I’m giving a fun talk later this month at Cornell. I am (an invited!) speaker at an x-ray diffraction workshop. This will be a particularly fun meeting as it’s not a presentation of what everyone has been doing. Rather, we’re supposed to give talks about what we would be doing if we had more light! The CHESS facility at Cornell has been pushing the frontier of photon based science with the idea of an Energy Recovery Linac. Should it work (and be built), we could expect an increase in around 1000 times the flux over current continuously operating sources (as opposed to pulse sources such as the LCLS).

There are several good candidates for material to include in such a talk, but some of the most obvious are experiments which we have tried that have *almost* worked.

There’s one particular example that has frustrated me to no end. We attempted to see get a signal from a Pt surface undergoing an oscillating production of CO2 from CO and O2. That system itself is just fascinating and part of the reason Ertl won a Nobel prize in chemistry a few years ago. What we’re trying to do is develop a new tool that could be used to look at similar systems (particularly though those at higher pressure where we current techniques have more difficulty).

We tried the experiment once and got some promising early results. Things looked pretty good for a first attempt. We went back though later and spent a week on this system with limited success. It was hugely disappointing (though that entire run was still a success as we managed to reproduce the initial result of a different experiment). We struggled to understand what went wrong, and I think I largely understand the problem. However in the face of getting it “to sort of work” once and “not to work at all” once, we can’t really write it up and say, “LOOK, COOL SCIENCE!” As often happens with experimental science in general, when we go back to verify and repeat, we end up “experimenting” ourselves right out of a publication instead of into one.

So this all happened 18 months ago. Why am I frustrated now? Well, before I can give a talk really saying, “here’s something we tried that almost worked,” I need to make absolutely certain (AGAIN) that it did in fact not work. So I’m back running over the same old ground trying to repeat the analysis in an independent fashion (obviously just looking at my old notebooks and repeating the process exactly will give the same answer. The point is to approach it fresh and unbiased).

It’s frustrating in that I’m going through all the same emotions as it doesn’t work; frustrating in that it almost worked; frustrating in that you almost see the signal, maybe you do; frustrating that it’s so close, but just beyond what we can do conclusively enough to say, “yes or no.” Frustrating because it’s “MAYBE” and in (good) science that always means NO, NOT YET.

How many times have I repeated, “Science is hard.” on this blog? sigh... Fun, but not easy.

I should go back to posting nice pictures on the blog. We’ve got several new ones.