Analogous systems have, however, not yet been set up for neutron beamlines. There, high pressure, high temperature is typically applied in multi-anvil set-ups^23,24 or within dedicated graphite heaters for Paris-Edinburgh cells.⁷¹ Yet, there are many scientific questions that would require the higher pressures afforded by the DAC but require the sensitivity of neutrons to light elements. For example, in materials science, a laser-heated neutron DAC could be invaluable for identifying the mechanisms of the formation of metal superhydrides.⁷² In geosciences, it would, for the first time, allow addressing the questions arising from the hydrous earth minerals and the possibility of large water reservoirs within the mantle, e.g., Ref. 73 and therein.

To facilitate such laser-heating in a neutron DAC, several key issues need to be addressed, particularly the sample volume required and the duration of the heating. During laser-heating in a DAC as used at a synchrotron, the laser beam creates a “hotspot” in the order of several tens of μm. Due to thermal conduction, this hotspot is subject to large temperature gradients. Modern high pressure beamlines at synchrotrons, however, often allow for sub-micron x-ray beams, that allow measurements of any portion of the sample. Significantly larger sample volumes are required for neutron scattering. Any in situ neutron scattering experiment in a DAC requires us to probe the entire sample under pressure. Consequently, the entire sample needs to be heated to one constant temperature. This can be facilitated through the use of a defocused CO₂ laser (λ = 10.6 µm). The advantage of a CO₂ laser lies in the fact that it couples to wide-bandgap materials, including diamonds. This cannot be achieved with a near IR (λ ≈ 1 µm) fiber laser, since the latter does not couple with diamond. We have previously used CO₂ laser heating on a range of materials^74–76 and have demonstrated its use without the typically added insulating materials or pressure media.⁷⁷ We, therefore, suggest the use of a defocused CO₂ laser beam to couple with both the sample and the diamonds. This technique significantly reduces the temperature gradient in the sample. While this may not yield ultra-high temperatures to 4000 K or above as possible with focused lasers, we anticipate temperatures up to ∼1500 K.

Since the samples need to be heated for sufficiently long times for in situ neutron studies, special techniques are required: For continuous heating to temperatures up to about 1000 K, the diamond cells have to be externally cooled and the diamond anvils have to be protected with inert gases. Another method to reach higher temperatures is flash heating, which was introduced recently.⁷⁸ There, a rectangular pulse heats the sample in short flashes, for several tens of ms. This is sufficient to heat the entire sample volume within a DAC and measure its temperature spectroscopically. This enables short “hot” periods followed by short “cold” periods, which will serve to avoid issues with the disintegration of the gaskets and anvils. Depending on the sample, accompanying finite element calculations will verify the entire sample within a neutron DAC can be heated uniformly. Moreover, the use of the event-mode at SNS allows for the identification and sorting of neutrons scattered during the “hot” and “cold” periods. Despite such a stroboscopic heating approach, it is thus possible to build up statistics for the sample scattering only at high temperature. Indeed, this stroboscopic approach has previously been successfully used for laser-heating experiments using levitators on SNS’s Nanoscale-Ordered Materials Diffractometer (NOMAD)⁷⁹ and will aid the implementation for the laser-heated DAC.

Based on these considerations, we present a concept for a set-up for a laser-heated neutron DAC, which is shown in Fig. 3. This is based on existing neutron DACs (with single-crystal diamonds)^51,52 as used at ORNL’s dedicated high pressure diffractometer, Spallation Neutrons and Pressure (SNAP). The conceptual drawing uses, however, the DAC introduced in Sec. II A. The set-up is designed for a beam-through anvil geometry. It will use the newest generation DACs that allow for sub-millimeter culet diameters (i.e., typically 600–800 µm culets) and thus relatively small sample sizes (i.e., typically 250–350 µm diameter). The necessary collimation typically added upstream of the anvil is not included for clarity here, However, as the collimation diameter steps down as it gets closer to the anvil, we expect that it will not be an issue to adapt to a laser-heating set-up. We suggest splitting the CO₂ laser beam to facilitate heating from both sides. Therefore, the laser-beam is directed onto the sample position using a mirror set-up, and the beam is focused to a diameter of the same magnitude as the sample diameter. A spectrometer (placed off the neutron beam path to allow for continuous operation during neutron collection) is added for temperature measurement.

