Siliceous and Delicious

Monday

For all of its frustration, the last year actually ended on a high note. As I was sitting in the end-of-year lab meeting, reflecting on my disappointment with the year gone by, I got a phone call from Jon Wilson, my former labmate. In a stunning and completely surprising turn of events, he told me of a small research project he and Woody had decided to work on—based on the very topic on which I wrote my first ever term paper at Harvard, for Andy’s paleobotany class. And they want me to be on the paper. Best of all, they want to do all the grunt work of analysis, measurements, and so forth, leaving me to do some of the broad reading, background research, and paper-crafting.

That phone call was a huge ego-boost, on a couple of different levels. First of all, they had decided to write a paper on a topic I found interesting, and had at the time been very disappointed to find nobody else seemed to care about. Now, all of a sudden, two people whose judgement I respect showed an interest in a matter I had obsessed over for weeks! Secondly, they had asked me to help them—which means they must have felt that I had something valuable and worthwhile to contribute. That gesture alone was a much-needed relief after many months of feeling completely useless. It’s a small, small step, but even this feels worlds apart from floating in space without any connections to the rest of the world.

So, I decided to start the new year on a positive note, and began by re-reading the term paper that I wrote for paleobotany those many moons ago, which I had sent to Jon and Woody before leaving for Vienna. It got me excited about the topic again, and got the rusty wheels of the old brain crank grinding slowly into action. I wrote a lengthy email responding to the notes Jon had sent me, with a few ideas and some pertinent references. Then there was a mountain of other email to answer—after all, I’ve been off email duty for two weeks… Got a long-overdue email of my to-do list, which was to introduce myself to my Austrian Scientist Mentor. The program started back in September (!), but the vicious combination of my shyness and my conflicted relationship towards (a) Austria, (b) returning to Austria, and (c) my PhD and place in academia, had paralyzed me into months of inactivity… and today I leapt over my own shadow, bit the bullet, and just wrote the damn thing. It’s a ridiculous hurdle to celebrate, but it’d been weighing me down and I’m glad I did it.

I booked time on the FIB for Friday morning, giving me a couple of days this week to get samples ready for zapping.

Tuesday

Started by trying to find out about what fluorescent substance I should use to try messing around with the CLSM approach in my diatom morphospace project, but could find nothing. The methods the germans suggested are complicated, and one of the major points of feedback from my G3 progress report meeting was to do things more simply and in a generally more slapdash fashion. So, I just emailed Jacques to ask his opinion. We’ll see what he comes up with.

Sneaking off to a coffee break paid off today: I bumped into Rama, the postdoc who works on silicon isotopes, and talked to him again about my radiolarian-diatom silicon isotope measurement idea. I managed to enthuse him sufficiently to secure a further meeting, so perhaps this project will actually materialize after all! I need to find out if I can get samples for this from Dave, though, because otherwise I will need to make a request from ODP and sample preparation would be a pain.

Jacques wrote back almost immediately and suggested I steal some of the fluorescent dye from his lab. I went and filled up a little vial of a mysterious dark red powder labeled “Fluorescein”. Spent a good hour trawling the internet for any sort of information on how to actually use this stuff (it’s a very, very common lab compound), but there was absolutely nothing. It’s apparently one of those things that’s just so obvious, nobody will actually tell you how to use it. Well, I’ll just have to mess around with it and figure it out myself. Let’s see what happens.

Well, I used too much fluorescein, and by the time I was finished rinsing and filtering the sample there was very little diatom material left—I only found two or three shell fragments on the slide I prepared—but it worked! I booked some last-minute time on the CLSM, took my smear slide down there, and boom! The frustules fluoresced when zapped with a 488 nm laser. There’s lots of details to work out, less fluorescein to be used, and more care to be taken when cleaning the sample afterwards, but this is quite a success.

While I was pottering about in the lab, I also prepared an SEM stub for Friday’s FIB session. All in all, I got a lot done today!

Wednesday

Broke with my schedule for one last day—after DSA this morning, spent some time sketching out a brief (2-page) outline of my ideas on the silicon isotope study I had talked to Rama about. After lunch, ambled down to his office and walked him through it, and he seemed to really like it! It’s a simple idea, and although it rests on many assumptions, it’s potentially quite doable, and quite interesting. As with everything in science (so I’ve learned), a great many things will have to work out just right for this project to materialize, but the key thing is that Rama agreed at the end of our conversation that it was worth giving it a shot. It was worth sacrificing that time away from the rigid prescriptions iCal had waiting for me: it felt like real science! And the alluring foretaste of working on a project I would actually enjoy is a sweet, sweet thing.

