Related resources
Search for item elsewhere
University researcher(s)
Magnetic Properties Of Two-Dimensional Materials: Graphene, Its Derivatives and Molybdenum Disulfide
[Thesis]. Manchester, UK: The University of Manchester; 2014.
Access to files
-
FULL-TEXT.PDF (pdf)
Abstract
Graphene, an atomically thin material consisting of a hexagonal, highly packed carbon lattice, is of great interests in its magnetic properties. These interests can be categorized in several fields: graphene-based magnetic materials and their applications, large diamagnetism of graphene, and the heterostructures of graphene and other two dimensional materials. In the first aspect, magnetic moments can be in theory introduced to graphene by minimizing its size or introducing structural defects, leading to a very light magnetic material. Furthermore, weak spin-orbital interaction, and long spin relaxation length make graphene promising for spintronics. The first part of this thesis addressed our experimental investigation in defect-induced magnetism of graphene. Non-interacted spins of graphene have been observed by intentionally introducing vacancies and adatoms through ion-irradiation and fluorination, respectively. The defect concentration or the magnetic moments introduced in this thesis cannot provide enough interaction for magnetic coupling. Furthermore, the spins induced by vacancies and adatoms can be controlled through shifting the Fermi energy of graphene using molecular doping, where the adatoms were alternatively introduced by annealing in the inert environment. The paramagnetic responses in graphene induced by vacancy-type defects can only be diverted to half of its maximum, while those induced by sp3 defects can be almost completely suppressed. This difference is supposed that vacancy-type defects induced two localized states (π and σ). Only the latter states, which is also the only states induced by sp3 defects, involves in the suppression of magnetic moments at the maximum doping achieved in this thesis. The observation through high resolution transmission electron microscope (HR-TEM) provides more information to the hypothesis of the previous magnetic findings. Reconstructed single vacancy is the majority of defects discovered in proton-irradiated graphene. This result verifies the defect-induced magnetic findings in our results, as well as the electronic properties of defected graphene in the literatures. On the other hand, the diamagnetic susceptibility of neutral graphene is suggested to be larger than that of graphite, and vanish rapidly as a delta-like function when graphene is doped. In our result, surprisingly, the diamagnetic susceptibility varies little when the Fermi level is less than 0.3 eV, in contrast with the theory. When the Fermi energy is higher than 0.3 eV, susceptibility then reduces significantly as the trend of graphite. The little variation in susceptibility near the Dirac point is probably attributed to the spatial confinement of graphene nanoflakes, which are the composition of graphene laminates. In the end of this thesis, we discuss the magnetic properties in one of the other two dimensional materials, molybdenum disulfide (MoS2). It is a potential material for graphene-based heterostructure applications. The magnetic moments in MoS2 are shown to be induced by either edges or vacancies, which are introduced by sonication or proton-irradiation, respectively, similar to the suggestions by theories. However, no significant ferromagnetic finding has been found in all of our cases.
Layman's Abstract
This thesis addressed the investigations of the magnetic properties of graphene and MoS2. The first part of our work is associated with defect-induced magnetism and the diamagnetism of graphene. The defects have been introduced by ion irradiation, fluorination, and annealing in the inert environment in this thesis. The former introduces vacancies and the latter two methods are using to attach sp3 defects to graphene. The other part is associated with the magnetic properties of MoS2, one of the potential two-dimensional materials which can be possibly applied for graphene hetertostructural devices. A Quantum Design MPMS SQUID magnetometer was used to measure the sensitive magnetic responses. Although the magnetometer is currently one of the most sensitive pieces of equipment available, milligrams of graphene laminates were required to produce detectable magnetic signals. Graphene and MoS2 laminates have been made by liquid-phase isolation of HOPG crystal and bulk MoS2 powder in NMP. The atomic-resolution micrographs were performed by hardware aberration-corrected FEI TITAN 80-300 operated at 80 kV, to study the geometry of defects in ion-irradiated graphene. In particular, chapter 3 displayed the experimental data associated with magnetic responses induced by point defects in graphene, supporting directly many theoretical studies. Fluorine adatoms which form sp3 bonds with carbon atoms were introduced by exposing graphene to fluorine atomic gas decomposed by XeF2. On the other hand, vacancy-type defects in graphene are created by ion-irradiation. Both types of defects