The Crystallographic Revolution
Total Page：16
File Type：pdf, Size：1020Kb
8 CHIMIA 2014, 68, Nr. 1/2 Crystallography, past, present, Future doi:10.2533/chimia.2014.8 Chimia 68 (2014) 8–13 © Schweizerische Chemische Gesellschaft ‘Seeing’ Atoms: The Crystallographic Revolution Dieter Schwarzenbach* Abstract: Laue’s experiment in 1912 of the diffraction of X-rays by crystals led to one of the most influential discoveries in the history of science: the first determinations of crystal structures, NaCl and diamond in particular, by W. L. Bragg in 1913. For the first time, the visualisation of the structure of matter at the atomic level became possible. X-ray diffraction provided a sort of microscope with atomic resolution, atoms became observable physical objects and their relative positions in space could be seen. All branches of science concerned with matter, solid-state physics, chemistry, materials science, mineralogy and biology, could now be firmly anchored on the spatial arrangement of atoms. During the ensuing 100 years, structure determination by diffraction methods has matured into an indispensable method of chemical analysis. We trace the history of the development of ‘small-structure’ crystallography (excepting macromolecular structures) in Switzerland. Among the pioneers figure Peter Debye and Paul Scherrer with powder diffraction, and Paul Niggli and his Zurich School with space group symmetry and geometrical crystallography. Diffraction methods were applied early on by chemists at the Universities of Bern and Geneva. By the 1970s, X-ray crystallography was firmly established at most Swiss Universities, directed by full professors. Today, chemical analysis by structure determination is the task of service laboratories. However, the demand of diffraction methods to solve problems in all disciplines of science is still increasing and powerful radiation sources and detectors are being developed in Switzerland and worldwide. Keywords: Bragg · Laue · Service crystallography · Structure determination · X-ray diffraction 1. Introduction chemists, the indestructible elements of circumstantial.This conundrum was solved Mendeleev’s Periodic Table hinted at the by X-ray crystallography whose develop- Atomic theories postulating that mat- existence of atoms that were linked to form ment is one of the crucial scientific revo- ter is composed of discrete components molecules. The theory of atomic valence lutions because it provided a ‘microscope’ date back to antiquity.[1] However, before and schemes of bonding, particularly in with atomic resolution that made atoms the early 20th century atoms could not organic chemistry, was proposed in the and their arrangements in space visible. It be seen, they were invisible even for the 1850s by Kekulé, but atoms were invisible is perhaps important to explain here what highest-resolution instruments available. and ideas about their nature were vague. exactly is meant by ‘microscope’, ‘seeing’ During the 19th century, physical observa- (iii) The physicists’ theories of atoms were and ‘visible’. Crystallographic results in tions accumulated indicating the existence closest to our modern understanding. The the form of pictures consisting of spheres of atoms, or of a grainy submicroscopic physics of gases led to Avogadro’s propo- or ellipsoids connected by rods which rep- structure of matter: (i) Crystallographers, sition that equal volumes of different gases resent molecules or structural frameworks finding that the orientations of the faces at equal temperature and pressure contain are today ubiquitous. These pictures are not of polyhedral crystals could be represent- the same number of particles. Loschmidt quite the same as direct observations with a ed by proportions of small integer num- in 1865 determined the ‘size of air mol- high-resolution microscope; they only rep- bers (law of rational indices, Miller indi- ecules’, i.e. Avogadro’s number. In 1897, resent the average locations of atomic cen- ces), postulated the periodicity of crystal J. J. Thomson showed that cathode rays tres and some information on atomic ther- structures. But the lattice theory had little are composed of particles which he called mal vibrations and possibly other types of impact on physicists and chemists, the na- electrons, and then proposed the plum pud- disorder. The rods represent interpretations ture of the periodically repeated object or ding model of the atom. Atoms as counta- of short interatomic distances as chemical ‘atom’ being mysterious. Lattice theory ble items were first ‘seen’ by Rutherford bonds. Thus, bonds are inferred, not ob- turned out to be incapable of predicting when he identified α-rays to be composed served. Modern charge-density analysis physical properties of crystals. (ii) For of He ions. Rutherford’s atom model con- with accurate diffraction data may furnish sisting of a tiny nucleus surrounded by direct evidence on interatomic bonding, orbiting electrons was proposed in 1911. but this is certainly not an easy, efficient In 1913, Bohr published the first version experiment. Direct visualisation of atoms of his quantized planetary-like atom mod- had to wait for high-resolution electron mi- el. All these models developed by physi- croscopy and scanning probe microscopy. cists had little impact on the chemists at The discovery of X-ray diffraction by the time of their publications because they crystals is due to Friedrich, Knipping and gave no insight into chemical reactions, Laue in 1912,[3] the idea having occurred *Correspondence: Prof. D. Schwarzenbach even though Bohr hoped his model to be to Laue in a discussion with Ewald on the Ecole Polytechnique Fédérale de Lausanne of chemical relevance.