Once such a set-up is established some material optimization is most likely required. Specifically, current cells often use metal seats for the diamonds. For extended heating of the diamond, it may be preferable to replace those with high-temperature ceramics. Finally, current neutron diffraction experiments have successfully used Re and W gasket, gasket materials that are also used for the laser-heated DAC. Due to the use of a defocused laser heating the anvil and the long duration of the flash heating, it will most likely become necessary to also optimize gasket materials, e.g., create custom-designed mixes of metal and diamond powders.

Furthermore, temperature measurement through optical means is typically easier at ultra-high temperatures rather than at the temperatures to 1500 K anticipated here. This issue could be overcome through temperature measurement via neutron resonance spectroscopy. In the past, this has been successfully used for in situ temperature measurement under high pressure, high temperature in the Paris-Edinburgh cell.⁸⁰ It would be highly interesting to adapt this approach to the neutron DAC, not just to enable reliable temperature determination at lower temperatures but also as an independent mean for temperature verification in a laser-heated DAC.

Overall, these above considerations and the presented concept suggest that laser-heating in a neutron DAC to at least 1500 K will be possible. While this may not be the same temperature maxima as possible using small samples and focused lasers, this will open many avenues for science questions related to materials sciences and earth and planetary sciences.

Development and synthesis of novel engineering materials frequently requires high temperatures often also at simultaneous moderate pressures only together with precise in situ measurement of phase transitions and physical properties and access to non-equilibrium states. The demands for such in situ characterization at temperatures above 3000 K are driven by the development of new materials for hypersonic applications, nuclear propulsion systems, fuel coatings in gas-cooled nuclear reactors, and plasma containment in nuclear fusion reactors.

More specifically, testing of ab initio predictions and the outputs of materials design schemes [e.g., the CALculation of PHAse Diagrams (CALPHAD) methodology] requires precise information about phase equilibria across the entire composition and temperature space. Levitation techniques were developed to provide containerless processing, free from heterogeneous nucleation sources as well as crucible contamination, at temperatures above 3000 K with minimal thermal gradients due to multi-directional laser heating. These are becoming more common over the last several decades. By offering unobstructed optical access to samples, they provide some of the best terrestrial methods for the study of density,^81,82 viscosity,⁸³ surface tension, resistivity, and ac susceptibility at ultra-high temperatures. The possibility of unobstructed horizontal access to the sample makes these extreme environments ideal for integration into beamlines at scattering facilities. Such deployments have allowed some of the first structural measurements into undercooled liquid metals,⁸⁴ in situ metastable phase diagram mapping,⁸⁵ and the self-diffusion behavior of various undercooled liquid alloys.⁸⁶ To cover the full range of materials, a large variety of levitation schemes exist, all suitable for different material classes: for instance, electromagnetic (EML) and ground-based electrostatic (ESL) levitators excel with conductive materials, aerodynamic levitators (AL) are perfect for ceramics, acoustic levitators (AcL) are best for low density liquids around ambient temperature conditions, and various hybridized schemes combine the strengths of the different techniques.

