Hard radiation for crystallography

Using single-crystal data collection strategy for powder diffraction 

Introduction

Single crystal diffraction typically uses hard radiation such as molybdenum in order to determine the precise crystalline structure of materials. Besides the ability to distinguish between inequivalent (hkl) reflections, the precise crystal structure determination is also due to the fact that one can reach short d spacing values with such radiation. Furthermore, a typical data collection strategy is to measure the weak reflections longer in order to have an improved determination of their intensities. And what about powder diffraction?

With the emergence of new technologies for the detection of hard X-radiation, it becomes possible to carry out hard radiation experiments without loss of efficiency in detection. This is made possible nowadays using CdTe technology (see Figure 1) such as implemented in the latest Malvern Panalytical detector GaliPIX^3D. The irradiated volume in powder diffraction is a key issue since higher volume yields enhanced particle statistics. Therefore, the use of high-energy radiation is beneficial as the irradiated volume of the sample can be increased. Additionally, this allows for transmission experiments preventing any preferred orientation and suppresses the typical fluorescence typically from transition metal ions. Here we present a detailed investigation of the crystal structure of Fe(IO₃)₃ from powder data. We compare our results with previous ones published in the literature based on single crystal and powder data measured with Cu radiation [1,2]. Using Mo radiation with the Debye Scherrer geometry, we show how accurate and meaningful results can be obtained thanks to the combination of Mo radiation, GaliPIX^3D detector and an optimized data collection strategy

Experimental

The X-ray diffraction measurements shown here were conducted on the Empyrean, Malvern Panalytical's multipurpose diffractometer, equipped with a focusing mirror, a high- resolution molybdenum tube, 0.02° Soller slits and a GaliPIX^3D detector. The powder sample of Fe(IO₃)₃ was loaded in borosilicate capillary of 0.7 mm diameter. In order to optimize the data collection, a variable counting time strategy has been applied similar to the case of single crystal diffraction.

Figure 1. Absorption efficiency for various radiations comparing the classical Si technology with the newly developed CdTe technology available with the GaliPIX^3D detector

Results/explanation

Fe(IO₃)₃ has already been reported in the literature [1,2]. Looking at these reports, several points are not very satisfactory:

1. Isotropic atomic displacement parameter being zero within the error bars or arbitrarily fixed
2. Low precision of the atomic positions of the lightest atoms (σ = 0.002-0.003 for oxygen atoms)

Using the structural model previously reported, we refine the crystal structure of Fe(IO₃)₃ using the data collected with Empyrean. Figure 2 shows the resulting powder diffraction pattern and its Rietveld refinement. First of all, one can notice the almost flat difference curve (blue curve in Figure 2) indicating the excellent fit obtained using the HighScore suite [3]. This excellent fit is confirmed by the Rwp value of 2.4. Table 1 reports the atomic coordinates derived from the Rietveld refinement. As expected, we obtained an excellent precision of the position of the iodine atom but also of the lighter atoms such as the oxygen atoms. In particular, the error bars on these lighter atoms are 10 times better than previously reported (σ = 0.0002-0.0004 compared to 0.002- 0003).

Table 1. Atomic coordinates obtained from the Rietveld refinement for Fe(IO₃)3. Fe(z) has been fixed to zero in order to fix the origin in this polar space group.

Furthermore, all isotropic atomic displacement parameters are positively defined. Most importantly, we can notice that the O₃ oxygen atom vibrates significantly more than the other oxygen atoms. Looking at the crystal structure (see Figure 3), one can notice that the O₃ atoms are facing an empty channel and thus are expected to vibrate more than the other oxygen atoms. Such meaningful Biso determination is made possible thanks to the use of molybdenum radiation and a variable counting time data collection strategy. 

Figure 2. Rietveld refinement of Fe(IO₃)₃ carried out with the HighScore suite [3]. Space group P63 with cell parameters a = b =9.2361(1) Å, c = 5.23882(7) Å; Rwp = 2.4, GoF =3.4 .

Figure 3. Crystal structure representation of Fe(IO₃)₃. Green atoms are iron atoms, purple atoms are iodine atoms and red atoms are oxygen atoms. The O₃ atom is facing the empty channel. 

Conclusion

The use of molybdenum radiation for collecting data with variable counting time coupled to the use of the GaliPIX^3D technology allows to obtain a precise and accurate structure determination from powder data. This high accuracy of the structure is further supported by meaningful Biso values derived from the Rietveld refinement of light atoms.