[2] latter’s doctoral thesis. This is one of the IPSB - Cristallographie Thus, ideas about atoms remained quite most influential experiments whose reper- Le Cubotron (BSP) CH-1015 Lausanne disparate and depended on the scientific cussions cannot be overestimated. It was E-mail: [email protected] discipline, evidence on atoms was largely acclaimed immediately at the time of pub- Crystallography, past, present, Future CHIMIA 2014, 68, Nr. 1/2 9 lication as the first experimental proof of NaCl, diamond and ZnS are essentially 3. Crystallography to Become a the lattice periodicity of crystal structures, determined by symmetry, and they could Method of Chemical Analysis and it also established the wave nature be guessed directly from the diffraction of X-rays. Its history is well researched pictures without knowledge of diffraction The theoretical basis having been estab- and well documented.[4] But none of the physics. The theory of X-ray intensity and lished, the field grew with the development original authors were interested in crystal of what is today called data reduction to of laboratory equipment. The production structures and chemistry. The revolution derive structure amplitudes was formulat- of X-rays was simplified by the invention of X-ray crystallography is due to William ed in the immediately following years. In and perfection of easy-to-use sealed X-ray Lawrence Bragg who first realized, at age 1913 the sphere of reflection and the notion tubes that became commercially available 23, that Laue’s X-ray diffraction photo- of the structure factor by Ewald, in 1914 in the 1920s. Many types of single-crys- graphs contained information on the spatial the temperature factor by Debye (up to a tal X-ray cameras, such as the success- arrangement of the atoms in the crystal.[5] factor of 2, corrected for later by Waller), ful Weissenberg and precession cameras, the need for integrated intensities and the were invented in the 1920s and 1930s. In Lorentz factor, and two-beam dynamical the 1950s, 3- and 4-circle single-crystal 2. The First Crystal Structures diffraction by Darwin. In 1915, X-ray scat- diffractometers appeared, first for neutron tering by a gas and the atomic scattering crystallography in the USA, and for macro- In the publication[3] describing the first factor to be calculated using Bohr’s atomic molecular crystallography in England. X-ray diffraction experiments, Laue pro- model by Debye were described. Powder The powder method was developed with posed the Laue interference equations, diffraction was invented by Debye and highly efficient X-ray monochromators for and used them to interpret the diffraction Scherrer in 1917 and applied to the deter- film cameras and diffractometers. The con- photographs of cubic ZnS assuming a mination of the structure of graphite, and struction of new radiation sources such as primitive lattice and the presence of five independently at the same time by Hull for more efficient laboratory sources, synchro- different, well-defined wavelengths in the the determination of the structure of iron. trons and free-electron lasers, and of new X-ray beam. W. L. Bragg realized that this The model of the ideal mosaic crystal and two-dimensional area detectors continues interpretation was not entirely convincing, the theory of secondary extinction was for- to this day. and that these photographs could be inter- mulated by Darwin in 1922 in connection The heavy calculations necessary for preted more satisfactorily by supposing the with an experimental determination of the the solution of ever more complex crys- lattice to be face-centered and the incom- scattering factors of Na and Cl in NaCl by tal structures, mainly the summation of ing beam to contain X-rays with a contin- W. L. Bragg, James and Bosanquet. The Fourier series and the refinement of struc- uum of wavelengths that were reflected first space-group tables useful for structure tural parameters by least-squares meth- from the lattice planes. This specular re- determination were published by Niggli in ods, were major obstacles for now routine flection he described by the equation now 1919 (see below). work until a few decades ago. Modern known as Bragg’s law, 2dsinθ = nλ, where W. L. Bragg’s discovery of crystal struc- X-ray crystallography developed with the d is the interplanar spacing, θ the reflection ture determination gave him and his father invention of ever more efficient computers. angle, λ the wavelength and n an integer.