The D8 DISCOVER, equipped with the INCOATEC IµSHigh Brilliance Microfocus source, is an innovative X-ray diffraction (XRD) instrument, perfect for characterizing multipurpose modern materials. This article demonstrates the use of this system in an in-plane grazing incidence diffraction (IP-GID) configuration, which boosts the signal from very thin epitaxial and polycrystalline layers, and enables probing properties in the sample surface plane. This article shows how the resolution of the instrument in αI, φ and 2θ is defined by measuring a Si wafer with (001) orientation. The article also presents measurements of a polycrystalline copper thin film on a (001) Si wafer, in addition to measurements of a series of epitaxial SrTiO3 on (001) Si thin films.
Non-coplanar GID, or IP-GID, is used to examine the near-surface region of samples (10 or fewer nanometers under the air-sample interface). This method is carried out by the use of the high intensity of the total external reflection condition, and performing Bragg diffraction from planes that are almost perpendicular to the surface of the sample. Figure 1 illustrates the experimental geometries utilized for coplanar and non-coplanar diffraction. Conventional lab systems for performing IP-GID make use of a primary beam conditioned by a single multilayer mirror or polycapillary optic. Such systems may form low divergence perpendicular to the film, ensuring superior depth penetration control, and a relatively high divergence parallel to the film (in the scattering direction), resulting in poor peak resolution.
Figure 1. Coplanar diffraction geometry and with θi and θD (b) Non- coplanar diffraction geometry with incident angle (αI) and exit angle (αI) from the sample surface with 2θIP-GID the Bragg scattering angle.
Figure 2 shows the configuration to perform IP-GID, the details of the configuration are presented in Table 1. The IµS, integrated with the MONTEL optic, generates a high-intensity primary beam with a spot size diameter of 1 mm and equatorial and axial divergence of less than 0.1°. This enables exceptional depth control and equatorial resolution, because of the symmetric divergence. A 0.2° equatorial soller is used in 0D mode in front of the LYNXEYE XE, producing a diffracted beam with a resolution of 0.2°. Figure 3 illustrates a schematic diagram of geometry of the IP-GID performed using the IµS source.
Table 1. IP-GID instrument setup for the D8 DISCOVER with IµS
||IµS Microfocus (Cu)
||Centric Eulerian Cradle (CEC)
||0.2° equatorial Soller collimator
||LYNXEYE XE (0D)
Figure 2. D8 DISCOVER with IµS and LYNXEYE XE detector configured for IP- GID
Figure 3. IP-GID geometry with the IµS, MONTEL optic, and secondary equatorial soller collimator.
Instrument Resolution Characterization Using a (001) Si Wafer
A 25 x 25 mm piece of a (001) Si wafer was utilized to characterize the instrumental resolution of the conventional (single Göbel mirror), as well as the IµS IP-GID system geometries. Figure 4a displays a scan of the (004) Si in-plane reflection at the inclination angle αI. The intensity and resolution provided by the IP-GID geometry - based on the IµS - is better when compared to the conventional technique. Figure 4b illustrates a phi scan of the (004) Si in-plane reflection whose FWHM is established to be 0.07°. This result is consistent with the equatorial divergence of the MONTEL. The conventional IP-GID geometry with 0.3° FWHM is consistent with respect to the 0.3° incident equatorial soller. Figure 4c presents a coupled 2θ-ω scan of the (004) Si in-plane reflection, and clearly depicts the Kα1 and Kα2 splitting in the improved IP-GID result (FWHM of Kα1 = 0.14°), whereas the conventional IP-GID instrumental geometry is too wide to view the splitting.
Figure 4. IP-GID measurements of a (001) Si wafer performed using the conventional IP-GID method and the IµS IP-GID: (a) Scan of the inclination angle (αI), (b) Phi scan and (c) coupled 2θ-ω scan.
Measurement of a Polycrystalline Cu Film on Silicon Substrate
Multiple coatings are coated on the materials to modify the surface properties, while maintaining the bulk properties. This could induce electrical conductivity along an insulating surface or to alter the surface finish. The thickness of the coatings is in the order of microns to nanometers. For layers with micro-thickness GID is usually employed by maintaining the incident beam angle to be below 1°, while using the detector to perform a scan over a broader range of angles. The low incident beam angle prevents the beam from penetrating into the substrate. However in a traditional Bragg-Brentano coupled geometry, the incident angle is scanned according to the detector angle, increasing the penetration depth of the incident beam during the progression of the scan (Figure 5).
Measurement of a Polycrystalline Cu film on Silicon Substrate
Many coatings are applied to materials to alter the properties of the surface while maintaining properties of the bulk. This could be to provide electrical conductivity along an insulating surface or to change the surface finish. These coatings can range in thickness from microns to nanometers. For micron thick layers a geometry called Grazing Incidence Diffraction is commonly used, where the incident beam angle is kept very low - typically below 1° - while the detector is scanned over a wide range of angles. Since the incident beam angle is kept low, the beam does not penetrate to the substrate. In a classic Bragg-Brentano coupled geometry the incident angle is scanned with the detector angle, resulting in an increase of penetration depth of the incident beam as the scan progresses. It is important to note the direction being probed in the material. In a standard coupled scan, the direction being probed is normal to the surface of the material. In the grazing incident method, the direction being probed is close to normal to the surface at the beginning of the scan, but tilts towards the plane of the sample surface as the scan progresses. In IP-GID geometry the angle of inclination to the surface (ai) is fixed near the critical edge of the material, while a detector scan is performed in a direction parallel to the surface of the sample. This results in extremely low penetration depth and an enhancement of the signal coming from the surface. In this case surface sensitivity is achieved, and the direction being probed is in the plane of the surface of the sample. Figure 4 shows a scan of a Cu thin film on (001) Si. The normal bisecting geometry scan shows only the (111) and (222) Cu peaks along with the (004) Si peak indicating the Cu  axis being oriented normal to the surface of the sample. The Grazing Incidence Diffraction scan was collected with a fixed incident angle of 1°. All Cu reflections are now present as the direction being measured is no longer normal to the surface. The Si (004) reflection is no longer present, as the direction being probed is not normal to the surface of the sample. The In-Plane Grazing Incidence Diffraction scan shows not only the Cu peaks, but also peaks coming from a surface oxide phase of Cu2O.
