Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program

Electrical and Computer Engineering

Year Degree Awarded


Month Degree Awarded


First Advisor

Amir Arbabi

Subject Categories

Electromagnetics and Photonics | Electronic Devices and Semiconductor Manufacturing | Nanotechnology Fabrication | Other Electrical and Computer Engineering


The invention and advancement of optical devices have tremendously changed our life. Devices such as cameras, displays and optical sensors are now an integral part of our lives. Moreover, with the rapid growth in new markets such as virtual reality (VR), augmented reality (AR), autonomous vehicles and internet of things (IoT) the need for optical devices is expected to grow considerably. Recent advances in nano-fabrication techniques have spurred a new wave of interest in optical metasurfaces. Metasurfaces are arrays of wisely selected nano-scattereres that generate desired transformation on the incident light. Metasurfaces provide a new platform for the development of a new class of optical elements which have planar form-factor and can be potentially low-cost. There are many complex interactions in a typical metasurfaces, which adds to the complexity of their design. For structures where modeling such complex interactions between the meta-atoms are not possible, optimization is the only alternative to enhance the performance. Gradient descent is the most efficient method for this type of optimization, since it utilizes the gradients derived using adjoint technique. Adjoint optimization was mostly used in context of topology optimization, however, topology optimization requires high resolution in representation of the meta-atoms topology. Also the optimized devices may not be readily fabricable. Instead of designing free-form structures, we propose and experimentally demonstrate an adjoint optimization method based on parameterized rectangular meta-atom. One intriguing feature that distinguishes optical metasurfaces from conventional optical components is their multifunctional capability. However, multifunctional metasurfaces with efficiencies approaching those of their single-functional counterparts require more degrees of freedom. In chapter III we will describe design of 2.5D metasurfaces and will describe the implementation of parameterized adjoint method for such structures. The adjoint optimization is a powerful method to achieve higher performance for metasurfaces however, it relies on full-wave simulations so the maximum size of device which can be optimized by adjoint optimization technique is limited by the available computational resources. In chapter IV, several techniques will be proposed to increase the performance of larger structure which can not be optimized with adjoint optimization. The methods show reasonable accuracy and high performance. And finally in chapter V, we present systematic approach to explore the large degrees of freedom in meta-atom shapes, and will develop an efficient multi-wavelength, dispersion engineered, and multi-angle metasurfaces that are robust to fabrication errors. The success of the presented method leads to optimal metasurface designs and will enable a planar architecture for high performance optical systems with complex meta-atoms.


Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.