Both neutron scattering facilities at ORNL, SNS, and the High Flux Isotope Reactor (HFIR) have deployed ESL³ and AL⁸⁷ levitation schemes for user experiments for years. For example, they have used stroboscopic methods to study short-lived undercooled liquid states⁷⁹ and inelastic scattering to determine the behavior of the Van Hove function in high temperature liquids.⁸⁸ Early commissioning was performed at SNS’s Engineering Materials Diffractometer, VULCAN, and HFIR’s Wide-Angle Neutron Diffractometer (WAND²). Nowadays, most structural studies are performed at SNS’s Nanoscale-Ordered Materials Diffractometer (NOMAD) with both AL and ESL, while ESL is also regularly deployed to SNS’s Wide Angular-Range Chopper Spectrometer (ARCS) and SNS’s Cold Neutron Chopper Spectrometer (CNCS) for inelastic and quasi-elastic scattering. Further aero-acoustic and electromagnetic levitators are planned for deployment in the next several years, with an additional ESL being developed for integration at HFIR’s High Intensity Diffractometer for Residual stress Analysis (HIDRA) for creep measurements.

Beyond the individual application of high temperature, levitators can also be combined with other extremes (e.g., operating inside a 4 T superconducting magnet for accurate thermal conductivity measurements⁸⁹) and complementary measurement techniques (e.g., in situ Raman spectroscopy during crystallization⁹⁰). Such levitation experiments can provide troves of simultaneously acquired data that frequently cannot be obtained by conventional methods. Furthermore, the use of multi-extremes can also aid in technical challenges due to sample stability. While all of these levitation systems are often capable of reaching well over 2000 K, the stability of the sample is often the limiting factor. Evaporation at ultra-high temperatures either prevents access to temperatures typical of synthesis or limits steady-state measurements to timescales unfeasible at current neutron sources due to typically long acquisition times, particularly for spectroscopy.

One thrust of high temperature experimental development is thus the combination of ultra-high temperatures (up to 4000 K) with elevated gas pressures (∼100–200 bars = 10–20 MPa), primarily for reducing evaporation rates. It has been demonstrated that inert gases at moderate pressures can reduce the evaporation rates of metals⁹¹ and molten salts.⁹² By simultaneously applying moderate pressure during high temperature application through the various levitation techniques, it would be possible to develop a complementary suite optimized for different materials at ultra-high temperatures. Making the instruments available to the user community both on the beamline for diffraction and inelastic scattering and offline to prepare for beamtime or collect complementary thermophysical property measurements will increase the efficiency of beam time use and aid the technique development.

One possible implementation is the combination of EML with a moderate pressure environment. EML creates a levitation force through the interaction of a high frequency magnetic field (generated by specially optimized coils) with the induced eddy currents in the sample. Therefore, samples need to be conductive, and resistive heating of the samples is strictly linked to the levitation force. Unlike ESL, EML is naturally stable and requires no complex positioning system. For a mobile extreme environment meant to deploy multiple beamlines per cycle under operation from many different users, this is a critical feature. Intrinsic heating is generally used in combination with a cooling gas to control sample temperature. With the addition of a heating laser along the vertical axis, ultra-high temperatures can be generated. Additionally, the lack of dependence on charge (as in ESL) or the need to create a specific type of flow (as in AL) would more readily enable simultaneous moderate pressures. Indeed, experiments using EML have already demonstrated the effect of evaporative reduction under the atmosphere onboard the International Space Station (ISS).⁹³ 

A basic schematic of such a system is portrayed in Fig. 4. The large open angle in the horizontal direction allows flexible use of vertical-bore scattering geometries on scattering instruments. Samples will be inserted in the position from the bottom of the coil using motorized sample changers, which will include a gas jet for additional control of temperature and sample rotation. The laser heating will be accomplished from the top to enable electromagnetic levitation of semiconductor materials and to complement induction heating to achieve temperatures above 3000 K. The most refractory carbides, nitrides, and borides are electrical conductors and can be electromagnetically levitated without preheating (Fig. 4).

To enable operation at pressures up to 200 bars and temperatures above 3000 K, the levitator chamber will require robust walls and windows, active cooling channels for heat dissipation, and significant laser safety concessions to make the instrument available both on and off the beamline. Most EML coils are made of water-cooled copper tubing. Alternate materials will need to be investigated, both for neutron compatibility and temperature stability. The extent of evaporation rate suppression will need to be studied with well characterized standards to ensure that convective mass loss does not overwhelm any benefit from increased pressure. The effects of ionization of the argon gas will need to be characterized.

