My current research is part of an ARPA-e OPEN-2021 project to develop a suite of sensors to characterize nitrous oxide emissions from agricultural land. This work builds on the soil nitrate sensors developed in PANDAS, and expands to a greater variety of sensors. The lead PI is professor Whendee Silver, a biogeochemist in the Environmental Science, Policy, and Management department at UC Berkeley.

Nitrous oxide (N2O) is a potent greenhouse gas which is difficult to measure, and which is known to have high spatial and temporal variation. N2O emissions are correlated to measurable soil conditions which can be used in emissions models. This project aims to create a suite of sensors to measure soil variables--such as moisture, temperature, pH, oxygen, and several ion concentrations-- which are correlated with N2O emission, then use the generated data to understand the relationships between these soil properties and N2O emission.

I'm working on several types of sensors for this project: improving stability over time and sensor-to-sensor variability of nitrate sensors, integrating printed temperature sensors with chemical sensors, designing waterproof packaging for field trials, and developing electrochemical O2 gas sensors. 

Another current project, funded by the California Department of Food and Agriculture, and in collaboration with Isaya Kisekka's group at UC Davis, seeks to monitor nitrate levels in groundwater around tomato fields and almond orchards in California's central valley. Drinking water from wells in the central valley commonly contains levels of nitrate which exceed public health standards. Monitoring how nitrate flows from fertilizer application to runoff or leaching will provide information to help agricultural communities use fertilizer more efficiently and protect their drinking water. 

For this project, we are improving the stability and lifetime of the nitrate sensors, and preparing them to be installed at several depths under the soil. 

The heart of my PhD work was part of the Precision Agriculture using Networks of Degradable Analytical Sensors (PANDAS) project, an ARPA-e OPEN-2018 project to develop wireless, battery-free, biodegradable soil moisture and soil nitrate sensors. Lead by Dr. Greg Whiting at CU Boulder, the project was a collaboration between CU Boulder, UC Berkeley, and Kansas State.

We sought to create biodegradable nitrate sensor nodes, which imposed specific design constraints. To achieve biodegradability, we kept the nodes as simple as possible, omitted onboard batteries or solar panels, and relied on passive RFID for communication and energy harvesting.

I worked on all parts of the node: passive wireless power and data communication using UHF-RFID; biodegradable materials processing; and soil nitrate sensing. With knowledge of each of these components, I turned to system integration and field deployment considerations, testing the nitrate sensors in soil and measuring RFID communication in and around soil and crops. 

In my Masters project, I studied how different solution processing techniques influence the functionality of flexible antennas in the 915MHz range. I learned about inkjet, stencil, spray, and screen printing, as well as RF antenna design, simulation, and characterization. The antennas I was working with were envisioned as part of the "human intranet," a network of wirelessly connected, unobtrusive,  wearable sensors for health monitoring or medical applications. In addition to the printing itself, I studied methods for connecting conventional rigid silicon electronics to the flexible printed antennas such that the connection was mechanically sound and electrically well matched at the frequency of interest. 

I later applied what I learned about antennas operating around human bodies to the problem of passive RFID wireless communication around wet soil.

My undergraduate research, in Dr. Wataru Nakagawa's lab at Montana State University, focused on the fabrication of optical nanostructures to control the polarization of NIR light. I used conventional silicon micro/nano fabrication techniques like resist spinning, electron-beam lithography, metal evaporation, wet chemical etching, and reactive-ion etching to create structures such as a reflective quarter waveplate and a filter for both wavelength and polarization. I also used AFM and SEM to inspect my structures and optimize fabrication processes. 

One application of these filters was to monitor the thermodynamic phase of clouds (ice crystals vs water droplets) from ground-based stations using scattered sunlight. Cloud thermodynamic phase information is important data for climate models, but is challenging to measure passively from the ground. My senior capstone project involved creating a prototype cloud polarimeter from commercially available components.

This project inspired me to learn more about sensors for environmental monitoring. I loved being able to test our prototype outside--actually taking cloud data. I was inspired by the potential impacts on understanding climate that my sensors could offer. 

In order to make biodegradable sensors, biodegradable conductors are needed. Many types of biodegradable conductors exists, but many are significantly less conductive than conventional metals. Antennas for wireless power and data transmission need high conductivity for good efficiency. Printed zinc is a highly conductive, degradable option. 

Previous studies have shown that treating  printed zinc traces with acetic acid removes the native oxide on discrete zinc microparticles and creates a conductive trace. In this work we modified ink composition and optimized the treatment process to increase conductivity and improve stability over time. We used the degradable conductor to make RFID antennas with similar performance to our previously published screen printed silver antennas. 

C. Baumbauer, A. Gopalakrishnan, M. Atreya, G. Whiting, A. C. Arias, Advanced Electronic Materials, 2024

[paper]

Measuring nitrate in soil is very difficult today, though high spatial and temporal resolution data would help increase efficiency of fertilizer use and reduce harmful downstream consequences of over-fertilization. Printed potentiometric nitrate sensors offer a promising option for distributed, in-soil nitrate measurement because they are produceable, have no moving parts, are robust, and are simple to read out. 

Here, we characterize both the ion-selective electrode and the printed reference electrode in varying concentrations of nitrate and in solutions that contain relevant concentrations of interfering ions found in soil. We also demonstrate that their sensitivity to nitrate in soil is equivalent to their sensitivity to nitrate in aqueous solution.  

C. Baumbauer, P. Goodrich, M. Payne, T. Anthony, C. Beckstoffer, A. Toor, W. Silver, A. C. Arias, Sensors, 2022

[paper]

Passive RFID tags allow the identification information of an object to be wirelessly tracked by a central reader, without on-board power on the object. Passive RFID tags with integrated sensors provide additional information about the condition of the objects.  Such technology could be very useful in shipping and warehouse management. The size and rigid form factor of the associated antennas limits their practical deployment today. 

Here, we demonstrate printed, flexible RFID antennas with compact geometry. We characterize how printing technique impacts antenna behavior, and offer three compact antenna designs. Finally, printed antennas and printed sensors are integrated with a commercial chip to form a flexible RFID sensor tag. 

C. Baumbauer, M. Anderson, J. Ting, Akshay Sreekumar, J. Rabaey, A.C. Arias, and A. Thielens Scientific Reports, 2020 10, 16543.

[paper]

Printed interdigitated capacitive humidity sensors are deceptively simple. They rely on the fact that most substrate materials absorb water in humid environments, which increases the relative permittivity of the material, which can be read with a capacitor. Their simplicity means many reports of printed interdigitated humidity sensors exist in the literature, but currently, it is challenging to compare the performance of humidity sensors printed on different substrates because of a lack of standard sensor characterization and fabrication. 

In this paper, we directly compare the performance of a range of printed sensors that are fabricated and tested via a standard method. The relationship between humidity and capacitance is not a linear one, so reports of linear sensitivity alone do not adequately represent sensor behavior. Moreover, response time, temperature, and prior exposure to high- or low-humidity conditions all impact sensor behavior. 

E. Wawrzynek, C. Baumbauer, and A. C. Arias Sensors, 2021 21, 19.

[paper]