Developments in methodology, instrumentation and computational power have ensured easy and fast structural analysis using X-Ray Powder diffraction (XRPD) data. While the technique has almost become a routine process, its results can vary in terms of accuracy.
Although a number of factors can impact the accuracy of structural analysis, this article addresses instrumentation only. By measuring a reference and an unknown sample on a Stoe Stadi MP diffractometer equipped with a MYTHEN detector, performance of the diffractometer is assessed and the level of accuracy that is possible in modern structure analysis is established.
High-Accuracy Data by Single-Photon-Counting
Factors such as high resolution, good counting statistics, simple background and symmetric peaks are used to characterize high quality XRPD data. However, in order to acquire such a data, it is important to optimize a number of instrumental parameters. With the scope on MYTHEN, single-photon-counting detector, integrated in Stoe Stadi diffractometers, this work addresses two questions: (i) what is single-photon counting and (ii) how does this technology enhance accuracy in structure analysis using X-ray powder diffraction data?
Figure 1. MYTHEN 1K detector system. Image credit: Dectris Ltd.
Table 1. Stoe stadi MP diffractometer
Tube |
Cu |
Detector |
MYTHEN 1K, 1000 µm |
Monochromator |
Ge 111 |
Geometry |
Debye-Scherrer |
Mode |
Scanning |
Radius [mm] |
190 |
Software |
WinXpow |
Conventional detectors are based on indirect detection where X-rays are first converted into visible light, which is then converted into charge. This charge is then accumulated, and the signal is obtained by integration over time. Low efficiency, poor resolution and noise inherent to this detection principle could be overcame only with new technologies.
MYTHEN is a silicon microstrip detector that operates in a single-photon-counting mode. In this process X-rays are detected directly, and each photon is individually processed and counted, resulting in the following benefits:
- Their quantum efficiency is significantly higher since they are optimized for X-rays, instead for visible light
- No dark current is produced because individual photon counting does not rely on charge accumulation
- No readout noise
Single-photon-counting works in the following way: X-ray photons interact with the detector sensor material and generate a charge. When an electrical field is applied across the sensor, the charge is forced to the readout chip, where each charge is processed separately. Its energy is evaluated against a threshold set by the user. In the event the energy is greater than the threshold, the photon is deemed acceptable for counting in a 24-bit counter. Such an energy discrimination makes it possible to suppress noise and fluorescence. Readout in 24-bit mode allows a dynamic range as high as 1:16.8*106, and happens very fast: MYTHEN silicon microstrip detector, readout times are only 300 μs.
The basic unit of the MYTHEN detector is a module. Each module includes 1280 parallely aligned silicon microstrips, each acting as individual detector, and counting up to 106 photons per second. Specific dimensions of the strips are responsible for other unique features of the detector. Strip width of 50 μm and its corresponding point-spread function of one strip cause signal to be “confined” the strip and shows no blurring. This results in minimal contribution of the detector to peak broadening, thereby allowing high-resolution measurements. Strip (sensor) thickness correlates with efficiency for certain energy range. Three available sensor thicknesses combined with strip length of 8 mm ensure optimal signal-to-noise ratio and efficiency over a wide range of X-ray energies (5-40 keV).
In a module, all 1280 strips detect X-rays in parallel fashion, what increases angular coverage and global count rates. A MYTHEN detector comprising a single 1280-strip module is known as MYTHEN 1K. These detectors are integrated in Stoe Stadi diffractometers. Table 2 sums up the detector’s specifications.
Table 2. Features of MYTHEN 1K
Sensor material |
Silicon |
N (strips) |
1280 |
Strip width [µm] |
50 |
Sensitive area [mm2] |
(1280x50 µm)x8 mm |
Max. count rate per strip [phts/s] |
> 106 |
Dynamic range [bit] |
24 (1:16.8 million) |
Energy range [keV] |
5-40 |
Point spread function |
1 strip |
Sensor thickness [µm] |
320, 450, 1000 |
Max. frame rate [Hz]
|
22
|
In order to understand how single-photon-counting technology enhances accuracy in structure analysis using X-ray powder diffraction data, two organic samples, one reference and one of unknown crystal structure were measured using Stoe Stadi MP diffractometer equipped with MYTHEN 1K detector. The data obtained was utilized to assess the instrument’s performance and to point out what level of structural detail can be seen using the XRPD data.
Please click here if you would like more information on the product in this article or a quote
Data Accuracy: Evaluation of the Performance of Stoe Stadi MP Diffractometer
In these studies, D-mannose (Figure 2) was selected as a reference sample for the following reasons:
- Crystal structure of D-mannose (P212121) was solved using single crystal diffraction data. The asymmetric unit comprises two molecules, differing in the orientation of their CH2-OH groups. However, the structure was not complete: some hydrogen atoms were missing, and some carbon atoms presented non-positive-definite values of atomic displacement parameters (ADP). Hence, it was interesting to observe what level of accuracy of the "backbone structure" can be achieved using XRPD data, and whether this data could explain non-positive definite ADP values.
- In contrast to sucrose, a commonly used reference sample, commercial D-mannose is believed to have a more uniform crystallite size, so sharp peaks stemming from a few individual large crystals in the sample could be avoided.
Figure 2. Structural formula of D-mannose. Image credit: Dectris Ltd.
First, a D-mannose sample was ground in a mortar and then filled into a capillary and measured on a Stoe Stadi MP diffractometer (Table 3).
