Real-Time Observation of Stacking Faults in Gold Shock Compressed to 150 GPa

A microscopic-level understanding of the high-pressure states achieved under shock compression, including comparisons with static compression, is a long-standing and important scientific challenge. Unlike hydrostatic compression, uniaxial strains inherent to shock compression result in plastic deformation and abundant lattice defects. At high pressures ($>50\mathrm{GPa}$), the role of shock-induced deformation and defects remains an open question. Because of the nanosecond time scales in shock experiments, real-time in situ observations of shock-induced lattice defects have been challenging. Here, we present synchrotron x-ray diffraction measurements on laser-shock-compressed gold that provide the first unambiguous in situ measurements of stacking faults (SFs), likely formed by partial dislocations, during shock compression. SF abundance increases monotonically with shock compression up to 150 GPa, where SFs comprise almost every 6th atomic layer. Our results show that SFs play an important role in the plastic deformation of face-centered-cubic metals shocked to high stresses, providing a quantitative benchmark for future theoretical developments.

Popular Summary

The highest-pressure states of matter achievable in the laboratory are obtained using shock-wave compression (as encountered, for example, in high-velocity impacts and explosions). Unlike hydrostatic compression, shock-wave compression of solids results in plastic deformation and abundant lattice defects. Although quantitative measurements of these defects during shock compression are essential to understanding the shocked states of materials, real-time measurements have posed a significant challenge because of the very short (nanosecond) timescales of the experiments. Here, we report the use of recently developed experimental capabilities to quantitatively observe, in real time, the formation of lattice defects during shock compression.

For face-centered-cubic (fcc) metals, shock-induced deformation is expected to give rise to the formation of stacking faults, defined as alterations of the stacking sequence of atomic layers. Though theoretically expected, real-time measurements of stacking faults under shock compression of these metals have not been reported so far. Coupling in situ synchrotron x-ray diffraction with laser-shock-compression experiments, we have measured the abundance of stacking faults generated in gold, a representative monatomic fcc metal, shock compressed to high stresses reaching 150 GPa. Our results show that stacking-fault abundance increases monotonically with shock compression, so that almost every sixth atomic layer is a stacking fault at 150 GPa.

These results provide new insight into the important role of stacking faults during shock-induced deformation in fcc materials and will serve as a quantitative benchmark for future theoretical developments.

Figure 1

Schematic diagram for in situ nanosecond x-ray diffraction measurements in shock-compressed gold and representative results. A 351-nm laser pulse is focused on an aluminized Kapton ablator generating a shock wave which then propagates through the Au sample. XRD patterns corresponding to the ambient Au sample and to the Au sample partially shocked (35% of Au unshocked during XRD measurement) to 116.9 GPa are shown on the left. The diffraction rings are from the Au and the localized spots are from the single crystal LiF window; a few of the LiF spots are identified in the ambient XRD pattern. The LiF spots are easily identified as they are the same in both XRD patterns because the shock wave had not yet entered the LiF window during the XRD measurement in the laser shock experiments. For the XRD pattern from the partially shocked gold, the diffraction signal from the unshocked portion of the gold is clearly visible in azimuthally integrated 1D line profiles as shown in the Supplemental Material. Laser interferometry is used to measure the gold/LiF window interface velocity from which the shock stress is calculated. Laser drive temporal pulse shapes result in states of constant stress behind the shock wave, as verified by the measured particle velocity histories.

Figure 2

Shocked gold XRD results. Representative measured and simulated x-ray diffraction line profiles for gold at ambient conditions and for four different shock stresses. Line profiles measured in the shocked states have both the background and the ambient portion of the XRD line profile subtracted. Simulated line profiles (red dashed lines) for shock-compressed Au use lattice parameter ($a$) and stacking-fault probability ($\alpha$) values that give the best match to the measured line profiles. The simulation of the ambient line profile uses the ambient Au lattice constant $a=4.0782\AA$ and $\alpha =0$. The simulated line profiles are convoluted with a Lorentzian function whose width increases linearly with increasing shock stress (see SM [41]).

Figure 3

Stress-volume states of shock-compressed gold. Stresses are calculated from velocity interferometry data. Unit cell volumes are calculated from $a^3$, where the lattice constant $a$ is determined directly from a fit to the (111) diffraction peak (blue circles) or the (200) diffraction peak (green squares), without considering the effects of stacking faults. The red diamonds are from a combined fit of the full diffraction profile, comprising all five peaks, that incorporates shifting and broadening of the diffraction peaks due to stacking faults. The gray band is the Hugoniot curve for gold, determined from continuum measurements (Refs. [7, 19, 20]; see SM [41]); the width of the band indicates the experimental uncertainty.

Figure 4

Stacking-fault probability versus volume compression. Stacking-fault probability α increases monotonically with volume compression. Corresponding longitudinal stresses and calculated temperatures are shown at the top. The calculated temperatures are determined using the equation of state from Ref. [17]. At the lowest stresses ($\sim 40\mathrm{GPa}$), and as discussed in the SM [41], potential contributions of strength effects on the stacking-fault probabilities cannot be ruled out.