Shock response and degassing reactions of calcite at planetary impact conditions

Calcite (CaCO 3 ) as a planetary material is a source to the atmospheric carbon dioxide through degassing by high-velocity impact events. Revealing the behavior of calcite in the extreme pressure and temperature conditions is required to understand the impact-induced degassing phenomena. Here we report laboratory investigations of shock- compressed calcite beyond the impact velocity of 12 km/s (faster than escape velocity from the Earth). The present precise shock measurements elucidate the shape of the calcite Hugoniot curve continuously passing through the melting and metallization states up to a pressure of 1000 GPa (= 10-million atmospheres) or a corresponding impact velocity of 30 km/s, allowing us to predict the post-shock residual temperatures and the dominant carbon oxide species in the impact aftermath. These predictions suggest that CO emission is much more dominant than CO2 at the impact velocities of ∼ 10 km/s and above, affecting the planetary atmospheric chemistry, greenhouse processes, and environmental changes during planetary evolution.

impact-induced degassing of carbonates plays a critical role in the planetary carbon cycle that affects the formation of planetary atmosphere and ocean and the habitability.
Calcite (CaCO 3 ) of carbonates is a typical component in chondritic asteroids 5 and rocky planet surfaces.
In fact, the large-scale, catastrophic impact evidence in places abundant in carbonates is visible also on the present Earth as the Chicxulub crater. The shock-compressed calcite behaviors have been investigated only up to 100 GPa pressure, corresponding to an 8 km/s impact velocity. In this study, the Hugoniot pressure-density-temperature (P-ρ-T) relation of calcite was precisely determined to 1000 GPa using the decaying shock method, allowing us to predict the post-impact residual temperatures and the degassed greenhouse gas species.
We performed laser shock experiments on targets illustrated in Fig. 1a and obtained the Hugoniot for calcite shocked between 200 and 1000 GPa (Table S1) using the impedance mismatching technique. Typical images recorded by the velocity interferometer system for any re ector (VISAR) for shock velocity (U s ) and streak optical pyrometer (SOP) (see Method section) for temperature (T) are shown in Fig. 1b and 1c. Figure 2a shows the relationship between the shock velocity and the particle velocity (u p ) determined in the present and previous studies. The previous results showed the linear relation; U s (km/s) = (3.71 ± 0.03) + (1.43 ± 0.01) u p (km/s) 6 and the extrapolation is in excellent agreement with the SESAME model.
The present results are well-approximated also by a linear relation; U s (km/s) = (3.79 ± 0.33) + (1.26 ± 0.03) u p (km/s). This linear relation appears to shift downwards parallel to the previous linear extrapolation.
From the present U s -u p data, the Hugoniot pressure-density relation was determined through the Hugoniot conservation relations, as shown in Fig. 2b. Our experimental results demonstrate that the SESAME model overestimates shock pressure by more than 100 GPa or density by more than 0.5 g/cc when particle velocity is above ~6 km/s corresponding to ~12 km/s calcite collision velocity. Since the previous experimental data below 100 GPa agree with the SESAME model, the large gap between the present results and the SESAME suggests that any state changes occur with a signi cant volume change in the 100-200 GPa pressure range. Figure 3a shows the measured optical re ectivity (R) as a function of the shock pressure. The calcite re ectivity along the decaying shock-velocity history was extracted from the VISAR data. The shock pressure is estimated from the measured shock velocity using the calcite Hugoniot determined in this work. Calcite is found to be re ective over ~200 GPa and the re ectivity steeply increases over ~350 GPa due to the metallization. Similar tendency was observed for shocked quartz and some other insulators. Figure 3b shows the decaying shock pressure-temperature relationship on calcite, including the previous experimental results 7 and the EOS models 8 . Below ~350 GPa, the obtained Hugoniot P-T curve follows the predicted melting line 9 . The trend of the P-T relationship changes at ~350 GPa, suggesting that the melt completion of calcite along its Hugoniot occurs at ~350 GPa. Above 350 GPa, calcite would be an electronically conducting uid.
The previous experimental results agree with the SESAME model, but not with the ANEOS and PANDA models. These show that the decomposition reaction of calcite into CaO and CO 2 in the solid phase does not appear in Hugoniot as predicted by the ANEOS and PANDA models at ~80 GPa, because SESAME model does not consider the decomposition reaction. Sekine et al. suggested that the decomposition of shocked MgCO 3 occurred at 107 GPa from their recovery experiment results and the Hugoniot and released state measurements 10 . Considering the binding energies of MgCO 3 to MgO and CO 2 , CaCO 3 may decompose at slightly higher shock pressure than MgCO 3 . Due to the melting and this decomposition reaction, therefore, the previous data points may show a small deviation from the SESAME Hugoniot along the predicted melting line of CaCO 3 . The Hugoniot P-T data in the previous and present studies can be smoothly connected along the melting line. No optical re ectivity observed below 200 GPa, as shown in Fig.3a, is consistent with the presumption that the melting and the decomposition reaction dominate in the 100-200 GPa pressure range. Finally, in the pressure range of 100-350 GPa along the melt line, the reactions of melting, decomposition, and dissociation of calcite may sequentially occur.