Figure 5. Normal “Bragg-Brentano” Diffraction (blue), Grazing Incidence Diffraction (red) and In-Plane Grazing Incidence Diffrac-tion (black) measurements using the IµS IP-GID geometry of a copper film on (001) Si.
In addition, it is vital to make a note of the direction being probed in the material. For a standard coupled scan this direction is always normal to the material’s surface. For the grazing incident technique, the direction being probed is almost normal to the surface while beginning the scan. However it tilts towards the plane of the surface of the sample when the scan progresses. In the IP-GID geometry the angle of inclination with respect to the surface (αI) is fixed close to the material’s edge while performing a detector scan in a direction parallel to the sample surface. This leads to very low penetration depth and improved signal emitted from the surface. Here the surface sensitivity is accomplished and the direction being probed is located towards the plane of the sample surface. Figure 4 illustrates a scan of a Cu thin film on (001) Si. The usual bisecting geometry scan can show only the (111) and (222) Cu peaks along with the (004) Si peak, demonstrating orientation of the Cu  axis normal to the sample surface. The GID scan was obtained at a fixed incident angle of 1°. All Cu reflections are now enabled as the direction being probed is no longer normal to the surface. The Si (004) reflection is now absent as the direction being probed is not normal to the sample surface. The IP-GID scan provides the Cu peaks, as well as the peaks received from a surface oxide phase of Cu2O.
Measurement of Epitaxial SrTiO3 on Si (001)
SrTiO3 is usually utilized as a buffer layer to perform the growth of perovskite-based multiferroic materials like PMNPT or LaAlO3 on Si. There is a huge lattice mismatch between SrTiO3 (a = 3.905 Å) and Si (a = 5.437 Å). This results in a 45° in-plane rotation of the SrTiO3, which aligns the SrTiO3 (100) axis with the Si (110) axis (5.437/v2 = 3.845 Å). Figure 6 illustrates this in-plane relationship, where the (200) SrTiO3 reflection demonstrates a clear 45° rotation relating to the (400) Si reflection.
Figure 6. Phi scan of the in-plane (200) SrTiO3 and (400) Si reflections measured using the IµS IP-GID geometry showing the in-plane 45° rotation of the SrTiO3 relative to the Si.
SrTiO3 films with different thicknesses deposited by molecular beam epitaxy, with thicknesses spanning between 8 and 100 nm were examined for in-plane relaxations effects. For the (220) Si reflection, Phi and αI were optimized. Then, a 2θ-ω scan was obtained and αI was optimized for the SrTiO3 (200) reflection to enhance the signal received from the film. Figure 7 depicts the difference in optimization of αI on the Si and SrTiO3 reflection for the 40 nm film. The result of a 2θ- ω scan obtained from 44°-49° 2θ, with 0.025° step size at the rate of 5 seconds per step for a total scan time of 17 minutes, is shown in Figure 8. Figure 9 illustrates the result of a similar analysis carried out on the Si (400) and SrTiO3 (220) reflections. In-plane reciprocal space maps for the SrTiO3 (200) and Si (220) reflections on the 8 and 100 nm films were collected. The in-plane alignment of the SrTiO3 related to the thickness is consistent. The crystalline quality and in-plane mosaic spread, which are respectively related to the film spot height and width, are decreased in the thicker film (Figure 10)
Figure 7. 2θ-ω scans of the in-plane (200) SrTiO3 and (220) Si reflections with αI optimized for SrTiO3 (red) and Si (black) measured using the IµS IP-GID geometry.
Figure 8. 2θ-ω scans of the in-plane (200) SrTiO3 and (220) Si reflections for a series of SrTiO3 film thicknesses ranging from 8nm to 100nm measured using the IµS IP-GID geometry.
Figure 9. 2θ-ω scans of the in-plane (220) SrTiO3 and (400) Si reflections for SrTiO3 film thicknesses ranging from 8 to 100nm measured using the IµS IP-GID geometry.
Figure 10. In-Plane Reciprocal Space Maps of the (200) SrTiO3 and (220) Si reflections for the 8nm and 100nm thick SrTiO3 on Si samples measured using the enhanced IµS IP-GID geometry.
The D8 DISCOVER with IµS can be used to collect data in an IP-GID geometry from different samples such as a bare Si wafer, fiber textured copper film on Si, and epitaxial SrTiO3 films on Si. The IµS, combined with MONTEL optic, produces a beam with very high flux density and low divergence, providing an ideal configuration for IP-GID by enabling a strong coupling with the surface and maintaining high resolution of the equatorial scattering plane. Configuration of a Centric Eulerian Cradle stage in the D8 DISCOVER with IµS enhances flexibility, eliminating the need for dedicated hardware for IP-GID. This system is highly suitable for laboratories handling modern multipurpose materials research.
This information has been sourced, reviewed and adapted from materials provided by Bruker AXS Inc.
For more information on this source, please visit Bruker AXS Inc.