By deploying this system to the ORNL’s diffractometers and spectrometers, we expect to dramatically expand the possible range of high temperature studies and present them to the user community as a routine service. For instance, a pressurized EML alone would be able to extend the stable temperature range of various high entropy or compositionally complex alloys well into the liquid state. Stabilization of elements, including Mg or Mn, would ensure stoichiometric stability over extended measurement. The addition of stabilizing forces materials would extend the possibilities to carbides and nitrides, compositionally complex alloys, and even molten salts. Finally, a variety of levitation-based sample changers exist,^3,82 which could be adapted to this multi-extreme environment to take advantage of the increased neutron flux associated with projects such as the construction of the STS and the associated Proton Power Upgrade (PPU) at the SNS that will also increase flux for FTS instruments.

As detailed in this Special Topic issue, eight new instruments will be built as the first suite of STS instruments. While eventually all instruments will use various complex and/or extreme environments, several instruments will initially play key roles in the use of (multi-)extremes. Specifically, two STS diffractometers will focus on the use of (multi-)extremes in a range of studies centered largely on quantum materials, VERDI, a Versatile Diffractometer with full polarization analysis capabilities, and PIONEER, a high-resolution single-crystal polarized neutron diffractometer. Furthermore, CENTAUR, the small- and wide-angle diffractometer/spectrometer, will enable measurements over a wide range of length scales for a diverse range of science areas. Finally, the CHopper spectrometer Examining Small Samples, CHESS, is a direct geometry spectrometer focused on samples with very weak scattering or very small in size. The concepts for multi-extremes will be implemented in various forms across this instrument suite as detailed now. It should, however, be noted that certain applications will require instrument capabilities not selected yet. These may become available in later instrument suites such as through the TITAN proposal.

One key challenge that needs to be addressed across the instrument suite for the deployment of these (multi-)extreme environments is the need for alignment. As sample and beam sizes are further and further reduced, it is critical that samples within their extreme environments can be scanned across the beam to optimize the sample signal. This has been well developed at synchrotron sources where beam (and sample) sizes can be submicrometers. Traditionally, this has been less critical at neutron sources where beam sizes are typical of several cm² and sample sizes are often several cm³. For the measurement of smaller samples, it was in the past usually sufficient to cover any part of the extreme environment not containing the sample with neutron absorbing shielding (Cd, Gd, or B₄C). This is often insufficient for the very small samples contained with DAC, as we found when deploying DACs to beamlines beyond the high-pressure diffractometer SNAP. In addition to shielding or more sophisticated incident-beam collimation, it has become a necessary routine to also optimize the sample signal through several alignment scans across the neutron beam. This requires high-precision alignment fixtures implemented in the beamline (e.g., hexapods) with micrometer precision. Specifically, for the highest pressures, these high-precision alignments are unavoidable although lower pressures (on larger samples) can be made possible with less precision alignment through custom incident-beam collimation. Such custom incident-beam collimation has become a must for DACs used throughout the facility and will certainly be equally necessary for multi-extremes, such as simultaneous high pressure, ultra-low temperature, and high magnetic fields.

The versatile diffractometer VERDI will be optimized for studies on magnetism in single-crystal and powder samples.³² This will be facilitated through a very small Q_min of 0.1 Å⁻¹ (d ≈ 65 Å) and through its full polarization capabilities. It will further be optimized for studies under the extremes of high pressure, high magnetic field, and dilution temperature. The instrument will be designed to accommodate a dedicated ∼14 T magnet. This magnet is customized for full Q-coverage and optimized out-of-plane coverage.

Naturally, science on VERDI will not only require these individual extremes but also the various combinations thereof. Thus, the magnet neutron DAC could become a key multi-extreme environment for VERDI. Since the instrument will feature a customized magnet, it will be possible to co-optimize so that it will be compatible with the 30 mm neutron DAC presented here.