Implications for natural impacts: carrier of atmospheric components
The adiabatic release paths from Hugoniot state of calcite were calculated using the Mie-Grüneisen equation of state (see Method section) at pressures of 100, 200, and 300 GPa based on our experiments in Fig. 4a (pink, green, and blue diamonds). The Hugoniot temperature T H and the residual (post-shock) temperature T R are shown in Figs. 4a and 4b, respectively. After calcite decomposed into CaO and CO 2 , the chemical reaction CO 2 = CO + 1/2O 2 occurs depending on temperature. The CO/CO 2 ratio produced from calcite shocked at 100, 200, and 300 GPa were estimated to be 6, 10, and 11, respectively, at a total pressure of 1 atm as a function of the residual temperature according to the equilibrium gas-phase diagram 11 . This result shows that the amount of CO generated from CaCO 3 increases with increasing shock temperature (shock pressure), because CO is more thermodynamically stable than CO 2 at high temperature.
According to the previous sample recovery experimental results, calcite should be completely degassed at shock pressures above 70 GPa, although there are some variations in the threshold pressure dependent Chicxulub impact reported the total gas amount by the shock-induced volatilization as (0.8-7.6) × 10 16 mol 15, 16 . If the CO/CO 2 ratio from shocked calcite is ~10 at an impact pressure of 200 GPa (in a case of our results), the amounts of CO and CO 2 gases produced by the impact become (0.7-6.9) × 10 16 mol and (0.1-0.7) × 10 16 mol, respectively. According to the previous studies using a light-gas gun, the CO/CO 2 from single crystal calcite shocked at peak shock pressures of below ~100 GPa was estimated to be about 0.1 17 . The residual temperature of the present study was much higher than those of the previous studies. Kawaragi et al. (2009) reported that a CO dominant gas (CO/CO 2 = 2.02) increased ~3 K in temperature compared with that in case of only CO 2 in the Chicxulub impact event 18, 19 . CO is known to produce CH 4 and tropospheric O 3 with the strong greenhouse effect through photochemical reactions and has more intense indirect-radiative-forcing to the heat uptake than CO 2 20 . Our results suggest that the indirect greenhouse effect of CO should be taken into account in considering the climate change after the high-speed planetary impact events. Figure 4b shows that CO/CO 2 ratio in released gas from calcite and the residual temperature as a function of impact pressure. The collision velocity required to generate shock pressures of 100-300 GPa was calculated to be 7-14 km/s for ordinary chondrites 21 . These velocities are typical in planetary impacts 22 . The results suggest the CO gas was produced more signi cantly in large-scale planetary impacts than the previously estimated. In natural impacts, CO production proceeds more e ciently than our experimental conditions, since meteorites should relatively increase temperature due to certain amounts of porosity 23 . Especially, it is possible that more CO than expected had been supplied on the Earth by planetary impacts during the LHB period.
Our results may prompt a reconsidering of whole planetary evolution processes on the planetary surface, especially the estimation of the atmospheric compositions and the greenhouse effect, and the seawater temperature. Furthermore, a large amount of CO generated from impacted carbonate rocks may contribute to model the global warming, the planetary atmosphere chemistry, and the biological mass extinction 24 on the Earth triggered by bolide impacts.

