Noncollinear (NC) magnetism is common in nature, especially when there exist geometrical frustration and chemical imparity in the system. In this work we studied the NC magnetism and the response to external magnetic fields in surfaces and interfaces of transition metals by using an semi-empirical tight-binding (TB) method that parameterized to the ab initio TB-LMTO calculations. We implemented this method to study two systems. The first one is the system of 6 Mn monolayers on Fe(001) substrate. Due to the complex structure and magnetic properties of Mn, we found 23 collinear magnetic configurations but only one NC configuration. The collinear ground state has a layered antiferromagnetic (AFM) coupling which agrees with previous experiments and calculations. In the NC configuration the local AFM coupling in the Mn layers is preserved, but the surface is 90 degree coupled to the substrate. Similar to the experiment in CdCr2O4, we obtained a collinear plateau in the NC evolution of the average magnetic moment in Mn slab under external magnetic fields. Another is the system of a Cr monolayer on a stepped Fe(001) substrate. As expected, the local AFM coupling in the interface of Cr and Fe are preserved. However, the edge Cr atoms is about 90 degree coupled to their nearest Fe neighbors. We also simulated the procedure of adding more Cr coverages gradually to a Cr bilayer coverage. As coverages increase, the magnetic moments in the Cr interface reduce, and the collinear plateau becomes wider as coverages increase. However, the saturation fields in both the two systems are extremely high, around 10 kT.We expect that when the effect of temperature is taken into account, and in some proper systems, the saturation fields could be largely reduced to the scale that can be implemented in experiment, and our study may shed light on information storage devices with ultrahigh storage density.

The magnetocaloric effect (MCE) is the temperature change in a magnetic material due to a change in an applied magnetic field. How the MCE behaves in different magnetic materials and at different phase transitions is fundamental to understand. The driver of the MCE is a change in entropy which has multiple contributions: magnetic, lattice, and electron. In this thesis the MCE is studied in a simple antiferromagnetic (AFM) model andin a realistic noncollinear spin glass Neodymium model using Monte Carlo and Atomistic Spin Dynamics simulations. For the simple AFM system, no clear results were achieved, indicating that MCE in AFM materials is not due to a change solely in the magnetic entropy. For the complex magnetic material Nd, a more clear result is seen, indicating that frustration in the system might be important to the MCE in noncollinear materials. Nd results also signify more phase transitions than previously reported.