We anticipate that the majority of studies in such a neutron DAC on VERDI would be conducted in a beam-through-gasket geometry. There, experiments would be conducted in a similar manner to past DAC studies on SNS’s Elastic Diffuse Scattering Spectrometer, CORELLI. There, custom incident-beam collimation was mounted directly on the DAC.⁹⁴ This has allowed for the successful measurement of ∼0.07–0.1 mm³ single-crystal samples within a DAC on CORELLI.^34,50,95 Furthermore, the WISH instrument at the ISIS Neutron and Muon Source (UK), an instrument that is conceptually very similar to VERDI, has also developed a successful high pressure program using a hybrid DAC.^29,96,97

With the smaller beam size available at VERDI and through the high-brightness cold neutrons provided by the STS, it is thus expected that experiments in the magnet DAC would be highly successful. We anticipate that even if sample sizes are reduced compared to past CORELLI experiments (reduction to ∼0.01 mm³), it will be entirely feasible. Implementing this magnet DAC on VERDI could thus conceivably enable pressures up to 20 GPa at a simultaneous 14 T.

The PIONEER high-resolution single-crystal diffractometer is designed to measure very small single crystals, on the order of 0.001 mm³ or less.³³ These small sample sizes will be enabled through the increased peak brightness of the STS and the use of advanced Montel mirrors. This would make PIONEER also a good candidate for the use of the magnet neutron DAC, whereby the cell would allow for single-crystal sizes of ∼0.01 mm³ when used with a 2 mm culet diameter. Furthermore, PIONEER could also benefit from its variable scattering directions since it will have a horizontal as well as a bottom detector array.

Interestingly, such small sizes anticipated are usually regarded as x-ray sample sizes rather than neutron sample sizes. While this capacity enables many new science directions, it also has very interesting implications for (multi-)extreme environments: Extreme environments developed for synchrotron x-ray sources could be deployed on PIONEER.

This is of particular interest in high pressure where large differences exist between x-ray and neutron capabilities. In order to accommodate the necessary large sample volume needed for neutron diffraction, achieving high pressures in neutron DACs necessitates high loads (up to 10 tonnes). To withstand such high loads, cells need to be strong and sturdy, seat materials need to be strong and even springs for temperature stability need to be of sufficient thickness. This prevents shrinking the cells beyond a certain point. This is different for DACs used at synchrotrons where sufficient loads of just a few 100 kg are applied through simple screws. Such cells are more versatile, are significantly smaller, and are readily adapted to low-temperature and/or high magnetic fields. Indeed, the so-called miniature turnbuckle DACs made from plastic have even been shrunk in size to below 10 mm diameter, which makes them compatible with magnets capable of well above 50 T.⁹⁸ For in situ x-ray measurements, DACs made from CuBe are also easily combined with low temperatures, high magnetic fields,⁹⁹ and they are readily available in the user program.

The use of such standard DACs rather than neutron DACs on PIONEER could be beneficial for single-crystal neutron diffraction studies. The use of x-ray (or miniature) DACs has previously been demonstrated for single-crystal neutron diffraction^100,101 at low pressures and for magnetic structure studies on powder samples to ∼45 GPa.¹⁰² In these cases, however, measurement times were very long (1 day or longer). In contrast, the optimization of PIONEER is estimated to yield a full dataset for a 0.001 mm³ crystal within several hours. While sample size requirements may not allow for the ultra-high pressures above the megabar, sample sizes possible on PIONEER should enable up to ∼50 GPa at reasonably fast measurement rates.

Furthermore, the same cells are readily adapted to the simultaneous use of the high magnetic field, low temperature as routine at synchrotrons or for many ex situ techniques. It, thus, may be interesting to test such magnet-compatible x-ray DACs with the high-field magnets to be used on PIONEER.

