Looking At Nothing

I think it is safe to say that “everyone” is familiar with electron microscopy for looking at small things, but the same cannot be said for small-angle scattering. Introducing more people to the wonderful world of small-angle scattering may be one of the ways of getting more support for and interest in the technique. With this in mind, and with some contacts in the PR department, we managed to get last-minute approval to do a demonstration at the public day at NIMS.

The team of the 2013 NIMS public day with Martin, me, Novi and Ryo. Missing from the picture is Julian.

The plan was to have people measure their own hair diameter using a laser in a similar way as demonstrated before in these youtube videos. As I was not expecting more than a few visitors per hour, 5 minutes per visitor was considered a suitable amount of time. To help them, we prepared a step-by-step instruction sheet and some “tools” that the people could cut out and use for their experiment. The procedure was as follows:

1. Visitors would be lured in by our blatant advertising to “measure the diameter of your hair using a laser!” (stuck in big letters on the wall, taking up lots of floor space and having a big screen with the program running on it. Also, we had foreigners at our desk in an otherwise very much Japanese event. There was even some Dutch candy!). Visitors would then be given an instruction sheet and some tools.
2. They would cut out the tools (one target and one holder for their hair), find a hair and stick it on the paper. Surprisingly, almost everyone could spare one (or had one cut off by one of the staff), and they very much enjoyed the bit of manual labour of cutting out pieces of paper, sticking hair on it and so on.
3. Going over to the laser, where the staff mounted the hair in front of the laser (which was a bit finicky work to get it to sit right) using a couple of paper clamps. The laser pointer was forced “on” using another paper clamp holding down the button (the laser pointer itself managed about half a day on a set of batteries, surprisingly).
4. Clamping the target onto a clipboard stuck to a tripod, and aligning the central black spot on target with the laser pointer beam, the scattering from a hair becomes clearly visible. The clipboard was about 2m from the hair and laserpointer. At this point I would always take a moment to explain that the central spot was from the laserpointer, but the oscillations were coming from their hair (in my best Japanese).
5. We would then ask them to draw arrows in the center of the oscillation spots. This is where it would most often go wrong, as people were unsure on where to draw the points and my Japanese would not be sufficient to explain this. Nevertheless, they had fun drawing points.
6. Taking that piece of paper, I would take them to measure the required values: the distance from hair to the target (in meter, with a tape measure if there was time), the number of spots from the target center they could see, the distance from the target center to that last spot (in mm), and reading the wavelength from the laser (green=532 nm, now forever etched in my memory).
7. Filling in these values in a Python program (available here) they were then given the diameter of their hair. In the beginning this was set to display the diameter in micrometer, but after an hour or two it became evident that that was not very well known with the public. We therefore changed this to mm. We had our Japanese staff stand at the computer, so he could explain a little about common hair diameters (European: 40 micron, Japanese: 80 micron with the occasional outlier at 110 or 120 micron). Significant digits and uncertainty were a bit of an issue, some wrote down all nine digits resulting from the calculation.

Before I tell a bit more about the experiences on the day, here are the tools you can use (conveniently in English and Japanese):

• The explanation sheet in English and Japanese
• The “tools” sheet in English and Japanese
• The Python program with text in English and Japanese

As this was the first time we did this, we had no idea what to expect. We shared the big lecture hall with three other demonstrations, which helped to draw people to us. We set up the demonstration in a slightly inefficient way, whereas we should have considered the flow of people through the steps. This was exacerbated by the sheer number of visitors. Visitors tended to come in groups of four or more (with the occasional elementary school class to make things really hectic), after which it would be quiet again. Overall, we “served” an estimated 70 people during the 6 hours, which were about 50% schoolchildren and about 50% adults. We had two to four people staffing the set-up at any time.

