Previous Article in event
Previous Article in session
Next Article in event
High throughput powder diffraction
Published:
11 November 2020
by MDPI
in The 2nd International Online Conference on Crystals
session Software/Tools to Deal with Crystal and Crystallographic Issues & Teaching Crystallography
Abstract:
X-ray powder diffraction is a powerful technique to probe crystalline materials, their purity and their structural characteristics. When coupled with synchrotron sources, it is also able to detect the subtlest details in situ, operando or in time-resolved studies.
The high flux of these sources enables also very high-throughput data collections, where over 10.000 samples can be analysed in a single day. This requires a combination of quick sample preparation, delivery, data collection and, potentially, processing. Techniques and methodologies will be presented, alongside relevant issues and opportunities such high-throughput analytics involves.
Keywords: high-throughput;combinatorial;diffraction;materials
Comments on this paper
Oriol Vallcorba
13 November 2020
Few questions about these measurements
Hi Nicola, nice work and very interesting. Thank you for the presentation.
I have a few of questions regarding these measurements:
1) What size of the beam you use with the 2D detector? It is another parameter that affects particle statistics...
2) Did you actually used the protein crystallization plates in this study? how big is the entrance of the "holes" where each sample is mounted and the beam is passing through? aren't they too thick (I mean the size in the beam direction)? I did not understand clearly if they were only a possibility for static measurements or they can actually be used in this application. Otherwise a custom solution will always be required for powder diffraction studies since, in most of the cases, we need to increase particle statistics somehow (moving, rotating, vibrating, ...).
3) You propose a solution to measure this huge amount of samples, which is working nicely and produce good data. However I am concerned about the capability of companies or research groups to produce and prepare such amount of samples and deal with the data analysis. Regarding the latter, do you also provide a solution (specially to companies, proprietary research) for the data analysis? What type of analysis or what did the company do with the 50000 sample data (if you are allowed to tell, of course)?
Thank you very much,
Best regards,
Oriol.
I have a few of questions regarding these measurements:
1) What size of the beam you use with the 2D detector? It is another parameter that affects particle statistics...
2) Did you actually used the protein crystallization plates in this study? how big is the entrance of the "holes" where each sample is mounted and the beam is passing through? aren't they too thick (I mean the size in the beam direction)? I did not understand clearly if they were only a possibility for static measurements or they can actually be used in this application. Otherwise a custom solution will always be required for powder diffraction studies since, in most of the cases, we need to increase particle statistics somehow (moving, rotating, vibrating, ...).
3) You propose a solution to measure this huge amount of samples, which is working nicely and produce good data. However I am concerned about the capability of companies or research groups to produce and prepare such amount of samples and deal with the data analysis. Regarding the latter, do you also provide a solution (specially to companies, proprietary research) for the data analysis? What type of analysis or what did the company do with the 50000 sample data (if you are allowed to tell, of course)?
Thank you very much,
Best regards,
Oriol.
Nicola Casati
13 November 2020
Dear Oriol,
the beam size plays a role only in the peak shape, for geometric reasons, nevertheless this is a circle of 0.2 mm diameter or a square of 0.3 x 0.3 mm, typically, depends a bit on the project. The reasons it does not play a significant role is that the amount and speed of movement of the powder means you are statistically very homogeneous within the first few seconds, except for minority phases. It is this parameter that overwhelms the size of the actively probed part of the sample container.
We did not use crystallization plates, they are rather thick and this means broad peaks, on the other hand they are mentioned as they are the way to automatize screening in several beamlines and also home based machines. IN our case we have 'washer' type containers or plates. The entrance and exit 'holes of the sample container are 5-12 mm, depends a bit on the project again, but they are basically much larger than the beam + vibration amplitude. The thickness is 1.2 mm, one can work also with larger or smaller, depends a bit on the beam energy used, we typically work between 0.7 and 0.5 Å. Yes, I will never suggest to do static measurements unless you have nano powders.
There are companies and groups that are going into this direction, Stenman minerals was the partner in the 50.000 samples study, for example, regarding quantification from samples collected in a mine. Research tools are now commercialized for high throughput, I think there's a company called Chemspeed that makes these types of machines for home laboratories. When it comes to analytics, this can be batched in several programs, with fast computing it may last quite as long as the collection time, if you can define well the borders of your project. On the other hand clustering methods and other big data tools are now employed in several sectors, including diffraction, it always depends on the project of course.
Was this a sufficient answer?
Nicola
the beam size plays a role only in the peak shape, for geometric reasons, nevertheless this is a circle of 0.2 mm diameter or a square of 0.3 x 0.3 mm, typically, depends a bit on the project. The reasons it does not play a significant role is that the amount and speed of movement of the powder means you are statistically very homogeneous within the first few seconds, except for minority phases. It is this parameter that overwhelms the size of the actively probed part of the sample container.
We did not use crystallization plates, they are rather thick and this means broad peaks, on the other hand they are mentioned as they are the way to automatize screening in several beamlines and also home based machines. IN our case we have 'washer' type containers or plates. The entrance and exit 'holes of the sample container are 5-12 mm, depends a bit on the project again, but they are basically much larger than the beam + vibration amplitude. The thickness is 1.2 mm, one can work also with larger or smaller, depends a bit on the beam energy used, we typically work between 0.7 and 0.5 Å. Yes, I will never suggest to do static measurements unless you have nano powders.
There are companies and groups that are going into this direction, Stenman minerals was the partner in the 50.000 samples study, for example, regarding quantification from samples collected in a mine. Research tools are now commercialized for high throughput, I think there's a company called Chemspeed that makes these types of machines for home laboratories. When it comes to analytics, this can be batched in several programs, with fast computing it may last quite as long as the collection time, if you can define well the borders of your project. On the other hand clustering methods and other big data tools are now employed in several sectors, including diffraction, it always depends on the project of course.
Was this a sufficient answer?
Nicola
Oriol Vallcorba
13 November 2020
Thanks
Dear Nicola
Yes, many thanks!
Best regards,
Oriol.
(sorry there was no option to reply on your reply)
Yes, many thanks!
Best regards,
Oriol.
(sorry there was no option to reply on your reply)
