International
Tables for Crystallography Volume H Powder diffraction Edited by C. J. Gilmore, J. A. Kaduk and H. Schenk © International Union of Crystallography 2018 
International Tables for Crystallography (2018). Vol. H, ch. 2.5, pp. 119123
Section 2.5.2. Fundamentals^{a}Bruker AXS Inc., 5465 E. Cheryl Parkway, Madison, WI 53711, USA 
A conventional powderdiffraction pattern is displayed as the scattering intensity versus scattering angle 2θ or dspacing. A 2DXRD pattern contains the scatteringintensity distribution as a function of two orthogonal dimensions. One dimension can be expressed in 2θ, which can be interpreted by Bragg's law. The distribution in the dimension orthogonal to 2θ contains additional information, such as the orientation distribution, strain states, and crystallitesize and shape distribution. In order to understand and analyse 2D diffraction data, new geometry conventions and algorithms are introduced. The geometry conventions and algorithms used for 2DXRD should also be consistent with conventional XRD, so that many existing concepts and algorithms are still valid when 2D diffraction data are used.
The geometry of a 2DXRD system can be explained using three distinguishable and interrelated geometry spaces, each defined by a set of parameters (He, 2003). The three geometry spaces are the diffraction space, detector space and sample space. The laboratory coordinate system X_{L}, Y_{L}, Z_{L} is the basis of all three spaces. Although the three spaces are interrelated, the definitions and corresponding parameters should not be confused. Except for a few parameters introduced specifically for 2DXRD, many of these parameters are used in conventional Xray diffraction systems. Therefore, the same definitions are maintained for consistency. The threecircle goniometer in Eulerian geometry is the most commonly used, and all the algorithms for data interpretation and analysis in this chapter are based on Eulerian geometry. The algorithms can be developed for the geometries of other types (such as kappa) by following the same strategies.
Fig. 2.5.3(a) describes the geometric definition of diffraction cones in the laboratory coordinate system X_{L}, Y_{L}, Z_{L}. The laboratory coordinate system is a Cartesian coordinate system. The plane given by X_{L} and Y_{L} is the diffractometer plane. The axis Z_{L} is perpendicular to the diffractometer plane. The axes X_{L}, Y_{L} and Z_{L} form a righthanded rectangular coordinate system with the origin at the instrument centre. The incident Xray beam propagates along the X_{L} axis, which is also the rotation axis of all diffraction cones. The apex angles of the cones are determined by the 2θ values given by the Bragg equation. The apex angles are twice the 2θ values for forward reflection and twice the value of 180° − 2θ for backward reflection . For clarity, only one diffraction cone of forward reflection is displayed. The γ angle is the azimuthal angle from the origin at the six o'clock direction with a righthanded rotation axis along the opposite direction of incident beam (−X_{L} direction). A given γ value defines a half plane with the X_{L} axis as the edge; this will be referred to as the γ plane hereafter. The diffractometer plane consists of two γ planes at γ = 90° and γ = 270°. Therefore many equations developed for 2DXRD should also apply to conventional XRD if the γ angle is given as a constant of 90° or 270°. A pair of γ and 2θ values represents the direction of a diffracted beam. The γ angle takes a value of 0 to 360° for a complete diffraction ring with a constant 2θ value. The γ and 2θ angles form a spherical coordinate system which covers all the directions from the origin of sample (instrument centre). The γ–2θ system is fixed in the laboratory system X_{L}, Y_{L}, Z_{L}, which is independent of the sample orientation and detector position in the goniometer. 2θ and γ are referred to as the diffractionspace parameters. In the laboratory coordinate system X_{L}, Y_{L}, Z_{L}, the surface of a diffraction cone can be mathematically expressed aswith or for forwarddiffraction cones and or for backwarddiffraction cones.
