In-situ measurements are those where the sample-of-interest is studied in a scientifically-relevant environment (as opposed to an artificial environment convenient for the instrument; e.g. vacuum). Thus, proteins studied in their native liquid environment, or polymer films studied during thermal processing, are examples of in-situ.
Because x-rays and neutrons are strongly penetrating, scattering experiments are well-suited to being performed in-situ (this is one of the key advantages of scattering). Thus, materials can be probed under realistic 'native' conditions, or during relevant processing/transformation conditions. One can also define in-operando, wherein materials are studied during operation (e.g. a battery studied during charging/discharging). More abstractly, one can consider in-situ the study of materials under prescribed equilibrium conditions, whereas in-operando is the study during prescribed non-equilibrium (e.g. time-varying) conditions.
Many synchrotron beamlines now offer the ability to perform various in-situ experiments. This can be accomplished by having the sample environment sufficiently flexible so that users can bring custom-built sample-cells. Many beamlines also offer their own in-situ cells for common in-situ experiments (e.g. sample heating). However, in-situ experiments are of course always more complex than conventional scattering experiments, requiring additional time for setup, additional complexity with respect to instrument control, and additional care with respect to data analysis.
- Control ionic strength, pH, concentration, etc.
- Solvent exposure (solvent annealing)
- Chemical reaction
- Thermal cycling
- Thermal gradients
- Mechanical force
- Pressure (c.f. diamond anvil cell)
- Applied fields
- Visible light, IR, UV, etc.
Issues to consider
There are many complexities to consider when planning an in-situ experiment.
The materials used within the cell must, obviously, be able to resist the in-situ environment. The materials used must be able to withstand exposure to the selected solvent (this includes any glues or o-rings used within the cell). Aluminum and teflon are common choices for in-situ cells.
It is crucially important to consider the exterior dimensions of the sample cell. The cell's envelope must, of course, not intersect with any beamline equipment. It is important to also give some additional clearance for motion of the cell (required for alignment and sample translation). The cell's mass must also be within the limits of the motion hardware.
The introduction of an ambient environment typically increases the background signal, thereby worsening the signal-to-noise. The ambient material may also scatter substantially (c.f. diffuse scattering), thereby overwhelming the signal-of-interest.
The absorption of x-rays (or neutrons) must be carefully considered. Although x-rays penetrate quite strongly, they will be attenuated by sufficient path-length of solvent, or sufficiently-thick sample windows. With respect to solvent, the attenuation can be calculated, although one must also consider the possibility of non-absorption effects like multiple scattering. E.g. a thick path-length through a disordered or strongly-scattering material will extinguish the direct beam, and make the scattering data very difficult to analyze.
Custom sample cells must include windows that can allow the incident and scattered beams to transit relatively unimpeded, while safely containing the internal environment. In general, for x-ray experiments the windows must thus be x-ray transparent, while being able to hold against a pressure differential. Some materials that are frequently used:
- Kapton is fairly transparent and introduces a weak (but non-zero) diffuse scattering halo.
- Interestingly, certain plastic-wrapping materials (e.g. for food wrapping) have been found to be superior to Kapton, in terms of pressure-holding ability and low-scattering-background (e.g. 'nelophan').
- Beryllium (Be) has a low absorption and scattering, but introduces safety considerations due to toxicity of Be dust.
- Mica can be made thin and sufficiently strong to act as a window material over small apertures.
- Glass (silicon dioxide, quartz, etc.) can be used if sufficiently thin. E.g. an x-ray beam of energy 10-20 keV can penetrate through the ~0.100 mm walls of a capillary or flow-through cell.
- Silicon, if sufficiently thin, can also be used. Hard x-rays (15 keV or higher) can penetrate through ~0.5 mm of material (i.e. a typical wafer thickness) with acceptable losses.
- A 12 µm mylar foil (with 100 nm Al coating) is used to protect Pilatus detectors. This could potentially be used as a window material.
Any in-situ cell must be designed so that the exit-window does not block scattering angles of interest. The scattering angles one requires of course depend on the x-ray energy and the q-range one wishes to access (i.e. the size-scale of the structures one is measuring).
When designing an in-situ cell, one must know what the external environment will be. For instance: will the cell be used in a vacuum chamber at the beamline (in which case it must hold its internal environment against low external pressure)? In-situ cells are easier to use if the beamline can offer an ambient-air environment for the cell to sit in (however, this increases background, and may not be appropriate in some cases).
Many in-situ cells will require connections: e.g. to flow-through vapor or solvent, or electrical connections for heaters or sensors. An additional complication is the sample-environment at the beamline. If a cell is being used in vacuum, then all connections will have to couple to feed-throughs available on the sample chamber (this requires planning!). Even if a cell is being used in ambient-air, the hoses an wires must be able to reach from the custom-cell, and well-away from the measurement position. If the driving equipment (reservoirs, readout electronics) need to be accessible during an experiment, then they will actually have to sit outside the shielded hutch of the endstation. The hoses/cables will thus have to run through one of the hutch labyrinths. This means that the cable-runs may be much longer than one initially assumes!
Synchrotron beamtime is extremely valuable. Without proper planning, a beam-run can be entirely consumed just in trying to get a custom cell to behave properly. It is always best to test the cell before-hand; e.g. filling the cell in a lab-space and testing its ability to resist solvent, hold vacuum, etc. Any interfaces to computer-systems should also, ideally, be tested and debugged prior to the beam-run.
- Vanessa K. Peterson and Christine M. Papadakis Functional materials analysis using in situ and in operando X-ray and neutron scattering IUCrJ 2015, 2 (2). doi: 10.1107/S2052252514026062