Materials And Methods
Hugoniot and temperature data of calcite The Hugoniot relations on calcite have been determined experimentally at pressure below 95 GPa 6 . On the other hand, only the theoretical model (SESAME 7331) is available at pressures above 200 GPa, and there are some data on the U s -u p relations from the calculation models (SESAME, ANEOS, and PANDA 8, 16 ). Thus, we require experimental data above 100 GPa to understand the dynamic behaviors in natural impacts. Extreme pressure and temperature conditions over 200 GPa and 5000 K will be generated at impact velocities of chondritic meteorites over 10 km/s using the Hugoniot relations 21,25,26 .
The shock temperature on calcite was measured up to a pressure of 160 GPa in the previous study using gas gun 7 . Temperature data of shocked calcite above 200 GPa are required to understand the reactions on calcite. Extrapolations of the temperature calculations by the ANEOS and PANDA models showed high temperatures relative to the SESAME data at high pressure over ~100 GPa. For example, the temperatures at 300 GPa by the ANEOS and the PANDA models are 8000 K and 3000 K lower than that by the SESAME model, respectively.
Laser shock experiments and optical measurements of pressure and temperature The laser shock experiments on calcite were conducted with GEKKO XII laser facility at Osaka University. The GEKKO facility is a neodymium-doped glass laser (the basic wavelength: ω 1053 nm) system. The total 12 beams of three 2ω beams with a diameter of 1000 μm and nine 3ω beams with a diameter of 600 μm were used in this experiment. The temporal shape of the laser pulse was approximately square with a 2.5 ns duration [full width at half-maximum (FWHM)] and rise and fall times of 100 ps each. The total energy and intensity of laser were 300−800 J and 10 12 −10 13 W/cm 2 on the target in the present study, respectively.
We used ~1 cm natural single crystals of calcite CaCO 3 (density: 2.71 g/cm 3 ). The crystals were polished to be 4 mm square plates (500 μm thick) by a manufacture. The target assemblage used in the present study is shown in Fig. 1a. All materials of the target were glued with an ultraviolet curing resin. Typical resin was ~1 μm thick.
Shock velocity and the thermal emission at the shock wave front of samples were measured by using the Velocity Interferometer System for Any Re ector (VISAR) 27 and Streak Optical Pyrometer (SOP) 28 , respectively. As quartz becomes a high optical re ectance above 150 GPa 29 , we can measure the shock velocity. And if calcite becomes re ective at high shock pressure, we can measure both shock velocities on the target simultaneously. Although refractive index of a crystal varies with the crystal orientation, the refractive index used in the present study is 1.66 and 1.55 for calcite and quartz, respectively. Velocity sensitivities [velocity per fringe (VPF)] of two line-imaging VISARs (line-VISARs) in this study were 7.39 and 4.47 km/s in vacuum. Based on the quartz U s measured by VISAR, the pressure (P), particle velocity (u p ), and U s for Al were obtained by the impedance matching for the interface between Al and quartz. The experimental error of a shock velocity was less than 5 % of the velocity sensitivities.
After that, the u p of calcite was determined by the impedance mismatching method 30  respectively 31,32 . Shock pressure and density (ρ) in calcite were determined from the U s and u p using the basic conservation relations 33 . Uncertainties in impedance mismatching method were propagated through a Monte Carlo method 34 , hence one-sigma uncertainties were taken to be standard deviation of the distributions. In addition, the re ectivity of shocked calcite was simultaneously obtained from VISAR measurements at the probe laser (532 nm). The re ectivity of shocked quartz was cited from a reference 35 .
The SOP measures the thermal emission from the shock around a 455-nm wavelength with a narrow bandwidth (38-nm FWHM). The brightness temperature is determined using the gray-body Planck spectrum: where h is Planck's constant, c is the speed of light, λ is the wavelength of the radiation, T is the temperature of the Planckian source, K B is the Boltzmann constant, and ε is the emissivity of the shocked matter. This equation can be reduced to a practical form at the narrow band wavelengths 28 , as described by Miller et al. (2007).
where T 0 and A are calibration parameters that depend on the SOP optical system and C is the CCD (charge-coupled device) counts obtained by the streak camera system. The emissivity ε is calculated from the Kirchhoff law, ε = 1 − R, where R is the shock-front re ectivity. We used quartz as the standard material because the relations between T-U s and R-U s have been precisely measured 29,36 . We determined the value of A from several calibration shots on quartz to measure U s and SOP simultaneously. More details of the SOP measurement can be found in the previous studies 25,26 .