Fig. 2.5.3(b) shows the diffractionvector cone corresponding to the diffraction cone in the laboratory coordinate system. C is the centre of the Ewald sphere. The diffraction condition can be given by the Laue equation aswhere s_{0} is the unit vector representing the incident beam, s is the unit vector representing the diffracted beam and H_{hkl} is the reciprocallattice vector. Its magnitude is given asin which d_{hkl} is the dspacing of the crystal planes (hkl). It can be easily seen that it is the Bragg law in a different form. Therefore, equation (2.5.2) is the Bragg law in vector form. In the Bragg condition, the vectors s_{0}/λ and s/λ make angles θ with the diffracting planes (hkl) and H_{hkl} is normal to the (hkl) crystal plane. In order to analyse all the Xrays measured by a 2D detector, we extend the concept to all scattered Xrays from a sample regardless of the Bragg condition. Therefore, the index (hkl) can be removed from the above expression. H is then a vector which takes the direction bisecting the incident beam and the scattered beam, and has dimensions of inverse length given by . Here 2θ is the scattering angle from the incident beam. The vector H is referred to as the scattering vector or, alternatively, the diffraction vector. When the Bragg condition is satisfied, the diffraction vector is normal to the diffracting lattice planes and its magnitude is reciprocal to the dspacing of the lattice planes. In this case, the diffraction vector is equivalent to the reciprocallattice vector. Each pixel in a 2D detector measures scattered Xrays in a given direction with respect to the incident beam. We can calculate a diffraction vector for any pixel, even if the pixel is not measuring Bragg scattering. Use of the term `diffracted beam' hereafter in this chapter does not necessarily imply that it arises from Bragg scattering.
For twodimensional diffraction, the incident beam can be expressed by the vector s_{0}/λ, but the diffracted beam is no longer in a single direction, but follows the diffraction cone. Since the direction of a diffraction vector is a bisector of the angle between the incident and diffracted beams corresponding to each diffraction cone, the trace of the diffraction vectors forms a cone. This cone is referred to as the diffractionvector cone. The angle between the diffraction vector and the incident Xray beam is 90° + θ and the apex angle of a vector cone is 90° − θ. It is apparent that diffractionvector cones can only exists on the −X_{L} side of the diffraction space.
For twodimensional diffraction, the diffraction vector is a function of both the γ and 2θ angles, and is given in laboratory coordinates asThe direction of the diffraction vector can be represented by its unit vector, given bywhere h_{L} is a unit vector expressed in laboratory coordinates and the three components in the square brackets are the projections of the unit vector on the three axes of the laboratory coordinates, respectively. If γ takes all values from 0 to 360° at a given Bragg angle 2θ, the trace of the diffraction vector forms a diffractionvector cone. Since the possible values of θ lie within the range 0 to 90°, h_{x} takes only negative values.
A typical 2D detector has a limited detection surface, and the detection surface can be spherical, cylindrical or flat. Spherical or cylindrical detectors are normally designed for a fixed sampletodetector distance, while a flat detector has the flexibility to be used at different sampletodetector distances so as to choose either high resolution at a large distance or large angular coverage at a short distance.
The position of a flat detector is defined by the sampletodetector distance D and the detector swing angle α. D and α are referred to as the detectorspace parameters. D is the perpendicular distance from the goniometer centre to the detection plane and α is a righthanded rotation angle about the Z_{L} axis. Detectors at different positions in the laboratory coordinates X_{L}, Y_{L}, Z_{L} are shown in Fig. 2.5.4. The centre of detector 1 is right on the positive side of the X_{L} axis (onaxis), α = 0. Both detectors 2 and 3 are rotated away from the X_{L} axis with negative swing angles (α_{2} < 0 and α_{3} < 0). The detection surface of a flat 2D detector can be considered as a plane, which intersects the diffraction cone to form a conic section. Depending on the swing angle α and the 2θ angle, the conic section can appear as a circle, an ellipse, a parabola or a hyperbola.