The versatile and flexible small-angle and wide-angle neutron scattering instrument CENTAUR will boast a large dynamic range in reciprocal space.³⁴ It will use a high-flux neutron beam for high-throughput measurements or to enable measurement on small samples. To address the wide-ranging scientific questions it will enable, CENTAUR aims to accommodate a large range of complex environments typically used the soft matter and biological science yet also the extreme environments needed for pressure, high magnetic field, and temperature discussed here.

These characteristics make CENTAUR a very interesting candidate for the deployment of the magnet-friendly neutron DAC. Typical small angle neutron scattering (SANS) sample shapes are thin disks, which are replicated by the sample shape in a DAC in a beam-through-anvil geometry. For example, for a DAC with a 2 mm culet diameter, as mentioned in Sec. II A, the starting sample chamber would be a 1 mm disk, 0.2 mm thick in this configuration. Since CENTAUR is aiming to probe samples as small as this in beam cross section, DAC experiments appear feasible.

The simultaneous application of pressure and high magnetic field could have many interesting applications in the search for skyrmions and other spin textures. To date, it has been difficult to achieve this on existing SANS instruments although it has been attempted on EQ-SANS, for example, in the past.¹⁰³ One difficulty has been that sample sizes in a DAC are small and experiments will most certainly benefit from the high flux and small beam at CENTAUR. Furthermore, the scattering geometry of the pulsed magnet closely coupled to the DAC lends itself naturally to SANS with its preference for forward scattering. This could be another avenue to explore beyond a magnet-friendly DAC within a steady-state magnet.

Additionally, the complementary diffraction capabilities offered through the wide-angle scattering will also be highly beneficial for DAC experiments. This will allow us to probe the structure of the sample to verify it is still intact and to determine the pressure in situ through the pressure equation of state of the material. The combined small-angle/wide-angle capability together with the high flux thus presents a powerful opportunity for the multi-extremes of pressure, high magnetic field (and low temperature).

The direct geometry spectrometer CHESS will enable the simultaneous measurement of dynamic processes over a wide range of energy on very small samples (∼1 mm³).^35,104 This will be achieved through optimized beam size at sample position, a large detector coverage, and a targeted energy resolution range. CHESS’s focusing guide will yield a small beam size of ∼1 mm, allowing for unprecedented small samples in terms of inelastic neutron scattering (INS). Furthermore, CHESS will also incorporate a hexapod placed within its vacuum chamber in order to enable micrometer-precision alignment of small samples against the neutron beam.

These features make CHESS a prime candidate for many extreme environments and also the multi-extremes discussed here. Past evaluations have shown that the current iteration of SNAP diamond cells, including built-in collimation, are highly compatible with CHESS in the beam-through-anvil geometry. On SNAP, these DACs can be readily cooled to 5 K, which should be equally feasible on CHESS. Preliminary considerations also suggest that comparable sample sizes will be possible for many studies, suggesting that INS to ∼50 GPa may become possible.

Furthermore, CHESS will aim to enable the use of an 18 T cryomagnet. Since DACs are compatible with CHESS, it may be equally possible to deploy the magnet DAC within the cryomagnet (depending on bore sizes available). This set-up will, however, likely operate without the hexapod in place limiting alignment capabilities. Without micrometer-precision alignment samples typically have to be larger, which in turn limits pressures. Nonetheless, it will be interesting to explore high pressure, high magnetic field on CHESS.

Finally, despite the cell alignment capabilities, CHESS will not afford the accessibility to the sample stage that is available on SNAP, for example. Thus, the set-up of an in situ laser-heating system would be very difficult since this needs careful manual alignment. It may, however, be possible to set up an ex situ laser heating system directly at the beamline. There, the DAC could be brought up from the tank, a laser beam could be rastered over the sample for heating, after which the DAC would be taken back down for measurement. Such a set-up naturally would not afford simultaneous high pressure, high temperature conditions. It would, however, enable the synthesis of novel materials that only persist under pressure such as the synthesis of metal superhydrides with their record-breaking T_c.^62–64 

