Betavoltaic Energy Harvesting Sensors for Liver Cancer Radiation Therapy Dosimetry
[Thesis]
Richardson, Heather Renee
Amirtharajah, Rajeevan
University of California, Davis
2019
111
Ph.D.
University of California, Davis
2019
As liver cancer incidence rates continue to rise and associated death rates follow a twenty-year trend, increasing by three percent each year, targeted radiation therapy provides a method to combat cancer progression in cases where liver resection or transplant is not possible. Transarterial radioembolization uses microspheres that emit beta radiation to kill the cancer while minimizing exposure to healthy tissue. Dosimetry is the measurement of ionizing radiation dose absorbed by an object, and accurate dosimetry in radiation therapy is crucial for treatment optimization and total dose tracking for consecutive treatments. There are several challenges to obtaining accurate absorbed dose estimates using current dosimetry methods, including the use of imaging methods that rely on photon emission, which is fundamentally limited for a beta emitting radionuclide. Beta particles impinging on a silicon diode produce electron-hole pairs, a process known as the betavoltaic effect, which can be used to create silicon betavoltaic batteries and particle detectors. This research explores mm-scale, betavoltaic energy-harvesting sensors that accumulate dose from beta particles emitted by radiation therapy microspheres to perform dosimetry for liver cancer therapy. A MATLAB model for betavoltaics was developed using Monte Carlo particle simulations and 180nm CMOS process parameters to predict the charge collection efficiency of the betavoltaic cell. For an activity of 2.5 GBq, the model indicates a current generated by the betavoltaics of 124 pA/mm2 and power efficiency of 0.0133%. Standard CMOS processes were used to create the betavoltaic devices using integrated diodes in 180 nm bulk silicon and silicon-on-insulator (SOI) test chips. SOI diodes can be connected in series to achieve a higher open circuit voltage and are inherently less susceptible to radiation damage. Simulations indicate the diodes can achieve an open circuit voltage VOC of 408mV and a current density of 305 µA/mm2 for a 2.5 GBq activity yttrium-90 source, which is the activity for a typical target dose of radiation therapy for liver cancer. A CMOS dosimeter was designed in 180nm CMOS to convert the beta generated current to a digital output using a charge-to-frequency converter. The charge-to-frequency converter integrates an input current to produce an output voltage, compares the voltage to a threshold to generate a pulse which will then inject a charge with the opposite sign of the beta generated current into the input node. A successive approximation register was used to modify the amount of charge that is injected into the input node to increase the input range of the dosimeter. The simulated performance of the CMOS dosimeter for an input range of 1 nA to 100 nA demonstrated an expected output frequency of 0.8 - 35 Hz corresponding to a source activity of 0.1 mCi to 40 mCi.