Following chemical reactions and physical processes, by observing the atoms moving, stands as one of the main interest of solid state research. When the reactant and the product are single crystals, single crystal diffraction is the method of choice for detailed in situ studies of how matter rearranges upon external stimuli. Many reactions proceed via minimum atomic movements and geometrical changes, thus the crystallinity remains intact. Sometimes, however, the structure profoundly changes by twisting, bending, flopping of molecular fragments, and/or jumping or hoping of atoms and alterations of the crystal packing. When the geometrical movements of atoms are significant, the crystal cannot withstand the stress generated and therefore loses its crystallinity (Fig.1(a)). The disintegration hampers the applicability of single crystal diffraction, making these reactions very loosely and sporadically studied. The breaking of the crystals, however, is all but an obstacle for XRPD. Taken the recent advances of instrumentation and software for structure solution, powder diffraction calls the challenge to study these reactions in detail . XRPD is sensitive to structural changes of crystalline matter, thus by inspecting the changes of diffraction patterns collected in situ, it provides an unique opportunity of visualization of phase transitions (Fig.1(b)) and obtaining structural informations (Fig.1(c)).
In some cases when the generated stress is immense, there is a rapid energy transfer through the dense and ordered packing of molecular single crystals and they are self-actuated by elastic deformation and/or by a ballistic event. This self-actuation phenomenon of crystals is termed "thermosalient effect" (Hüpf-Effekt) and is extremely rare propensity of certain crystalline solids capable of bending, twisting, curling, rolling or locomotion . These systems are a prototype materials for novel, environmentally clean and efficient thermal-to-mechanical energy transformation. The capability of rapid energy transfer through the dense and ordered packing of single crystals evolves as a new platform for efficient dynamic elements, including artificial tissues and microfluidic elements. The fast energy transfer secured by efficient coupling between the thermal energy and mechanical energy stand as their main assets towards the design of fast actuators. The thermosalient effect is often accompanied by anomalous cell expansion, such as exceptionally large positive thermal expansion and/or negative thermal expansion, which adds a special interest to the structural science of these materials. To decipher the mechanism of molecular triggering and the driving force behind the thermosalient effect and to elucidate the interplay between thermodynamic and kinetic factors in its occurrence, we focused our attention on the first documented case of jumping crystal. In a short article from 1983 Etter and Siedle  briefly noted that when crystals of (phenylazophenyl)palladium hexafluoroacetylacetonate (PHA) are heated on one side they "literally fly off the hot stage". XRPD was the method of choice for explaining the structural features of PHA as a prototypic thermosalient active crystal.
The 2D projection of the scattered X-ray intensity of PHA as a function of diffraction angle and temperature (Fig.1(b)) reveals remarkably rich crystal chemistry – there are five polymorphs which can interconvert thermally into each other (Fig.1(c)). Upon cooling, form α transforms into form δ via second-order phase transition, further cooling results in sharp, fist-order phase transition of form δ into form ε. These thermally induced processes are mirrored on heating, when form ε transforms back to form δ, and subsequently into form α. On heating form α underwent fist-order phase transition to form γ, which is transformed into form β. Once form β is obtained, it does not change upon thermal treatment, indicating that this is the thermodynamically most stable form.
Two phase transformations of PHA results in significant physical changes of the single crystals. The transition of form α to form γ is accompanied by forceful jumps of crystals, while the transformation to form δ – form ε leads to single crystal disintegration (Fig.1(a)). It is apparent form Fig.1(c) that these phase transformation go along with sharp and substantial change of the unit cell volume and this immediate cell expansion/contraction contributes to the crystal habit alterations. Only one of the phase transitions, however, is thermosalient and the "jumping crystal" effect cannot be solely explained by unit cell changes, but by changes of the crystal structure. The crystal disintegration hampers the applicability of single crystal diffraction for structure determination, making XRPD a remaining method of choice. Figure 2 presents the Rietveld refinement plots of the crystal structures of ε and γ solved by powder diffraction data.
Plotting the crystal packing of all known structures (Fig.3(a)) reveals important details: beside the unit cell volume changes, columns of stacked PHA molecules in form α and form ε are similarly packed, while in form γ the columns are significantly spaced apart. The cooperative change of unit cell and crystal packing generates internal stress which results in thermosalient effect. In form β the PHA molecules are differently packed in respect to each other, as compared to the packing in other forms. Accordingly the ε-δ-α-γ transitions are reversible in nature, while the γ-β transition is irreversible.
From Fig.1(c) it is apparent that at least two phases of PHA undergo exceptionally large thermal expansions/contractions. Form β shows all-positive thermal expansion, whereas in form α one axis is strongly expanding, another axis shows moderate expansion and the third one thermal contraction (Fig.3(b)). The typical values for thermal expansion coefficients of molecular solids are 0·10–6K–1<α<20·10–6K–1. Interestingly, form α and form β show exceptionally large volumetric thermal expansion, αV=247.8(2)·10–6K–1 and αV=255.5(3)·10–6K–1, respectively. Studying the crystal packing, a direct link between the anomalous thermal expansion and the mechanical response is established, i.e. it is traced back to the extraordinary susceptibility of this material to internal strain.
In summary, XRPD is a powerful tool in studying different phenomena, such as the thermosalient effect and thermal expansion, which originate from the internal strain that develops in response to the strong thermal anisotropy of some exotic crystalline materials.