UC Berkeley

This thesis studies the unique properties of Cr and Cr-Al alloys; the first half focuses on Cr while the second half focuses on Cr-Al alloys. Both Cr and Cr-Al alloys have sharp features in their d bands which affect their magnetic properties and ultimately lead to anomalous electrical transport. Although the specifics of the element and the alloy are quite different, they are united by the sensitivity of their magnetic and electronic states to external perturbation. This thesis particularly focuses on the effects of nanoscale structure such as crystal defects, grain boundaries, and short- to medium-range chemical ordering, on both the magnetism and the electronic transport properties of Cr and Cr-Al.

Bulk chromium has an incommensurate spin density wave (ISDW) and has been widely studied as an archetypal band antiferromagnet. The ISDW results in a sinusoidal modulation of the antiferromagnetically aligned moments; this is due to delicate nesting of the Fermi surface which is easily disrupted by perturbation. Thus, the SDW transitions from incommensurate to commensurate (CSDW) or to paramagnetic with small amounts of dopant atoms (Mn or V) or with the application of pressure. These effects have been well studied in bulk Cr.

The 1988 discovery of giant magnetoresistance (GMR) in Fe/Cr multilayers, which was awarded the 2007 Nobel prize in physics, inspired further research on the SDW in Cr, and shifted the focus of that research towards films and multilayers, where variables such as thickness, strain, and disorder are crucial. Until recently, most studies of the SDW in Cr thin films have focused on ultrathin, epitaxial films; however many of the Fe/Cr multilayers studied in the literature are polycrystalline. In fact, the degree of disorder in a multilayer is an important variable, as some research has analyzed the effects of surface roughness on GMR. This thesis aimed to understand the SDW in polycrystalline Cr films such as those commonly used in GMR multilayers, where disorder and stress are the important variables.

Infrared reflectivity was used to measure the characteristic SDW pseudogap energies to distinguish the SDW state of Cr thin films grown under different deposition conditions (e-beam and sputtered at different argon pressures). The fundamental distinguishing properties of the films are stress and disorder, both strongly affected by the deposition conditions. Films with low stress and disorder are ISDW, like bulk Cr. Films with high tensile stress are CSDW, like Mn-doped Cr. Finally, films with high disorder, determined from the resistivity, have regions of both ISDW and CSDW. Importantly, all of the Cr films measured showed SDW signatures, showing that the SDW is quite robust even in highly disordered thin films. A low temperature magnetic phase diagram was created for Cr films.

The SDW in Cr also leads to anomalous features in the electrical resistivity due to resonant impurity scattering. This occurs when impurities form quasilocalized states within the SDW pseudogap. When the quasilocal states are near the Fermi energy, resonant scattering occurs and causes features such as very high residual resistivity and a resistivity minimum with temperature. This has been studied in bulk samples due to dopant impurities, and theorized to occur for lattice defects such as vacancies as well. However, the defect concentrations in bulk are very low so this was not observed until our measurements on polycrystalline films.

It was shown that Cr thin films show unusual and extremely deposition condition-dependent resistivity due to resonant scattering, such as residual resistivity ranging between 3 and 400 μΩ-cm, and significant resistivity minima at low temperature. Several experiments showed that these features are due to defects in the Cr lattice such as grain boundaries and vacancies. When a highly disordered, 400 μΩ-cm film with a significant minimum is annealed to 800°C, the resistivity is decreased by 10× and the depth of the minimum is decreased by 50×. On the other end of the spectrum, two low resistivity (< 10 μΩ-cm) samples grown in the same run but on different substrates show small but noticeable different sized resistivity minima (0.01 and 0.003 μΩ-cm) because one is polycrystalline and the other epitaxial.

The anomalous resistivity and SDW behavior observed in Cr films led to the study of Cr_1-xAl_x alloys. Cr_1-xAl_x exhibits semiconducting behavior for x ∼ 0.25. Initially, researchers studying Cr_1-xAl_x suggested that the SDW pseudogap, which eliminates about 30% of the Fermi surface in pure Cr, may eliminate the entire Fermi surface in Cr_1-xAl_x, leading to a complete gap. However, the SDW gap primarily affects d electrons, while conduction occurs primarily through s electrons, so this suggestion does not explain the observed behavior.

The peak resistivity occurs around x ∼ 0.25, suggesting a stoichiometric Cr₃Al compound could be responsible for the semiconducting behavior. Such a compound was suggested by a previous electron diffraction study, but the mechanism for affecting the transport behavior was not explained until now. The results of this thesis indicate that the semiconducting behavior in Cr_1-xAl_x is due to a combination of a stoichiometric Cr₃Al compound causing a hybridization gap on one part of the Fermi surface with the SDW gap eliminating another part.

The atoms in Cr₃Al are observed to occupy the sites of a bcc lattice, like Cr. Density functional theoretical calculations were performed to compare possible types of chemical ordering and showed that the Cr₃Al structure proposed from electron diffraction, a chemically ordered rhombohedrally distorted phase with ordering along the <111> direction, is the lowest energy of those considered. In addition, the band structure for this structure shows a pseudogap, consistent with the observed transport behavior of Cr₃Al.

Experimental results also support the importance of an ordered phase. Nonequilibrium thin films of Cr_1-xAl_x were grown at different substrate temperatures to vary the properties. Samples grown below 400°C are semiconducting, while samples grown above 400°C are metallic. This is consistent with the proposed 400°C phase boundary for the ordered Cr₃Al structure.

The SDW pseudogap also plays an important role in the semiconducting behavior, but this is difficult to measure experimentally. The Neel temperature of Cr₃Al is about 500°C, at which point the resistivity is already quite metallic. For this reason, two previous studies on the resistivity around the Neel temperature came to different conclusions about the role of the SDW pseudogap on the semiconducting behavior.

To clarify this, neutron diffraction was performed to study the SDW state of Cr_1-xAl_x films. Because of the sensitivity of the SDW state to deposition conditions in Cr thin films, and the significant variation in transport properties of Cr_1-xAl_x films grown at different temperatures, a change in the magnetic state may be expected to accompany the variation in transport properties in Cr_1-xAl_x. It was found that both metallic and semiconducting Cr_1-xAl_x films had robust antiferromagnetism, with Neel temperatures above the highest measured temperatures (∼ 600K).

Theoretical results quite clearly suggest that antiferromagnetism is a necessary condition for the semiconducting behavior. The density of states for antiferromagnetic and nonmagnetic Cr₃Al were calculated and show that the pseudogap is eliminated in the nonmagnetic case. Thus, antiferromagnetism was shown to be a necessary but not sufficient condition for producing the semiconducting behavior in Cr_1-xAl_x.

It is thus concluded that the semiconducting behavior in Cr₃Al arises from a combination of the antiferromagnetic pseudogap and a rhombohedral-type chemical ordering of the bcc lattice.