Ideas we have gained for next time include:

• Setting up the tables a bit better, so that there can be a more logical flow from stage to stage (whilst still keeping it compact), and reserving a separate table for the cutting and sticking-work.
• The set-up using commonly available bits and pieces worked in our favour. Some said they intended to repeat the experiment at home, showing that they noticed it consisted of nothing more than a laser pointer and some paper clamps.
• If we prepare for groups we can explain only once for a group of four or five.
• Staffed stages could be three stages (cutting, measuring and computation) or four stages (cutting, “laser”-ing, measuring, computation). Parallel laser stations would be preferred, perhaps of different wavelength.
• A microscope table can be included for people who want to verify (which should be encouraged if there is time).
• It would be nice if a form was printed with a picture of the measurement with maybe a picture of them, their hair oscillations and the “scientist”-guaranteed number (signatures and such).

I strongly encourage people to try this demonstration at their institute as it was good fun! We’ll do this next year again if we have the chance so do drop by if you have the chance!

A day later than planned, but here’s another post…

I am planning to speak at the TMS Magnesium Workshop Madrid 2013, some time during the conference lasting from May 21 to May 25. During this conference I will be focusing on the results we obtained for ex-situ and the preliminary in-situ results studying the growth of precipitates in MgZn alloys. Naturally, I will be heavily plugging SAXS and the MC analysis methods. If you are in Madrid or at the conference at that time and want to meet, please leave a message…

Secondly, For a recent conference I had the chance to make another poster. It has been a while since I posted any posters on this site, so if you want to get an idea what they look like these days, take a look here. The design is still a bit full of text but in a font large enough to read. Additionally, the text is laid out in a newspaper column-format to make it easier on the eye to move from the end of one line to the beginning of the next line (some info here). The whitespace at the right-hand side should make it easier to distinguish the section headers, and given that this poster was presented in Japan, where comics are read from right to left, this poster has section headers rightmost, followed by an image left of the header, and then two columns of text (which are read left-to-right, unfortunately breaking the mould). The introduction and conclusion section is typeset to a larger font as this is what most people are interested in, and the authors and affiliations is at the bottom (because that is what people should remember at the end). The space at the bottom should contain my business cards in a pocket so people can grab one if I am not around. Lastly, the background is meant to appeal at large distance for its colour, but set to a transparency which does not hinder reading at close distances.

Despite these considerations, the poster has not enjoyed much success, so for the next one I may employ a different style.

It has been a long time in the making, but now the day has finally come where the 1D Monte Carlo method has been published! To top it off, the publication is open access (courtesy of my current institute: NIMS), and has a wicked showcase document as supplementary material. Feel (very) free to check it out here!

In short: the MC method can retrieve form-free particle size distributions from isotropic scattering patterns, complete with uncertainties. There is more to it than that, but the details are in the paper. The Python 2.7 code is available in a Git repository as indicated in this post, available under a creative commons attribution sharealike license.

In the same J. Appl. Cryst. edition that my paper appears in, there is another paper which caught my interest. This is a paper by E. B. Knudsen, et al., which presents a bunch of code under the name of FabIO that can be used in Python to read in detector images (under a GPL license). There is even some code there for the more obscure of image formats, with improvements to follow. For me, this is part of what I was looking for to augment my own procedures, so I really appreciate it when others make their stuff available with a suitable license!

That FabIO package is part of larger programs for XRD and data reduction, respectively. Given that this is a program which is sustained by the efforts of a variety of user-driven institutes (ESRF, for example), I am very much looking forward to see which direction it will be taken into. One of the authors has assured me that if there is a need and an example (say, for supporting an imageplate image format or some of the stranger wire-arrays), they are more than willing to implement that particular image format. Naturally, that open-source project is available for collaborative efforts as well!

Hello everyone, and welcome to the remaining 90% of 2013 (as of February 4), surely you’ve made good use of the first 10!

I’ve been tasked with writing a review paper which made me come across several noteworthy papers. Two topics in particular surprised me: the history of instrumentation and the depth of detector corrections. In the hope that they may be of some interest to some of you, a few words on them might not be amiss.