Table 3. D-mannose: details of data collection
Scanning mode |
Step scan |
I [mA], U [kV] |
40, 40 |
λ [Å] |
1.5406 |
Detector, sensor thickness [μm] |
MYTHEN 1K, 1000 |
Resolution [°] |
0.015 |
Angular coverage [° 2θ] |
19.2 |
2θ range measured [°] |
0.56-101 |
dmin (used) [Å] |
1 |
2θ range [°] / t [s] |
0-21.935 / 360
18-37.935 / 720
35-54.935 / 1080
53-72.935 / 1440
72-101 / 1800 |
Using the XRS suite of programs, the crystal structure obtained from single crystal data was refined against the XRPD data. In order to evaluate the performance of the diffractometer and the quality of the data, Rietveld refinement was carried out without any geometrical restraints. ADPs were prescribed as isotropic values and were not refined.
Comparison of the atomic coordinates obtained from the XRPD data with those acquired from single crystal data, it was found that the average variation between the two sets of coordinates is in the range of ±0.06, as shown in Figure 3.
Figure 3. Results of the restraint-free Rietveld refinement presented as difference between the coordinates obtained from XRPD data and single crystal (SC) data. Image credit: Dectris Ltd.
The outliers seen in the plot most likely come from ambiguities in the difference Fourier map calculated for the last cycle of Rietveld refinement: residual electron densities around one of the molecules suggest disorder. An in-depth investigation of the difference Fourier maps and modeling of the D-mannose disorder is still ongoing.
Conducted experiments and calculations allow for conclusion that Stoe Stadi diffractometer equipped with MYTHEN detector produces high quality data:
- High quality XRPD data can support restraint-free Rietveld refinement. This way, model-bias is significantly reduced, what increases accuracy of the structure analysis.
- XRPD data can provide high accuracy structure analysis: XRPD results can match the results obtained from single-crystal data.
- XRPD data can provide the finest levels of detail such as disorder of light atoms.
Structure Determination: Tackling the Monomer-Trimer Ambiguity: MAAN
MAAN, 2-(methylideneamino)acetonitrile is a small molecule, reported to be monomeric specie, capable of polymerizing as cyclic trimer (Figure 4). The molecular and crystal structures of either the monomer or trimer have not been reported, and MAAN samples are characterized using a variety of spectroscopic techniques. Having in mind that MAAN can polimeryze and that distinguishing between the polymers using spectroscopic methods is sometimes difficult, one commercial MAAN sample, whose name suggested monomeric specie, was subjected to analysis.
Figure 4. Structural formulas of (a) monomeric and (b) trimeric MAAN. Image credit: Dectris Ltd.
First, infrared spectrum was measured and indicted that the material contains trimeric MAAN. However, it was not possible to discern the sample’s purity or the actual conformation of the polymer. Therefore, the molecular and crystal structure of the compound had to be established. Ground sample was filled into a 0.3 mm capillary, and measured under the conditions shown in Table 4.
Table 4. MAAN: details of data collection
I [mA], U [kV], λ [Å] |
40, 40, 1.5406 |
Detector, sensor thickness [μm] |
MYTHEN 1K, 1000 |
Resolution [°] |
0.015 |
Angular coverage [° 2θ] |
19.2 |
2θ range measured [°] |
0.56-91.93 |
dmin (used) [Å] |
1.07 |
2θ range [°] / t [s] |
0-21.935/1440
18-37.935/2880
35-54.935/4320
53-72.935/5760
72-91.935/7200 |
No. measured points
|
6129
|
Indexing of the XRPD pattern resulted in monoclinic cell, what suggested that the material contains only one phase. This information, and molecular structure constructed using Density Functional Theory (DFT) calculations were then used for structure determination in direct-space program FOX. The resulting structure was later refined in the XRS suite of programs. During the refinement process, geometric restraints were applied, but in order to allow flexibility of the molecule, their weighting factor was reduced methodically. The hydrogen atom positions were calculated in the program Mercury. ADP values of all the atoms were prescribed to expected values, and were not subsequently refined. In the last phase of refinement, a difference Fourier map was calculated. No suspicious electron densities maxima were observed. Results of the refinement are presented in Table 5 and Figure 5.
Table 5. Results of the Rietveld refinement for trimeric MAAN
a (Å) |
15.1876(9) |
b (Å) |
10.18249(8) |
c (Å) |
6.9466(4) |
β (°) |
90.9(3) |
Space group |
P21/n |
No. reflections |
625 |
No. parameters |
45 |
No. observations |
5211 |
No. soft restraints |
33 |
RF |
0.104 |
Rwp |
0.176 |
Rexp
|
0.059
|
Figure 5. Last cycle of the Rietveld refinement: measured pattern (black), calculated (red) and difference curve (blue). Image credit: Dectris Ltd.
Given that the crystal structure (Figure 6a) did not have any strong hydrogen bonds, it was assumed that the molecular structure must be close to the one calculated by the DFT method. Figure 6b shows a comparison of the molecular structure obtained from the refinement against XRPD data and the one calculated by DFT. Such a good fit demonstrates that it is possible to obtain accurate solutions even without prior knowledge about the material.
Figure 6. (a) Crystal structure viewed along c axis (b) comparison of molecular structure (blue) to the molecular structure obtained from DFT calculations (purple). Image credit: Dectris Ltd.
Conclusion
In this study, the two examples show that the combination of a Stoe diffractometer and the MYTHEN single-photon-counting detector can enhance structure analysis using the XRPD data.
The tests carried out on D-mannose demonstrate that the high quality XRPD data can allow for restraint-free refinement, resulting in structure that is comparable to single crystal case. In addition, this high quality data makes it possible to observe the finest structural details such as localized residual electron densities, which denoted anisotropic displacement of atoms, indicating possible disorder in the molecule. By reducing a model-bias, and detection of fine structural details, the combination of Stoe Stadi diffractometer and single-photon counting detector push the limits of accuracy that is obtainable from XRPD data.
This information has been sourced, reviewed and adapted from materials provided by Dectris Ltd.
For more information on this source, please visit Dectris Ltd.