A. J. S. Rayl has an excellent article, with just the right balance of humor and fact, on the Phoenix perchlorate business here. I was waiting for her to post before I added any commentary about it. I'll just add a few other random facts that I got from the press briefing and from email and phone conversations afterward. (If you haven't checked out the recording of my Ustream presentation today, watching that might make the following discussion a bit clearer.)
One of the more surprising things I heard in Tuesday's press briefing was TEGA instrument lead Bill Boynton saying that they had inconclusive evidence of the presence of perchlorate in their first sample because they hadn't been looking for chlorine during their heating cycle because they hadn't expected to see it. I had an instinctive negative reaction to this remark: how on Earth could you possibly send a mass spectrometer to Mars and then not look for some masses? There's actually a very good reason, and I'll try to explain it.
I usually think of TEGA as a mass spectrometer, which is an instrument that takes in gases and then splits them up by their mass and spits out a spectrum, basically a graph of mass of a molecule versus a count. Certain molecules or atoms produce peaks at certain points. Oxygen -- one of the things they'd be looking for -- would produce peaks at masses of 32 daltons (for molecular oxygen, O2) and 16 daltons (for a neutral oxygen atom). To "see" chlorine, they'd have to be "looking" for a mass of 35 daltons. I always assumed that mass spectrometers just kind of took a continuous count of all different masses of molecules, but that's not how TEGA works. Boynton told me that TEGA actually has to scan across the masses, and a full scan across its entire mass range takes 5 minutes. That's not too long, right?
Yes and no. TEGA isn't just a mass spectrometer. Before and during its mass spectrometry run, it also does something called Differential Scanning Calorimetry. To do mass spectrometery of solid samples, the first thing you need to do is to make a gas out of your solid, and to do that TEGA heats the solid sample to 1000 degrees Celsius. But before it does that, first it does calorimetry on two earlier heating cycles to lower temperatures.
These two lower heating cycles drive off all the water from the samples. They do that for several reasons. First, the presence of any water tends to swamp their sensor and generally fill the interior of their instrument with annoying sticky water vapor, so they want to totally dry out the sample before they try to do the thorough mass spectrometry. Second, water is, after all, one of the more interesting things they can study; they do lower heating cycles first to study the quantity of water present, and, if there's enough of it, the ratio of heavy water (water made with heavy hydrogen or deuterium, which is a hydrogen atom that has a neutron in the nucleus along with the usual single proton) to regular water, which gives them insight into how much atmosphere Mars has lost throughout its history.
To get at the amount of water in their sample, they slowly apply heat to the sample. The sample will heat up until it gets to the vaporization point of water, at which point the temperature will remain constant until all the water has been turned to gas (it takes lots of energy to turn water from solid to gas and as that's happening, the temperature does not increase). TEGA studies the stepwise way that the temperature rises and then stalls to diagnose the presence of not only water but also lots of other volatile species.
As different gases can evolve rather quickly during these early, low heating cycles, the TEGA team can't afford to do the full five-minute mass spectrometer scan during this calorimetry phase -- instead, they focus on a few masses of molecules that they predict will come off during the heating cycle, such as water and its cracking products, hydrogen and oxygen. The list is only 10 or 20 masses long, and the 35-dalton mass of chlorine was not initially on their list. So they didn't look for it in their first sample. They did, however, notice that oxygen was given off at reasonably high temperatures, a fact that made some on the team wonder if perchlorate might be the culprit.
Then, MECA got two samples that seemed to contain perchlorate. Needless to say, when the MECA team announced their result to the science team generally, the TEGA team decided they needed to be looking for chlorine. So they reprogrammed TEGA to look for chlorine during those early heating cycles. That program was in place when they analyzed their second sample, Wicked Witch -- and they didn't see any chlorine. More head-scratching ensued. That's where it stands now.
There are several possibilities to explain the apparent contradiction between the MECA detection and the TEGA non-detection. One: MECA's samples contained perchlorate, while TEGA's didn't. This is a bit tough to explain given that two of the samples, MECA's "Sorceress" sample and TEGA's "Wicked Witch" sample, came from pretty similar locations. But the distribution could be patchy. They can best test this possibility by taking one scoop of soil and delivering some to MECA and some to TEGA, an operation they have not yet tried.
Two: MECA could somehow be contaminated while TEGA is not. There are various reasons why this possible explanation is almost vanishingly unlikely, but they are re-looking at potential contamination sources, just to be thorough.
Three: the perchlorate could be one that does not give off chlorine at high temperatures. To address this possibility, they need to figure out which perchlorate salt they are dealing with -- potassium? Magnesium? Et cetera. They won't do that until the analysis of the MECA data is much more mature -- a process that may take months or more. But they'll get there eventually.
I had more to say on MECA -- but this is absolutely all I can write for tonight!