were found to induce remarkable spin-half paramagnetism, but no magnetic ordering could be detected even for the largest defect density at T=1.8K. This agrees with the theoretical prediction. The magnetic moment achieved in our case is 0.1% of the maximum hypothetic moment of ~1μB per carbon atom. The fluorine clusters on carbon sheets lead to the inefficiency of inducing magnetic moments by fluorine adatoms. Instead of the contributions from all the fluorine adatoms, magnetic moments only come from edges or vacancies of clusters. On the other hand, structural integrity of ion-irradiated graphene laminates limited the vacancy concentration. Although the achieved magnetic moment is much larger than those of the ferromagnetic findings reported in the other carbon systems, the distance between two localized spins (~8nm), obtained from the atomic-scale observations in chapter 5, is too large for magnetic ordering. In chapter 4, the magnetic moments injected by introducing sp3 and vacancy-type defects in graphene laminates were demonstrated to be controlled by molecular doping. Although no ferromagnetic coupling has been found in either the graphene laminates before or after doping, we were interested in controlling the paramagnetic centres in two types of defected graphene laminates. The paramagnetic centres induced by sp3 defects are completely suppressed when EF~0.5eV, while only half of the initial magnetic moments are eliminated when the Fermi energy of graphene with vacancy-type defects shifting to the same level. This observation matches the scenario of dual origins of paramagnetic centres in defected carbon lattices. Two localized states, π and σ, are associated with imbalance of two sublattices and the dangling bonds, respectively. In principle, the former is expected to be existed in both types of defects while the latter only risen by vacancy-type defects. The distinguishable suppression of the magnetic moments in two types of defected graphene on the effect of doping reflect the π states is closed to the Fermi energy, while the energy for pairing the spins in the σ states is much higher than those achieved by chemical doping. Furthermore, the almost complete suppression of spins in sp3 defected graphene is suggested to be useful in spintronics development. On the other hand, large orbital diamagnetism of neutral graphene and its rapid reduction when graphene is doped are expected. However, the sharp changes of diamagnetic susceptibility were not observed in the molecular-doped graphene. Instead, when the Fermi energy is less than 0.3eV, almost no variation shows in diamagnetic susceptibility of graphene. When the Fermi energy is higher than 0.3eV, a reduction in the susceptibility comparable to graphite displayed. This is attributed to the spatial confinement due to graphene laminates are composed by the nanometer-sized flake.In chapter 5, the defected graphene encapsulated by graphene cells were demonstrated to be feasible to observe the intrinsic point defects avoiding their damage or transformation by electron beams. In particular, proton-irradiated trilayer graphene and systematic simulations were shown to provide a reliable identification of the location and types of defects through high resolution transmission electron microscope. This method can view intrinsic defects when it has similar contrast as the outer cell. In addition, the required clean process of in-situ annealing prior to high resolution imaging is altered by drop casting nanocrystallites of MoS2 due to the possible transformations of defects by annealing. Reconstructed single vacancies are in majority of the defects observed in all of our samples. This observation explains the magnetic and electronic properties discovered in the ion-irradiated graphene. The contradiction of this result to the previous observations, where more divacancies and other more complicated defects have been discovered, is explained as encapsulation protecting the defect from e-beam sensitive transformation. In the chapter 6, MoS2 laminate has been investigated about its magnetic properties on the effect of edges and vacancies through the modification by sonication and ion-irradiation, respectively. In all our cases, no significant ferromagnetic ordering has been found. The total length of edges of nanoflakes in MoS2 laminate produced by sonication shows significant effect on the paramagnetic signals of MoS2 laminates. The increase of total edge length is proportional to the addition of paramagnetic centres. This agrees with the theoretical expectation that zigzag and missing sulphur edges can induce the magnetic moments in MoS2. Moreover, the effect of ion irradiation on the magnetic properties of MoS2 was also studied. Paramagnetic signals induced by ion-irradiation are detected. Low density defects create the magnetic moments of 1.82μB per defect which agrees with 2μB per defect in the theoretical expectation. The total magnetic moments increase when the defect density increases, but the magnetic moments per defect were decreased. The reason of the latter phenomenon is still unknown, requiring further microscopic studies.