Estimation of an adiabatic release state
The adiabatic release state of shocked calcite was determined to estimate the residual temperature at pressure release. Release isentrope was approximated by using Mie-Grüneisen equation of state 37 . At rst, the following Eq. (1) was given by combining the Rankine-Hugoniot equations with Mie-Grüneisen EOS where P, E and V are the pressure, internal energy and the speci c volume, respectively. Grüneisen parameter Г was assumed to be a function of V only, using Г 0 = 0.57 38 , The subscript H, S, and 0 denote the reference Hugoniot, release adiabat, and the initial state, respectively. The internal energy on an isentrope was determined by solving the ordinary differential equation (ODE), which gives where subscript 1 refers to shocked state. Substituting Eq. (3) into Eq. (1), we determined the adiabatic release state of calcite using the measured P 1 , V 1 , and the obtained calcite Hugoniot.
The residual (post-shock) temperature T R was given by Eq. (4) 39 where T 1 and V 2 denote the initial shocked temperature and residual speci c volume, respectively. Figure 1 Schematic setup and typical measurements by VISAR and SOP in laser shock experiments. a, The target assembly consisted of poly-propylene (CH), aluminum (Al), crystalline α-quartz (Qz), and calcite (CaCO3) from the drive laser side (GEKKO XII). The thickness was 30, 40, 50, and 500 μm for CH, Al, SiO2, and CaCO3, respectively. The CH was used as an ablator to generate a shock wave, and Al as the base plate, and SiO2 as the reference material for the impedance mismatching analysis and the temperature determination. The VISAR side of SiO2 was coated with an anti-re ected (AR) layer for a probe laser at the wavelength of 532 nm. Since AR coating could not be available for calcite and SiO2 has almost the same refractive index with calcite., the last SiO2 after calcite was used for the window material. b and c, VISAR line-imaging velocity record and inferred shock velocity (yellow curve, scale on right).

Figures
Corresponding SOP raw image and count pro les considering the emissivity (1-R) (white curve). The time zero is set to the arrival time of shock wave on the Al/CaCO3 and Al/Qz interfaces. The vertical width in images is about 300 μm.

Figure 2
Hugoniot relations of calcite. a, The Us-up data of this study are shown as light blue diamonds and connected linearly as the broken line. The error bars of this study are smaller than the symbol size. The open green circles indicate the experimental result 6 by Kalashnikov, (1973). The orange, red, and purple solid lines represent SESAME, ANEOS, and PANDA models, respectively 8, 16. b, The pressure-density data determined in this study are shown as light blue diamonds and connected by the broken curve, and compared with the SESAME model as orange solid curve.

Figure 3
The re ectivity at 532 nm and the temperature as a function of pressure. a, The green, blue, and orange plots are shots 42246, 42180, and 42232, respectively. b, The light blue plots indicates our result from the decaying shock experiment. The orange, red, and purple solid lines represent SESAME, ANEOS, and  The release isentrope and the residual temperature and the CO/CO2 ratio of shocked calcite as a function of pressure (impact velocity). a, The release isentrope paths (dotted lines) of shocked calcite from 100, 200, and 300 GPa in the magni cation of Fig. 3b. The diamonds of pink, green, and blue mean the start and the end points of releasing. b, The residual temperature (blue; left axis) and the CO/CO2 ratio (red; right axis) as a function of pressure. The impact velocity was calculated using the Us-up for the ordinary chondrite 21 proposed from Zhang and Sekine, (2007) at the vertical impact. The CO/CO2 ratio was estimated using the equilibrium gas-phase diagram as a function of temperature 11.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. v2NatComEarthUmedacalciteSM.docx