One of the things that will have to work out is getting samples for this venture. I suggested to Rama that we look at the record of the last 5 million years of the Benguela upwelling system off the southwestern coast of Africa. Partly this is because there is a number of other proxy studies looking at productivity at that time and place, and partly it’s because one of the most recent studies published on it was written by Dave, my MSc advisor in Berlin. My hope is that he will both have left over samples and be willing to send them to us to do the analysis! I’m not holding out too much hope, as his response before xmas was a little lukewarm, but there’s a possibility he was just in a rush to go gift shopping and will warm up to the idea now that the new year is upon us. We’ll see—I sent him another email describing what we wanted to do, and asking him if he was interested, and if he had samples to spare.

For the last part of the afternoon, decided to follow the schedule and spend a little bit of time on the old (and mostly discarded) diatom crushing project. One of the things long on my list had been getting a smoother edge on the glass coverslip I was using as a barrier against which to press the diatoms. One idea that had come up was to simply heat the edge of the coverslip in a flame, and see if the edge becomes smoother by melting a little and then re-solidifying. With my new-found lab confidence, off I went to try this little experiment. The result of my investigations: it seems that coverslips vary in how cleanly cut their edges are. Most seem to have two straighter edges and two rougher edges, presumably because of the way they’re made. Furthermore it seems that all coverslip edges have one rough, ragged side and one smooth, clean side. I was just unlucky enough last year to glue my coverslip to the slide with the ugly side facing up. So although the flame annealing exercise was fun, I think it probably was hardly necessary. I just need to glue the coverslip the right way.

As I was wandering about the lab donning my white lab cat and a pair of purple nites in search of a bunsen burner, I bumped into Andy, who asked how things were going. I gave him the progress report in the abstract, and decided to also mention the plans for the silicon isotope project I had just been hatching with Rama. It seemed to pique his interest, and he asked several good questions until he seemed to get what the point was behind the whole thing—and then gave his “well-I-think-that-would-be-quite-interesting” stamp of approval. Inside, I pumped my fist and loudly and exclaimed: “cha-cha-ching”!

In summary, then, another good day, progress made on the silicon isotope project front, including tentative approval of the project by my advisor, and insights gained on the intricacies of glass coverslips. Who knew science could be so titillating!

Thursday

Today began with a challenge thrown my way: Dave sent a reply to the email I sent yesterday, in which I had asked him if he wanted to collaborate on the silicon isotope project and supply samples from the Benguela area. His response was positive:

I think doing rad silica is an excellent idea and one I once toyed with doing together with Cinzia Cervato when I was back at the ETH.  Never went anywhere as we didn’t have Thierstein’s support but it is potentially very hi-impact science.

But he went on to say that he felt the Benguela material would not be the ideal starting point for making such measurements. Rather, he suggested starting by trying to deconvolve the signal in the modern, and how it is preserved in the sediments, by examining plankton tow material from different depths to constrain where the different taxa in the sediment are coming from, and then looking at Holocene core-top sediments. And he then suggested I contact Kozo Takahashi, the Japanese don of the radiolarian world, who has studied extant radiolaria and may be able to help. So, all in all it sounded a bit like he didn’t think my particular approach was a winning ticket—and I do see his point. The details of how the signal is preserved need to be worked out before this can be used as a reliable proxy. But I suppose my thinking (beyond wanting to get this project off the ground before Rama leaves Harvard!) is that it would be worthwhile first establishing whether there’s a signal there to be explained, and then seeking to explain the details of how it is preserved—rather than spending time and effort establishing how silicon isotope compositions behave from water column to sediment only to find there is no intelligible signal in the sediments anyway.

That said, Dave’s email did end on a positive note:

Happy to help in any way possible with the project should it take place, tho I think you should primarily choose a water-column specialist as a partner.

This leaves me unsure of how to progress. Sure, it would be great to study the water-column aspects of silicon isotopes in radiolaria, but that seems like it would be a far more time-intensive project than I can commit to at the moment. Perhaps I can assuage Dave’s concern by suggesting starting these studies in parallel, looking at a modern water-column-to-plankton setting at the same time as analyzing a fossil time series. This might ultimately be a good thing—if I could get my hands on those Benguela samples, I could get a study underway with Rama, and at the same time lay the foundation for further work and building contacts with other radiolarian workers. Especially in Japan, which would be a radically kick-ass place to visit.

The morning’s schedule called for work on the pre-Cenozoic radiolarian project, so off I went to the lab and spent a couple of very pleasant hours scanning the residues of one of the carbonate concretions from Ohio for radiolarians. It took a while to get myself set up, to get the microscope adjusted, and to get my eye in—but eventually I started to get a feel for what’s there. A lot of small quartz (?) grains, a lot of fluffy organic crap and a lot of other orangey and yellow mineral phases. But just before lunch I discovered a little whitish-clear ball with a couple of spines coming off it—could this be a radiolarian? It certainly looked like one. In my attempts to lift it out of its petri dish through the surface of the water layer it was sitting in—using a single eyelash attached to a wooden stick—I lost it again… Evidently, the eyelash is the tool of choice only for picking dry residue, not wet residue. My task over lunch: run to Bob Slate and obtain a fine paintbrush, which The Literature assures me is the preferable implement for transferring wet radiolarians from one place to another. The hunt is on!

