Resumen de Development and characterization of graphene-based electronic devices
This thesis encloses a set of studies centered in graphene, a two-dimensional material with extraordinary electrical, optical and mechanical properties. The research has been conducted from the technological point of view, on the boundary between material science and electrical engineering, basic physics and its applications. In the framework of the present research work, both basic material characterization, and micro- and nano-fabrication techniques have been adopted for the study of graphene.
To understand the importance of this material, it is inviting to go back to its discovery in 2004 by Geim and Novoselov. The general belief at the time, derived from previous thermodynamical calculations, was that two-dimensional crystals could not exist isolated in a stable way. Against this prejudgment, the research group from Manchester was able to achieve the remarkable goal of isolating graphene. Using rudimentary methods to exfoliate a single layer from graphite, they were capable to measure the electronic properties of graphene, in which they verified a linear energy-momentum dispersion. This dispersion relation lead them to discover its extraordinarily high charge carrier mobility, which indicates that electrons in graphene travel in a ballistic regime. Such a high mobility is a desirable attribute for electronic applications where high frequency is required, and it has been one of the reasons why graphene has drawn the attention of the scientific community.
Since this discovery, the amount of research groups focused on this material has increased rapidly, pushing forward the state of the art in the field very quickly. In the current stage of the field, several issues have aroused regarding this material. On the long-term, the most important issue is to find the niche application where the material excels. Graphene has proved to be an excellent solution for a variety of subjects, such as flexible electronics, energy harvesting and storage, and high speed electronics, to cite some. For now, it seems that none of them are ready for the market, either due to the cost, or to the fact that the existing solutions exhibit better characteristics than graphene. In order to solve this first issue, the production means need to be capable of fabricating cheaper graphene, in large quantities and of the best quality. Therefore, the optimization of the growth methods is a short term priority. Among the existing techniques, chemical vapor deposition (CVD) is one of the most spread ones, as highly ordered material in large quantities can be obtained. In this growth method, carbon is deposited on the surface of a metallic catalyst from a gaseous carbon source. Nevertheless, both the nucleation density and the growth rate need to be further improved. In this thesis, a preliminary study of the growth conditions is detailed, where an oxygen treatment of the catalytic substrate is proved to overcome these issues, in agreement with results published elsewhere.
A problem derived from the CVD growth method is the need to transfer the material from the catalytic substrate to a more suitable one, typically a dielectric. Even being continuously optimized, this technological process is commonly performed manually. Only big companies, such as Sony or Samsung, have been capable of developing an automatized version of the process. Even for these cases, the choice of substrate is limited, as these methods are based in roll-to-roll processing, and hence its application is restricted to only flexible substrates. In this thesis, an alternative automatic transfer method that accepts rigid target substrates has been designed and tested. When compared to the common manual method, it has been demonstrated that the developed transfer method improves the yield of the process, resulting in cleaner samples with less tensile strain. For reaching these conclusions, a complete processing methodology has been implemented at ISOM, where graphene field-effect transistors (GFETs) have been fabricated and characterized. This device structure is quite versatile in the case of graphene, as its electrical output characteristics are closely related to the material quality and its intrinsic physical properties, such as the electron and hole mobilities, or even the possibility to select the charge carrier type through a controlled electrical field.
Finally, the last high-priority problem faced in this thesis is how to standardize and ensure the quality of the graphene-based devices. This activity is essential for the electronics industry, where small quality issues during the processing steps or on the final products can drastically reduce the production yield, which may affect badly to competitiveness of any company from the sector. With help of precise, sensitive characterization methods, this issue can be overcome. For this purpose, electrical and optical characterization methods have been adapted and developed during this thesis. In them, a deep connection with physical models allows to extract quantitative information from the fabricated GFET devices. As a remarkable property of the electrical method, information from each carrier type can be obtained at once. Regarding the optical method, Raman spectroscopy mapping has been assessed on the GFET channel. By means of an algebraic transformation, a separation between doping and strain effects on the material allows to quantify both effects.
Once these characterization methods have been developed, our first step has been to test the results that they provide against data from the literature, with successful results. As the effectiveness of the methods has been ensured, devices from manually and automatically transferred samples have been characterized. The outcome of this characterization, as mentioned earlier, is that devices fabricated using the in-house automatic transfer method exhibit higher quality when compared to devices fabricated with the manual method.
The development of the technologies associated with graphene is still in its early stage, as its incorporation to the existing industrial applications is out of reach in the short term. Nevertheless, with the state of the art evolving on a daily basis, it has become a very promising and exciting research field.
