In 2009, the Phoenix Lander first detected high levels of perchlorate (ClO₄⁻; 0.62–0.67 wt%) and high ClO₄⁻/Cl⁻ ratios of 6.13 in three topsoil samples from the Mars arctic region (68.2188°N, 125.7492°W). This finding is intriguing because the perchlorate content is unexpectedly high. If the total chlorine is assumed to be composed of ClO₄⁻ and Cl⁻ in the samples, then ClO₄⁻ accounts for 85% of the total Cl and seems to be the primary species of chlorine on Mars.

Subsequent in situ investigations at other landing sites further highlighted the uniqueness of the Phoenix soils. The Curiosity rover at the Gale Crater (in the equatorial region) found ubiquitous oxychlorine species in aeolian deposits, soils and drilled samples of sedimentary rocks. In two aeolian samples, ClO_x⁻ comprises 20% of the total Cl in sample "Rocknest" and 15% in sample "Gobabeb". A similar estimation of 4% is also made for Viking soils through reanalysis of the Viking in situ analysis and experiments. In Martian meteorites, the ClO₄⁻/Cl_total proportion is 20% for "EETA79001" and 0.1% for "Tissint". In samples from terrestrial (hyper)arid analog sites (e.g., the Atacama Desert, the Antarctic Dry Valleys), the highest fraction (0.08%) is found in the Atacama Desert samples.

Therefore, how did the high ClO₄⁻ and particularly high ClO₄⁻/Cl⁻ ratios in the topsoil of Phoenix occur? Did it originate from high yields of ClO₄⁻ from Cl⁻or is it formed by ClO₄⁻ enrichment? Is ClO₄⁻ a primary phase on Mars? Are the high ClO₄⁻/Cl⁻ ratios present elsewhere on Mars?

We prefer an enrichment mechanism of ClO₄⁻ in the Arctic environment because the direct conversion rates of ClO₄⁻ from Cl⁻ are low in almost all simulation experiments, and no high ClO₄⁻/Cl⁻ ratios are detected elsewhere in situ on Mars, or in Martian meteorites, or in terrestrial analogue samples. Furthermore, deliquescence processes (hygroscopic dissolution of salts or the melting of salt-ice mixture eutectics to form a solution) seem to be the key to account for the perchlorate enrichment at the Phoenix site. However, without further quantifications, it is unclear how the deliquescence only results in perchlorate dissolution without dissolving other coexisting salts.

To evaluate the influence of deliquescence on ClO₄⁻/Cl⁻ ratios on the Martian surface, it is necessary to first understand the solid–liquid phase transition of  perchlorate-chloride-water ternary system at low temperature, which is yet unavailable. Therefore, we start with building the tool. We choose to focus on Mg- and Ca- perchlorate-chloride-water systems initially due to their strong hygroscopic characteristics and the low eutectic T compared to the Na- or K- system. We experimentally determine the essential solubility data at some low-T conditions, then   with other published solubility data at different temperatures, we construct and validate a thermodynamic model for the temperature range of 180 K ~ 300 K.

With the new thermodynamic model, we construct the deliquescence liquefaction surface in Mg- and Ca-perchlorate-chloride-water systems (Fig. 1) and determine the deliquescence RH-T boundary for ClO₄⁻/Cl⁻ fractionations. We find that there is always a specific range of relative humidity (RH) and temperature (T) in which perchlorate will dissolve preferentially over chloride, forming a solution rich in perchlorate but depleted in chloride, resulting in a brine with a ClO₄⁻/Cl⁻ fractionation signature. Meanwhile, the majority of chlorides remain in the form as solid salts. Above the upper boundary, both ClO₄⁻ and Cl⁻ salts will dissolve; below the lower boundary, neither salt will dissolve. No fractionation of ClO₄⁻ and Cl⁻ would occur in either case.

Fig. 1. Deliquescence surfaces of mixed hydrated Mg- and Ca- perchlorate/chloride salt mixtures calculated using the thermodynamic model.

Then, how would the deliquescence RH-T boundary for ClO₄⁻/Cl⁻ fractionation apply to the Phoenix site, Gale Crater, and Earth’s Antarctic Dry Valley (ADV), where ClO₄⁻/Cl⁻ have been reported? We find that the Phoenix landing site can sufficiently support the deliquescence of Mg- and Ca- mixed salts and result in ClO₄⁻/Cl⁻ fractionations in the brine, but the Gale Crater and Antarctic Dry Valley would not. This results are in good agreement with in situ sample analysis results.

Furthermore, we explore whether anywhere else on Mars has the potential to support ClO₄⁻/Cl⁻ fractionation, as observed at the Phoenix site. Using the LMD GCM atmospheric circulation model, we determine a global map of the duration of RH-T environments that can lead to the formation of ClO₄⁻/Cl⁻ > 0.2 brines in a Martian year (Fig. 2). The results show a clear difference between the northern and southern polar regions on Mars. The Martian north polar region provides the most prolonged suitable RH-T conditions. In contrast, the Martian south polar region (south of 60°S latitude) is similar to the equatorial region, which does not support the formation of ClO₄⁻/Cl⁻ > 0.2 brine. Some large basins, such as the Acidalia-Chryse Plain, the Utopia Plain, the Hellas Basin and the Argyre Basin, would support ClO₄⁻/Cl⁻ fractionation, although to a lesser extent. The positive topography, such as the southern highlands and Olympus Mons, is unfavorable in supporting ClO₄⁻/Cl⁻ fractionation.

Fig. 2 Global distribution of the fractional duration in a Martian year of RH-T conditions leading to the formation of ClO₄⁻/Cl⁻ > 0.2 during deliquescence on the surface of modern Mars. White areas do not support forming the ClO₄⁻/Cl⁻ > 0.2 brine at all.

Our work highlights the unique environment of the modern Martian arctic region. The high ClO₄⁻/Cl⁻ ratios found in Phoenix soil are not universal on Mars, and chloride salts should be the major and primary phase of total chlorine. Nevertheless, a high ClO₄⁻/Cl⁻ signature may be used to indicate deliquescence processes in cold and arid environments on Mars.

Our work demonstrates that understanding oxychlorine species on Mars and their environmental and climate implications requires integrating multiple research methods and multidisciplinary collaborations. This is also the original intention of this study to comprehensively adopt experiments, thermodynamic simulations and climate simulations to provide insights for in situ observations.

In particular, cryogenic brine phase diagrams are valuable tools for quantitatively evaluating deliquescence processes on Mars. It provides a new perspective from the “salt-to-brine” rather than the “brine-to-salt”. Pursuing the understanding of cryogenic salt-brine systems in extraterrestrial bodies provides a new active field for traditional phase diagram research. We are currently working on the chlorate-chloride-water ternary system at low temperatures to obtain new physicochemical constraints on the perchlorate-chlorate-chloride-water system for Mars.