The values of 2θ and γ can be calculated for each pixel in the frame. The calculation is based on the detectorspace parameters and the pixel position in the detector. Fig. 2.5.5 shows the relationship of a pixel P(x, y) to the laboratory coordinates X_{L}, Y_{L}, Z_{L}. The position of a pixel in the detector is defined by the (x, y) coordinates, where the detector centre is defined as x = y = 0. The diffractionspace coordinates (2θ, γ) for a pixel at P(x, y) are given by
The conic sections of the diffraction cones with a curved detector depend on the shape of the detector. The most common curved detectors are cylindershaped detectors. The diffraction frame measured by a cylindrical detector can be displayed as a flat frame, typically a rectangle. Fig. 2.5.6(a) shows a cylindrical detector in the vertical direction and the corresponding laboratory coordinates X_{L}, Y_{L}, Z_{L}. The sample is located at the origin of the laboratory coordinates inside the cylinder. The incident Xrays strike the detector at a point O if there is no sample or beam stop to block the direct beam. The radius of the cylinder is R. Fig. 2.5.6(b) illustrates the 2D diffraction image collected with the cylindrical detector. We take the point O as the origin of the pixel position (0, 0). The diffractionspace coordinates (2θ, γ) for a pixel at P(x, y) are given byThe pixelpositionto(2θ, γ) conversion for detectors of other shapes can also be derived. Once the diffractionspace coordinates (2θ, γ) of each pixel in the curved 2D detector are determined, most dataanalysis algorithms developed for flat detectors are applicable to a curved detector as well.
In a 2DXRD system, three rotation angles are necessary to define the orientation of a sample in the diffractometer. These three rotation angles can be achieved either by a Eulerian geometry, a kappa (κ) geometry or another kind of geometry. The three angles in Eulerian geometry are ω (omega), ψ (psi) and ϕ (phi). Fig. 2.5.7(a) shows the relationship between rotation axes (ω, ψ, ϕ) in the laboratory system X_{L}, Y_{L}, Z_{L}. The ω angle is defined as a righthanded rotation about the Z_{L} axis. The ω axis is fixed in the laboratory coordinates. The ψ angle is a righthanded rotation about a horizontal axis. The angle between the ψ axis and the X_{L} axis is given by ω. The ψ axis lies on X_{L} when ω is set at zero. The ϕ angle defines a lefthanded rotation about an axis on the sample, typically the normal of a flat sample. The ϕ axis lies on the Y_{L} axis when ω = ψ = 0. In an aligned diffraction system, all three rotation axes and the primary Xray beam cross at the origin of the X_{L}, Y_{L}, Z_{L} coordinates. This cross point is also known as the goniometer centre or instrument centre.

Sample rotation and translation. (a) Three rotation axes in laboratory coordinates; (b) rotation axes (ω, ψ, ϕ) and sample coordinates. 
Fig. 2.5.7(b) shows the relationship and stacking sequence among all rotation axes (ω, ψ, ϕ) and the sample coordinates S_{1}, S_{2}, S_{3}. ω is the base rotation; all other rotations and translations are on top of this rotation. The next rotation above ω is the ψ rotation. The next rotation above ω and ψ is the ϕ rotation. The sample coordinates S_{1}, S_{2}, S_{3} are fixed to the sample regardless of the particular sample orientation given by the rotation angles (ω, ψ, ϕ). The ϕ rotation in the goniometer is intentionally chosen as a lefthanded rotation so that the diffraction vectors will make a righthand rotation observed in the sample coordinates S_{1}, S_{2}, S_{3}.