CHESS will also support the development of a new electrostatic levitation furnace and possibly the pressurized levitators (with the implementation of a re-entrant well for additional containment). Previous inelastic scattering experiments at ARCS and CNCS with the Neutron Electrostatic Levitator (NESL) were able to probe the Van Hove function of liquid metals at a range of incident energies.⁸⁸ The analysis of the dynamic pair distribution function requires both a large Q-range and a broad swath of incident energies to optimize the resolution for different features. The repetition-rate multiplication features of CHESS will enable this type of experiment to be performed in significantly reduced time, which is critical when the regime of interest is a transient undercooled state. Additionally, quasi-elastic scattering performed at CNCS with NESL to determine the self-diffusion coefficients of species in undercooled liquid metals could be furthered at CHESS, where the increased brightness, the emphasis on small samples, and the cinematic acquisition modes will allow unprecedented access to the dynamics of the undercooled regime.

While currently selected instruments at the STS and also existing instruments at the FTS will enable the use of many multi-extremes, there are also conditions that will not be possible in the current suite, particularly multi-extremes that include very high steady-state magnetic fields (∼40 T). The multi-modal TITAN instrument addressed many of the issues that arise from the use of such a 40 T magnet and also considered co-optimization for multi-extremes.³¹ These considerations and co-optimizations included bore size in the magnet, accessibility to the sample space, and alignment capability in addition to magnet specific considerations. The overarching aim was to build a diffractometer/spectrometer that was optimized for the use of a wide range of multi-extremes, including all the multi-extremes proposed here.

During the conceptual proposal stage, simulations of the TITAN instrument included simulations of a sample contained within a simplified DAC. This DAC was equipped with single-crystal diamond anvils with 600 µm culets to allow for a sample size of ∼0.004 mm³. This size was recently identified as the smallest sample measurable under high pressure at SNAP although long collection times are necessary (e.g., ∼5 h on simple metals, ∼10 h on water ice). It was thus chosen to illustrate TITAN’s potential. For the simulation, TITAN was set up with a bandwidth of 2–9 Å (achieved through operation at 7.5 Hz) and its “wing detectors” with a geometry of 70°–110° horizontal coverage at an ±60° out-of-plane coverage. Due to the large interest in superhydrides, the sample simulated was LaD₁₁ (spacegroup P₄/nmm, 150 GPa¹⁰⁵). In this simulation, excellent data quality (i.e., sufficient statistics for a full quantitative Rietveld analysis) was already obtained after a 10 min measurement only. This highlights the potential in terms of brightness and focusing that will be enabled by the STS for high pressure neutron scattering. Furthermore, the TITAN concept explicitly called for fine alignment capabilities as well as for accessibility to the sample space. All these characteristics make an instrument like TITAN an excellent candidate for the implementation of the laser-heated DAC discussed in Sec. II B.

While the TITAN instrument was not selected as the first instrument, many of the lessons learned and concepts developed will be used further on. These include the multi-extremes proposed here that can be adapted to many other instruments. This also includes the concept for another spectrometer, ZEEMANS, that will focus on deploying a steady-state magnet with up to 25 T magnet with a 32 mm bore.³¹ The magnet neutron DAC proposed here is compatible with this magnet and instrument and will enable simultaneous pressures to ∼15 GPa, not just for diffraction but also for neutron spectroscopy.

