FINDING HYDROGEN BY SYNCHROTRON POWDER DIFFRACTION

Makoto Sakata,^a Hitomi Sakai,^a Eiji Nishibori,^a and Masaki Takata^a,b

^aDepartment of Applied Physics, Nagoya University, Chikusa, Nagoya 464-8603, Japan; ^bJapan Synchrotron Research Institute, Kouto, Mikazuki-cho, Sayo-gun, Hyougo 679-5198, Japan (sakata@cc.nagoya-u.ac.jp)

It is not well established whether hydrogen could be detected by X-ray diffraction. Most of works to finding hydrogen by X-rays are done by least squares refinement, which relies on R-factor only introducing atomic scattering factor of hydrogen. It is reasonable to expect that chemical bonding nature of hydrogen in crystals would be considerably different from each other. Therefore, in a rigorous sense, it may not be justifiable to use atomic scattering factor of hydrogen in the least squares refinement. In the case of Cytidine (C₉H₁₃N₃O₅), for example, atomic positions determined by various authors very well agreered except hydrogen[1].

Recently, a sophisticated analytical method, which is called imaging of diffraction data by the Maximum Entropy Method (MEM)[2], has been successfully applied to find hydrogen in crystals[3]. MEM does not require atomic scattering factors of any atoms. The previously analysed crystalline materials are basically ionic crystals, which should not be extremely difficult to find. In this work, we determined charge density distributions of hydrogen in two organic crystals, which are Cytidine (C₉H₁₃N₃O₅) and Monosodium L-Glutamate Monohydrate (C₅H₈NO₄NaH_2O) by analysing synchrotron powder data, which are all collected by the large Debye-Scherrer camera installed at BL02B2, Spring-8. For Cytidine, both RT and low temperature (100K) data are analysed. It is understood that finding charge densities of hydrogen in MEM map is much easier when low temperature data is analysed. For the L-Glutamate, therefore, only low temperature (92K) data are analysed.

From the MEM charge density map obtained in this study, it is found that first of all charge density of hydrogen can be detected in MEM map without any ambiguities, secondly charge densities of hydrogen bond is similar to weak covalent bond, thirdly non-hydrogen bond H shows lobe shape charge density distribution, fourthly in Cytidine case, no local maxima is recognized for charge density of hydrogen even at low temperature, while in L-Glutamate case, it shows local maxima very clearly. In addition to these results, very interesting charge densities of hydrogen are found for hydrate water molecule, which does not show slightest sign of lobe shape charge density distribution. This strongly suggests that water molecules of hydrate are in disordered states occupying quite a few crystallographic sites.

References

1 Furberg, S. (1950) Acta Cryst. 3, 325; Furberg, S., Petersen, C. S. and Rømming, C. (1965) Acta Cryst. 18, 313; Ward, D. L. (1993) Acta Cryst. C49, 1789-92; Chen, L. and Craven, B. M. (1995) Acta Cryst. B51, 1081-97.

2 Sakata, M. and Sato, M. (1990) Acta Cryst. A46, 263-70.

3 For example, Noritake, T., Aoki, M., Towata, S., Seno, Y., Hirose, Y., Nishibori, E., Takata, M. and Sakata, M. (2002) App. Phys. Lett. 81, 2008-10.