Our research can be best described as device driven materials innovation–developing new material growth and fabrication techniques and then turning those into functional electronic and photonic devices and circuits. We focus on a broad range of areas, such as energy harvesting, sensing, and computing.

Due to the inherently interdisciplinary nature, our research projects often contain a mix of materials growth, device fabrication, materials and device characterization, and simulation, requiring our group members to have both a broad understanding of each area, as well as significant expertise in a single area.



As the devices we make become increasingly more complex, the fundamental roadblocks to realizing our designs often are the material growth and fabrication techniques available. This thrust focuses on developing low-cost, scalable growth techniques to enable novel material systems and geometries for photonic and electronic devices.


Polycrystalline Materials for High Performance Energy and Sensing Devices


Above figure shows the process of growing ultralarge grain polycrystalline InP on a molybdenum substrate with high optoelectronic quality as shown in the Suns vs. ΔEf curves. Additionally, control of nuclei spacing above one mm and precise positioning is possible with this technique.


The thin-film vapor-liquid-solid (TF-VLS) technique is used here to make an InP thin film on a molybdenum foil substrate. The growth is done via first depositing a template layer of indium and silicon oxide on the desired substrate followed by heating in a phosphorous ambient. The phosphorous supersaturates the liquid indium, causing precipitation of InP while consuming the starting indium. Eventually, the deposited indium template will entirely convert to InP.

Significance :

Traditionally, growth optoelectronic quality III-V films have required a lattice matched substrate, a constraint that causes significant processing and cost challenges. TF-VLS enables high-quality III-V thin-films to be grown on a variety of substrates (silicon, metal foils, etc). This greatly opens up the potential applications for which III-V’s are viable. Furthermore, this technique is general, and can be used for a variety of compound semiconductors.

TF-VLS growth presents a methods of growing large-grain polycrystalling or single crystalline compound semiconductor thin films without the need to consider the epitaxial relationship with the substrate. Furthermore, it allows a variety of device geometries that would be challenging or impossible to fabricate with conventional vapor phase growth methods.

Key Publications:

R. Kapadia, Z. Yu, et al. “A direct thin-film path towards low-cost large-area III-V photovoltaics”, Scientific Reports, 3, 2275; DOI:10.1038/srep02275, 2013.

R. Kapadia, et al. “Deterministic nucleation of InP on metal foils with the thin-film vapor-liquid-solid growth mode”, Chemistry of Materials, 26 (3), 1340–1344, 2014.


Single Crystalline Materials for Electronic and Photonic Circuits

tf_vls-single-crystalline-materialsAbove figure shows the process of growing single crystalline InP on a silicon wafer. The scanning electron microscope images (bottom left) shows the growth occurs only where desired, while the electron backscatter diffraction (EBSD) image (bottom middle) shows that each InP mesa is a single color and hence a single crystalline orientation. The TEM demonstrates that even though the substrate is not lattice matched, clean lattice planes are visible at the interface.


By following a procedure similar to the TF-VLS approach for polycrystalline materials, but restricting the size of the group III template, it is possible to make high-quality single crystalline III-V’s on non-epitaxial substrates.


This technique enables direct growth of compound semiconductors on silicon, potentially circumventing the need for complex hetero-epitaxial growth methods. Allowing us to take advantage of the excellent optoelectronic and electronic properties of III-V’s while also leveraging the well established silicon processing infrastructure. This approach offers a potentially low-cost and scalable route towards high performance III-V optoelectronic and electronic devices to be integrated with silicon.

Key Publications:

K. Chen, R. Kapadia, et al. “Direct growth of single crystalline III-V semiconductors on amorphous substrates”, Nat. Commun., 7,1502, 2016.




Our focus in the device thrust is on demonstrating high performance devices from the scalable and low-cost materials approaches developed in our materials thrust.


High-performance devices grown directly on silicon


Above figure shows single crystal InP transistors and phototransistors grown directly on silicon wafers. Importantly, the device mobility of the III-V MOSFET is >700 cm without an optimized gate dielectric, and exhibits Ion of >120 mA/mm with a gate length of 3 microns and VDD = 2 V. The phototransistor shows ultrahigh responsivities between 20 and 700 A/W and specific detectivities close to 1E12 Jones. 


Thin (<200 nm) single crystalline InP channels were grown directly on a silicon wafer. These devices were then fabricated using a Ge/Au/Ni alloyed source/drain and zirconium oxide gate dielectric. Despite the unoptimized gate dielectric, high currents, rivaling long channel InP bulk MOSFETs were observed.


These demonstrations are strong evidence that the approaches developed here have the potential to create high-performance circuits without lattice matched substrates. This would simultaneously reduce the cost, and increase the scope of applications where these could be deployed.

Key Publications:

K. Chen, R. Kapadia, et al. “Direct growth of single crystalline III-V semiconductors on amorphous substrates”, Nat. Commun., 7,1502, 2016.

H. Ko, K. Takei, R. Kapadia, et al. “Ultrathin compound semiconductor on insulator layers for high performance nanoscale transistors”, Nature, 468, 286–289, 2010.


Integrated Photonics for Electron Emission

In this thrust, we explore the opportunities for ultra-fast and highly efficient control over electron emission using integrated photonics.


Above figure schematically demonstrates the vision for using optical cavities to tune electron emission. The combination of precise control over photon absorption with an understanding of electron scattering potentially enables new ultra-fast and efficient class of electron emitters.