What may come as a bit of a surprise, as it did to me, is that the history of SAXS instrumentation turns out to have been quite advanced right from the start. Current papers on instrumentation discuss the benefits of the technological advances from focusing optics, three-pinhole systems, crystal monochromators and so on, so here is a quote from O. E. A. Bolduan in 1949: “Investigators have attempted to solve the peculiar requirements of this field by use of longer wave- lengths to increase diffraction angles [3] crystals for monochromatization and collimation [4-7] high intensity x-ray tubes [8,9] or cameras with focusing by means of curves [sic] crystals [9,1o] or totally reflecting plates [11] to increase the level of useful radiation [...]“, with Bolduan himself discussing three-slit collimation to reduce parasitic scattering from slit #2. Publications of SAXS instruments on home-built rotating anode generators appear already in the 1950′s. After reading a lot of papers discussing the finer details of those instruments leaves me wondering what major improvements we have made in the 60 years since that paper, and one of them is of course in the field of detectors.

Before the advent of more advanced 2D wire-detectors, imageplates and CCD-based detectors, one was pretty much stuck with using either Geiger counters (0D) or photographic film for 2D images. The photographic film, though, behaves rather nonlinearly vis-a-vis incident radiation amounts, necessitating serious corrections before it could be used. 2D wire detectors were an improvement, but only suitable for quite low countrates (<100 Hz/pixel) and show a rather large point-spread function. CCD-based detectors brought improvements there but their dynamic range is typically not much higher than 3 to 4 orders of magnitude. Imageplates brought improvements with up to 6 orders of magnitude of dynamic range, but necessitate laborious read-out and erasure procedures and thus are ill-suited for time-resolved experimentation. Recently, direct-detection detectors appear to solve the majority of the issues of all prior detectors, at a hefty price (conveniently, the price-tag of the Pilatus detector comes permanently written on its housing: a Pilatus 100k costs about 100k-euro, a 300k costs 300k-euro and so forth).

Nevertheless, each of these detectors produces intensity that is in dire need of corrections. Reading up on these, here are my recommended papers for getting more insight into some of these. Firstly, there is S. L. Barna with a good list of corrections, with some focus on distortions found in CCD’s. Furthermore, in results shown by V. Le Flanchec that even image plate data needs to be corrected for read-out positional inaccuracies as well as some other effects. F. Né also indicates their darkcurrent corrections need special treatment, in particular to start counting from the time of their last erasure.  Other names to look for insights on this topic are, amongst others, P. Boesecke, T. Zemb, A. Rennie, P. Jemian, and many more, especially those in the CANSAS groups. Through their work, my list of corrections (of a month ago) has by now increased to a whopping 20(!) data corrections that could be considered applicable for small-angle scattering data. This space is too small to describe them all, but there will be space elsewhere later on this year, so stay tuned!

Lastly, in unrelated news: we applied the previously presented MC method to obtain radius distributions for isotropically oriented, rod-like precipitates in metals, the draft has been made publication-ready and can be found here. We are looking for a journal to publish this in at the moment.

Keep scattering!

Barna: Barna, S. L. and Tate, M. W. and Gruner, S. M. and Eikenberry, E. F., “Calibration procedures for charge-coupled device x-ray detectors”, Rev. Sci. Instrum. 70 (1999), pp. 2927

Le Flanchec: V Le Flanchec, D Gazeau, J Taboury, and T Zemb, ”Two-Dimensional Desmearing of Centrosymmetric Small-Angle X-ray Scattering Diffraction Patterns”, J. Appl. Cryst., 29 (1996), pp. 110–117

Bolduan-1949: Orvil E. A. Bolduan and Richard S. Bear, ”Effective Use of Collimating Apertures in SmallAngle XRay Diffraction Cameras”, J. Appl. Phys. 20, 983 (1949); doi: 10.1063/1.1698263