The day continued to be quite an exciting one. Having procured the requisite paintbrushes (the #0000 proving to be the most nimble and accurate), I spent a good chunk of time poking around the petri dish under the picking scope, and found not one but several more little whitish-clear balls with spines coming off them that looked promisingly radiolarian-like. There were also several spicule-like fragments that looked like they could well be radiolarians of the Palaeoscenidiidae or some other spicular group, or broken off spines of some other large form. I deftly transferred these finds onto an SEM stub, hastily booked time on the SEM and raced over to the CNS to sputter coat the sample.

After some challenges finding the microscope (it had been moved) and physically accessing the room (my card was not activated for the microscope’s new home), I finally got to look at the beauties I had just discovered. Alas, they did not look quite as good (and in many cases, not quite as radiolarian-like) under the SEM as they had looked under the picking scope. Most of the spherical radiolarians appeared broken, and covered in muck—either mineral overgrowths from recrystallization, or just loose material from the macerate clinging to the radiolarian skeletons. Either way, the resulting images looked remarkably unspectacular when compared to the images in the published literature (click to see a larger image):

But still, they’re things of beauty. Broken specimens, diagenesis, and sample preparation issues aside, I must rate this day a success—I finally found radiolarians in the rocks I’ve been dissolving since this summer, and got to feel the thrill of discovery when I eventually got my eye in and saw the first little glassy balls under the picking scope. There is plenty of more sample to pick through, more samples to process, and many more samples to discover in the literature and procure from other workers. It’s an excellent start!

Friday

Wow, what a productive week! This morning I managed to rush out of the house a little sooner than the past few days, because I had time booked on the FIB all morning. I arrived just in time to get myself set up, logged in, and my sample loaded on the machine via the Kubrickian “load dock”, all the while humming Zarathustra… As I was running through the “FIB Daily Adjust” routine, which makes sure the two beams (SEM and FIB) are correctly aligned on your sample, I came up against a problem: the routine tilts the stage to varying degrees and asks you to repeatedly find and center on a particular feature of interest you choose on the sample. Part of this routine tilts the stage to 54˚, an angle at which the sample disappears “over the edge” when using a 45˚-angled stub, as was my plan (in order to be able to image at 90˚ from the milling plane, see post from last week).

I puzzled about this for a good long time (many, many dollars of instrument time!), but as I could not reach the FIB engineer, I resorted to taking the sample out and starting over on one of my flat-mounted samples. Just as I was getting started, the FIB engineer showed up (ah, timing!), and very helpfully showed me how to adjust the beam coincidence without using the daily routine macro, thereby avoiding the “over the edge” problem. Now I was ready to blaze, and zapped through my first diatom in the hour of time I now had remaining. The result:

Another beautiful cut (as viewed by the FIB beam). Two passes of the beam make the cut—a first, higher energy one (the big rectangle), followed by a second, lower-energy one (the dark line at the top edge of that rectangle). You can see here that my second pass was not set up well—it doesn’t actually seem to have touched the sample at all… In future, I will need to make sure I am imaging with the same probe as I am using to mill (I believe I was imaging with the 6 nA probe while milling with the 1.5 nA probe in this case), and use the location of the previous mill shape in defining the new shape, in order to avoid misalignment.

In any case, while this was a neat cut, you might also notice there is another (large) diatom lying directly south of my cut individual in the picture above. This presented another problem I hadn’t considered before, although it’s pretty darned obvious if you think about it: when viewing diatoms side-on, there can’t be anything else in the way. Duh. So when I tried to image the feller I had just zapped, the big guy next to him was in the way!

Drat. That was silly! But an important lesson to learn. The take-home message is to a) make sure the samples are thinly dispersed on the stub, so it’s unlikely for one guy to be blocking another’s view of the electron beam, and b) to check the side-on view of a specimen before beginning the milling process to make sure it’ll be visible in full. The best I was able to do with the sample I had just cut (which wasn’t quite sitting vertically on the stub anyway, as you can see in the image above) was to image it from an angle (pretty much the same situation I had before this ingenious 45˚-stub experiment), such that the guy in front wasn’t blocking the view.

Frustrating? Surely. But this is certainly progress—I am trialling the FIB-SEM method, as per my project plan, and that means encountering unexpected obstacles and difficulties. And finding a way to surmount those is a success for sure. It’s been a big week, and I’ve achieved a lot. Not just on the level of actual work accomplished, but on a meta-level of a behavioral transition of how I’m going about work. In the past year(s) I would (and did) let the absence of clear instructions on how to use a dye stop me. I would have stopped when I found an interesting sample and agonized for weeks about whether or not to book time on the instrument. Just trying it and seeing what happens, just logging in and booking the next available slot on the instrument feels much better. For the first time in a long time, I’m satisfied with my progress. And I even look forward to coming back to work on Monday.

A little bit.