In 2DXRD data analysis, it is crucial to know the diffractionvector distribution in terms of the sample coordinates S_{1}, S_{2}, S_{3}. However, the diffractionvector distribution corresponding to the measured 2D data is always given in terms of the laboratory coordinates X_{L}, Y_{L}, Z_{L} because the diffraction space is fixed to the laboratory coordinates. Fig. 2.5.8 shows the unit vector of a diffraction vector in both (a) the laboratory coordinates X_{L}, Y_{L}, Z_{L} and (b) the sample coordinates S_{1}, S_{2}, S_{3}. In Fig. 2.5.8(a) the unit vector h_{L} is projected to the X_{L}, Y_{L} and Z_{L} axes as h_{x}, h_{y} and h_{z}, respectively. The three components are given by equation (2.5.5). In order to analyse the diffraction results relative to the sample orientation, it is necessary to transform the unit vector to the sample coordinates S_{1}, S_{2}, S_{3}. Fig. 2.5.8(b) shows the same unit vector, denoted by h_{s} projected to S_{1}, S_{2} and S_{3} as h_{1}, h_{2} and h_{3}, respectively.
The transformation of the unit diffraction vector from the laboratory coordinates X_{L}, Y_{L}, Z_{L} to the sample coordinates S_{1}, S_{2}, S_{3} is given bywhere A is the transformation matrix. For Eulerian geometry in matrix form, we haveIn expanded form:In addition to the diffraction intensity and Bragg angle corresponding to each data point on the diffraction ring, the unit vector h_{s}{h_{1}, h_{2}, h_{3}} provides orientation information in the sample space. The transformation matrix of any other goniometer geometry, such as kappa geometry (Paciorek et al., 1999), can be introduced into equation (2.5.10) so that the unit vector h_{s}{h_{1}, h_{2}, h_{3}} can be expressed in terms of the specified geometry. All equations using the unit vector h_{s}{h_{1}, h_{2}, h_{3}} in this chapter, such as in data treatment, texture analysis and stress measurement, are applicable to all goniometer geometries provided that the unitvector components are generated from the corresponding transformation matrix from diffraction space to the sample space.
Reciprocalspace mapping is commonly used to analyse the diffraction patterns from highly oriented structures, diffuse scattering from crystal defects, and thin films (Hanna & Windle, 1995; Mudie et al., 2004; Smilgies & Blasini, 2007; Schmidbauer et al., 2008). The equations of the unitvector calculation given above can also be used to transform the diffraction intensity from the diffraction space to the reciprocal space with respect to the sample coordinates. The direction of the scattering vector is given by the unit vector h_{s}{h_{1}, h_{2}, h_{3}} and the magnitude of the scattering vector is given by , so that the scattering vector corresponding to a pixel is given byThe threedimensional reciprocalspace mapping can be obtained by applying the normalized pixel intensities to the corresponding reciprocal points. With various sample orientations, all pixels on the detector can be mapped into a 3D reciprocal space.
References
Hanna, S. & Windle, A. H. (1995). A novel polymer fibre diffractometer, based on a scanning Xraysensitive chargecoupled device. J. Appl. Cryst. 28, 673–689.Google ScholarHe, B. B. (2003). Introduction to twodimensional Xray diffraction. Powder Diffr. 18, 71–85.Google Scholar
Mudie, S. T., Pavlov, K. M., Morgan, M. J., Hester, J. R., Tabuchi, M. & Takeda, Y. (2004). Collection of reciprocal space maps using imaging plates at the Australian National Beamline Facility at the Photon Factory. J. Synchrotron Rad. 11, 406–413.Google Scholar
Paciorek, W. A., Meyer, M. & Chapuis, G. (1999). On the geometry of a modern imaging diffractometer. Acta Cryst. A55, 543–557.Google Scholar
Schmidbauer, M., Schäfer, P., Besedin, S., Grigoriev, D., Köhler, R. & Hanke, M. (2008). A novel multidetection technique for threedimensional reciprocalspace mapping in grazingincidence Xray diffraction. J. Synchrotron Rad. 15, 549–557.Google Scholar
Smilgies, D.M. & Blasini, D. R. (2007). Indexation scheme for oriented molecular thin films studied with grazingincidence reciprocalspace mapping. J. Appl. Cryst. 40, 716–718.Google Scholar