Convergent Beam Electron Diffraction of Ordered L10 Nanoparticles

Abstract

We have determined the long-range order parameter of individual [001] oriented FePt L1 0 nanoparticles. Measurement of convergent beam electron diffraction intensities of single particles in scanning transmission electron microscopy (STEM) mode, and comparison of the measured intensities to multislice simulations of diffracted beam intensities allow determination of the order parameter of the particle.FePt films were prepared by co-sputtering high purity Fe and Pt targets onto SiO2/Si substrates using ultra high vacuum dc magnetron sputtering. Ex-situ anneals between 500°C and 800°C in one atmosphere of a flowing reducing gas, argon with 3% hydrogen, were used to stabilize the isolated cluster morphology and to induce ordering. The composition of the Fe-Pt deposit was measured by Rutherford backscattering spectrometry (RBS) and determined to be 50 ±0.5 atomic % Fe.The CBED patterns were simulated with a multislice approach [1] for [001] oriented particles under conditions matching the experiment. The multislice approach specifically includes important dynamical scattering effects. Thermal effects were included using the frozen phonon method with root-mean-square deviations from the atomic positions of 0.007 nm for Fe and 0.0063 nm for Pt (estimated from the international table for X-ray crystallography [2]). The order parameter is included by statistically weighting each site in the structure with the correct fraction of the atomic species found on those sites. An example simulated CBED pattern is shown in Fig 2. In Fig. 3 we show the simulated I 110 /I 220 intensity ratio as a function of changing order parameter (S) for several thicknesses close to the observed particle thickness. Also in Fig 3 are two examples of experimental intensity ratios and the corresponding range of order parameter.Single particle diffraction data was acquired in the STEM mode of the TEM. A convergent beam of 4 mrad convergence semi-angle was focused on the specimen, and CBED patterns were collected from the ordered particles. The 4 mrad angle was chosen to give minimal overlap between the diffracted disks. Fig 4 shows an [001] oriented particle CBED pattern after subtraction of an offparticle background pattern.Dynamical scattering effects are seen in the oscillatory nature and flattening of the curves in Fig 3. This introduces uncertainties in the measurements of order parameter. With those uncertainties taken into account, the order parameter for the two cases shown is determined to be S = 0.4±0.1 and S=0.62±0.24 [3].References

Particle size and convergent electron diffraction patterns of triangular prismatic gold nanoparticles

“…New progresses have been also extensively done in convergent beam electron diffraction (CBED) for nano-sized materials. Normal CBED patterns of volumetric materials show dynamic contrast within the bright discs, as well as diffuse Kikuchi bands and sharp HOLZ lines, but the CBED patterns of nanoparticles only show bright kinematical discs [8]. In addition to this, Tehuacanero-Cuapa et al [14], trying to obtain CBED patterns in Au decahedral nanoparticles of less than 30 nm in size, found that the electron beam perforated the Au nanoparticles, thus concluding that the CBED analysis should be performed with particles larger than 30-nm-in-size to avoid the perforation [14,15].…”

“…Recent advances in transmission electron microscopy (TEM) allow obtaining electron diffraction patterns of nanoparticles as small as 5 nm in size, allowing to study their individual structure [8][9][10]. In many cases, the electron diffraction patterns of nanoparticles show "forbidden" reflections when they are observed by TEM [11,12].…”

Dynamical Diffraction Simulations in FePt—I

“…Consequently, the traditional methodology of taking the ratio of the total integrated peak intensities of the superlattice and fundamental reflections according to kinematical scattering theory (Warren, 1990), as done in X-ray diffraction, no longer applies, and S determination becomes more complex. Simulations of electron transmission in crystalline specimens including dynamical scattering are necessary to determine S (Petrova et al, 2005 a , 2005 b ). Figure 1 demonstrates the difficulty of predicting S for [001] and [111] directions in FePt using transmission electron microscopy.…”