Here, we have introduced three concepts for the multi-extreme environments for use in future STS instruments and potentially also at other neutron sources at ORNL and beyond ORNL. Thereby, we propose to address the multi-extreme of high pressure, high magnetic field (and low temperature) through a 30 mm diameter DAC made from non-metallic NiCrAl. This cell is optimized for low-temperature stability through the use of a spring. It is also highly versatile in scattering geometry, making it suitable for a range of applications and, depending on sample volume needs, will accommodate maximum pressures of up to 20 GPa within a magnet. Its potential deployment for neutron diffraction on VERDI and PIONEER is discussed with its beneficial characteristics for SANS experiments on CENTAUR as well as potential uses for inelastic scattering on CHESS. Additionally, the use of a miniature coil-magnet placed directly around the anvil/gasket assembly is considered. This could provide high pulsed fields for CENTAUR, for example. Next, a concept for laser-heating a neutron DAC with a defocused CO₂ laser is described. The system benefits from the cinematic modes available at the SNS for stroboscopic measurements using flash laser heating. This could allow for simultaneous temperatures of 1500 K at pressures currently available in room temperature neutron DACs. A reduced version for offline heating is discussed for CHESS, while the full implementation will require the accessibility of an instrument such as the proposed TITAN concept. Finally, the third concept proposes a hybrid hyperbaric levitator system. There, the presence of inert gases under moderate gas will prevent sample evaporation. This will enable ultra-high temperatures up to 4000 K. The deployment of such a system on the CHESS instrument is discussed. These multi-extreme environments open new parameter space to be probed with in situ neutron scattering. They will thus be highly beneficial in future scientific endeavors at the STS and beyond.

We thank Jiao Y. Lin (ORNL) for the simulations of a DAC with a sample on TITAN. We thank the instrument teams of CHESS, PIONEER, VERDI, and CENTAUR for useful discussions. We are grateful for the contributions of Jooseop Lee and Justin Charmichael for the design and testing of the pulsed magnet and M. Rucker for coil winding. Last but not least, we thank all the researchers contributing to and supporting the TITAN proposal, particularly toward discussions on the development of multi-extreme environments, specifically Stanley W. Tozer (NHMFL), David Lipke (MST), Juergen Brillo (DLR), Takeshi Egami (UTK), Masa Matsuda, Yan Wu, and Malcolm Guthrie (all ORNL), and Collin Broholm (Johns Hopkins). This research used resources at the Spallation Neutron Source, a DOE Office of Science User Facility operated by the Oak Ridge National Laboratory (ORNL). This research also used resources of the Spallation Neutron Source Second Target Station Project at ORNL. ORNL is managed by UT-Battelle LLC for DOE’s Office of Science, the single largest supporter of basic research in the physical sciences in the United States.

This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

The authors have no conflicts to disclose.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

We collected neutron diffraction and preliminary transmission estimates from two gaskets, a 301 stainless gasket and a NiCrAl gasket. The NiCrAl used was Pascalloy, 3.3–3.8 wt. % Al, 39–41 wt. % Cr, balance Ni. It was heat treated at 740 °C for 12 h to a hardness of 58 Rockwell C. Both gaskets were 1 mm in thickness with an outer diameter of several mm.

The measurement was performed on SNS’s dedicated high pressure diffractometer Spallation Neutrons and Pressure (SNAP).¹⁰⁶ SNAP was set up with its focusing guide for a beam of 1–2 mm radius. Its choppers were set for a center wavelength of 2.1 Å at 60 Hz, yielding a wavelength band of 0.5–3.7 Å. The detector banks were placed at a center 2θ of 50° and 115°, respectively, for maximum Q-range. A high-efficiency N₂ beam monitor (Ordela, Inc.) placed 1.96 m downstream from the sample position in the transmitted beam was additionally used for transmission measurements.

The gaskets were mounted vertically standing up (i.e., perpendicular to the beam) so that transmission was measured for 1 mm thickness in both cases. An 850 µm incident beam collimator made from hexagonal boron nitride was used to center the beam on the sample and to ensure the same volume of material was exposed to the beam. This is a typical set-up used during DAC experimentation, which is why it was used here.

The resulting data were reduced in Mantid¹⁰⁷ and the diffraction data for Pascalloy are shown in Fig. 5. The transmission data simultaneously collected on the monitor covered a wavelength band of 0.92–3.56 Å, somewhat less than the overall wavelength band due to edge effects. Transmission between ∼80% and 90% was observed for both gaskets, 301 stainless and Pascalloy, for this wavelength band.