Experimental results on ELDICO ED-1

Sample:
Sodium calcium aluminium fluoride, Na₂Ca₃Al₂F₁₄

Goal:
ED-1 functional proof of concept

Approach:
A series of 600 diffraction patterns was collected spanning a rotation angular range of 120°, with an exposure time of 0.1 second per frame.

Challenge:
Keep the sample in the beam as accurate as possible through a sphere of confusion << 1 µm to obtain accurate unit-cell parameters and orientation matrix

Solution:
The data processing via state-of-the-art crystallographic software allowed to reconstruct the 3D diffraction pattern. The pattern could be indexed with a cubic unit cell and space group I213. The observed lattice parameters a=b=c deviate by only 0.36% with respect to the value of 10.257(1) Å reported in literature for the same material characterised with X-ray diffraction, indicating good accuracy on the geometrical parameters.

Sample:
Paracetamol, C₈H₉NO₂

Goal:
Confirm ability to measure organic compounds

Approach:
A single crystal was measured in continuous rotation mode for a total measurement time of 100 sec over 100° rotation – all at room temperature

Challenge:
Confirm that the ED-1 can measure fast enough to even keep sensitive organic samples alive, yet enabling ab-initio structure solution.

Solution:
An ED data set was successfully collected at room temperature on a nanocrystal of approximately 660 × 350 nm2. The data set obtained was integrated to a resolution of 0.8 Å. The structure was solved with direct methods in the monoclinic space group P21/n with a = 7.0603(9) Å, b = 9.496(2) Å, c = 11.648(2) Å; β = 96.86(1) °.

Sample:
Metaxalone, C₁₂H1₅NO3

Goal:
To benchmark the performance of the ED-1

About:
Sold as Skelaxin, it is a muscle relaxant, but its exact mechanism of action is still not known. Metaxalone exhibits various polymorphic forms with substantial effects on e.g. solubility and bioavailability: A-rac (triclinic, P-1), B-rac (monoclinic, P21/c and C-rac. Finally, A-R/S, is a conglomerate of the enantiomorphic forms A-S and A-R (orthorhombic, P212121). The latter however could be obtained as nanometer-sized needles only and thus required electron diffraction for accurate structure elucidation.

Approach:
A single crystal was measured in continuous rotation mode for a total measurement time of 62.5s over 125° rotation – all at room temperature

Challenge:
Improve on data quality over literature for nanometer-sized needles

Solution:
Many features of the data indicate a higher performance of the ED-1 diffractometer compared to the equipment previously used. This is clearly seen, not only, in the precision of the unit cell obtained (5.512(1), 10.564(3) and 19.906(7)), but also in many other qualifiers such as the lower Rint value of 24% and the much lower refinement R factors.

Sample:
Febuxostat, C₁₆H₁₆N₂0₃S

Goal:

Avoid tedious crystallisation – save time & money

Approach:

Determine the crystal structure of a third polymorph, which crystallises preferably in very thin needles

Challenge:

Using the tools of the APEX4© software package two twin domains were identified and integrated separately. With the merged dataset of both domains the structure could be refined to an R1 of 14.2% (wR2 of 37.9%) and a completeness of 86.7% (vs. 80.4% for only the main domain) could be achieved from one single run. This demonstrates that electron diffraction can not only handle twins, but can even draw a benefit from it.

Solution:

A single run from a single crystal was processed by PETS 2 up to a resolution of 0.91Å; indexing was successful, using all reflections, resulting in a final constrained orthorhombic unit cell of a=9.0310(6), b=13.7770(17) and c=26.4120(9) Å; solving the structure by direct methods was streightforward

Sample:
Tyrosine, C₉H₁₁NO₃

Goal:
Getting a crystal structure from a pseudo-twinned sample of tyrosine, i.e. two crystals measured at the same time.

Approach:
One single run spanning an angular range of 100° phi rotation was recorded in 1° steps with an exposure time of 0.5 s per frame.

Challenge:
Solving the structure of a twinned crystal from electron diffraction data having only 100 frames at hand.

Solution:
Using the tools of the APEX4© software package two twin domains were identified and integrated separately. With the merged dataset of both domains the structure could be refined to an R1 of 14.2% (wR2 of 37.9%) and a completeness of 86.7% (vs. 80.4% for only the main domain) could be achieved from one single run. This demonstrates that electron diffraction can not only handle twins, but can even draw a benefit from it.

Sample:
Histidine, C6H9N₃O2

Goal:

Determine the absolute configuration of a chiral compound

Approach:

Diffraction data was collected in steps of 0.5° per frame while the crystal was continuously rotated at 1° per second over a range of 120°.

Challenge:

Obtain data accurate enough to clearly observe intensity differences among Friedel pairs

Solution:

Data out to 0.7Å are 79% complete in P2₁2₁2₁ and superior in many aspects to already published data. Even at highest resolution <I/σ> is at least 3, indicating the absolute configuration can be determined