Professor Maria Merlyne De Souza

EEE Staff Photo Prof M De Souza

Contact Details

m.desouza@sheffield.ac.uk

Tel: +44 (0)114 22 25167
Mob: 07540 668 490

ORCID: 0000-0002-7804-7154

Qualifications

  • PhD (Engineering), University of Cambridge 1994
  • BEng (Electronics & Communications Engineering), Indian Institute of Science Bangalore 1988
  • BSc (Physics & Mathematics), University of Bombay 1985

Research Activities

  • CMOS in GaN and on-chip inductors for Power Management Integrated Circuits.
  • ZnO/Ta2O5 synaptic thin film transistors for logic-in-memory and neuromorphic applications.
  • RF Power Amplifiers in GaN for high linearity and efficiency.
  • Perovskite solar cells.

Responsibilities

  • Chair in Microelectronics, EEE Department, University of Sheffield (2007-)
Bio

I graduated with a B.Sc in Physics and Mathematics (1985) from the University of Mumbai, a B.E. in Electronics and Communications Engineering (1988) from the Indian Institute of Science, Bangalore and a PhD from the University of Cambridge (1994). I joined as a Junior Research fellow in ‘95, was promoted to a Senior Research fellow in ‘98 and was appointed Professor in Electronics and Materials at the Emerging Technologies Research Centre, De Montfort University in 2003. I joined the EEE department at Sheffield as Professor of Microelectronics in 2007. I work in multi-disciplinary research focused on the physics of devices, materials and their microelectronic applications in computing, communications and energy conversion.

Until now, microelectronics has relied on the versatility of silicon CMOS to deliver enhancement in performance by scaling the MOS transistor. I have worked on various aspects of CMOS such as reliability, high-k gate oxides and Indium for retrograde channels, first introduced in production at the 65 nm node. However, scaling (as we know it) is now nearing an end and alternate materials and device architectures are required for future semiconductor applications. Supervised learning for image and speech recognition, autonomous driving and medical diagnosis in Artificial Intelligence (AI) presently rely on CMOS based deep neural networks. These are inherently power-hungry due to a continuous exchange of information between the required large volume of memory and processing units. It is expected that such Von Neumann architectures will be replaced by neuromorphic systems that are more akin to a biological brain. Our team has recently demonstrated ZnO/Ta2O5 solid electrolyte thin film transistors with synaptic capabilities. I am interested in exploring such memristive devices in neuromorphic applications, electrochemical storage and flexible electronics for health.

My interest in more efficient semiconductors, smart materials and systems that leave a smaller footprint on the environment, spans to GaN for power and RF applications, that I have previously explored in equivalent silicon- based device technologies such as the IGBT and the RF LDMOSFET. These are driven by the automotive, aerospace, space, renewables, telecoms and consumer/industrial electronics sectors. Our recent work includes a new class of harmonic RF power amplifiers with record efficiency and output power prototyped using commercial GaN devices. We are also working towards a p-type MOSHFET and magnetic thin films for “CMOS in GaN” in power management integrated circuits and current sensors.

Research Projects

...

Completed Projects

...

Research Students

...