Fieldwork in the Atacama Desert, Chile

Sunday, 24 December 2017

Latest paper summary: Effect of Hydration State of Martian Perchlorate Salts on Their Decomposition Temperatures

Here’s an early Christmas present for everyone (so both of you who read this blog), our latest paper (my first out of this project) is now published online and completely open access for anyone to read.


If you can’t be bothered to read it, here is a summary:

Since the Viking landers back in 1976, numerous missions have gone to Mars and attempted to find organic matter on the surface. However, to this date, none have succeeded in finding anything other than simple chlorinated hydrocarbons. We expect there to be a detectable amount of relatively complex organic molecules in the Martian near surface from meteoritic input (meteorites are full of organic matter), hydrothermal processes and maybe even left behind as evidence of extinct (or, unlikely but not impossible, extant) life. This lack of detectable organics was always a bit of a mystery…

Note: the presence of organic matter does not necessarily mean the existence of life, organic molecules are just those which have carbon and hydrogen and they can be formed by numerous non-biological processes – including in deep space.

However, since they were accidentally discovered by the Phoenix mission back in 2008 we’ve known that that there are these salts on Mars called perchlorates. These salts are very rare on Earth, only being known to exist in significant quantities in the Atacama Desert and the Antarctic Dry Valleys; two of the driest places on Earth.  As the name suggests perchlorates have 4 oxygens to each chlorine ion and so are very highly oxidising. While (relatively) stable on the Martian surface, as soon as these are heated within the sample analysis oven of a Martian lander or rover these literally explode, giving off loads of oxygen into the analytical system (which should be a vacuum). This means that any organic matter which may be in the Martian soil is combusted, reacting with the oxygen present and breaking down to carbon dioxide and carbon monoxide,  and the few surviving scraps are heavily broken up and react with the chlorine in the perchlorate, making small chlorinated hydrocarbons. Because of this destruction the molecules we are interested in cannot be detected – This is the perchlorate problem.

My attempt to summarise the Perchlorate Problem on the Sketch Your Science wall at AGU this year

My job as a postdoc researcher at Imperial College is to attempt to find a way around this ‘perchlorate problem’ by researching how these perchlorates (and other similar minerals) react with organic matter. The first step of this, however, is to try to understand a little more about the Martian perchlorates themselves. And this is what this recently published body of work was concerned with.

Since it landed on the Mars in 2012 the Curiosity Rover has drilled 15 holes into the Martian surface and analysed the drilled sample to see what it is made of. One of the techniques used is looking at the gases given off by the sample when it is heated. However, every time the Curiosity Rover has analysed one of these sample holes, the temperature that oxygen gas is released from the sample is very different (Sutter et al., 2016). Most of the oxygen given off is believed to have come from these perchlorates (based on other gases given off at the same time). It has been found by previous laboratory experiments that different types of perchlorate salts (magnesium perchlorate, iron perchlorate, calcium perchlorate, etc) break down and give off oxygen at different temperatures (Glavin et al., 2015), and also break down at different temperatures depending on whether other minerals which may act as catalysts are also present in the soil (Sutter et al., 2014). Therefore it has so far been concluded that different types of perchlorates and/or catalyst minerals have been present at each drill site.

Curiosity on Mars (credit: NASA)
The first 14 drill holes Curiosity drilled on Mars (credit: NASA)

This is a bit problematic. It is believed that most perchlorate forms in the Martian atmosphere and falls out onto the surface (as it does on Earth) and so there is no known mechanism for having different perchlorates in different places – they should all be pretty much the same across the Martian surface.

It is known that perchlorates are highly hydrophilic – they will suck up any water available (they are used as drying agents in industry) and can change their hydration state (how many molecules of water are bound to each molecule of perchlorate) really easily. The temperature and humidity of the Martian surface changes massively throughout the Martian day and year. Photographs taken on Mars by Phoenix show growing blobs on the lander struts which have been interpreted as perchlorate goos as they absorb water throughout the Martian day. So, my boss, Mark Sephton, had an idea that the hydration state of the perchlorates may actually have an important effect on their breakdown temperature and set me to investigate this.

'Blobs' of perchlorate 'goo' moving across the struts of the Phoenix lander (Renno et al., 2009)

I did this by taking a single one of the perchlorate salts (magnesium perchlorate) and drying it out in the glassware drying oven to drive off water in an attempt to reduce its hydration state. After three weeks I removed some sample, flash heated it and analysed the gases given off using pyrolysis-GCMS at 100 °C temperature steps from 200-1000 °C to create oxygen (and other lesser gases) release-temperature profiles. A sample of the perchlorate was also left out on the lab bench to rehydrate and analysed in the same way after 24, 48 and 72 hours of exposure and rehydration. This whole experiment was repeated after another week and then after a fortnight so oxygen release profiles were created for 3, 4 and 5 week drying times and corresponding 24, 48 and 72 hour re-exposure samples.

It was confirmed that these drying and re-exposure experiments were definitely changing the hydration states of the perchlorates by testing them (by X-ray diffraction) next door in the Natural History Museum. This involved transporting them through crowds of tourists under an inert atmosphere (in a sealed lunchbox filled with nitrogen gas) to prevent further hydration state change.

What we found was that different hydration states did indeed affect the temperature of decomposition so that the oxygen release profiles were as different for different hydration states of this one kind of perchlorate as the previous studies had found for all the different kinds they tested. This, therefore, gave a much simpler answer to this puzzle: Curiosity has been analysing different samples with different hydration states of perchlorate in them. This makes sense as the different samples drilled have been analysed at different times of the Martian day and year and samples spend varying amounts of time stashed inside the temperature-controlled innards of the rover before they get round to being analysed – which could allow further changes in their hydration state. Our data show that is it possible for numerous hydration states of perchlorate to exist within a sample and this leads to multiple peaks in the oxygen release profile, some of the Martian samples have multiple peaks in their oxygen release profile and so we suggest that this is due to unstable mixtures of perchlorate hydration states being present on the Martian surface.

This figure from the paper shows how the data from this study (A) compares to select samples analysed on Mars by Curiosity (B) and various types of perchlorate from a study by Glavin et al. (2015) (C). It can clearly be seen that the variation in the samples of magnesium perchlorate that were dried out for various numbers of weeks and analysed in this study is almost as great as the variation in different perchlorates from the Glavin et al. study. So, different hydration states may offer a simpler explanation for the variation seen in the Martian samples. 
We tried to see if the oxygen release profile of the Martian samples (and therefore, based on our findings, the perchlorate hydration state) corresponded to the climate conditions at the time they were sampled at but did not find any relationship. There did seem, however, to be a vague suggestion in the data that they were related to the time of the year that they were sampled at, although more data would be needed to be sure about this.


All in all, we conclude that the hydration state of perchlorate salts is yet another thing to make making sense of Martian data yet more complicated.

No comments:

Post a Comment