UC Berkeley

Organic light-emitting diodes (OLEDs) are made from a growing library of man-made semiconductor materials. Unlike inorganic LEDs, OLED materials can be custom-designed and synthesized for desired emission characteristics. The bulk of OLED development to date has been for display and lighting applications, leveraging their tunable color characteristics and high efficiencies. Within display and lighting development, there has been a push to fabricate OLEDs over large areas of flexible substrates compatible with low-cost, high-throughput roll-to-roll printing technology. It is this flexibility and scalable processing that has more recently garnered attention for using OLEDs in wearable optoelectronic biomedical devices. This work focuses on the development of solution-processed, printed OLEDs specifically for biomedical applications.

As a proof of concept that solution processed polymer OLEDs can perform in a wearable biomedical device, green and red OLEDs spin-coated on glass substrates were implmented in an all-organic otoelectronic pulse oximeter to measure heart rate and arterial oxygen saturation. These OLEDs were able to generate a clear photoplethysmogram (PPG), the signal from which heart rate and oxygenation are derived, in conjunction with both inorganic and organic photodiodes. The fully organic opto-electronic sensor was able to accurately measure heart rate and oxygenation within 2% error of a commercially available hospital-grade pulse oximeter.

The next step from a proof of concept device was to develop OLED structures and processing more in line with fabricating flexible and scalable devices. Structurally, most OLEDs employ indium-tin oxide (ITO) as a transparent anode. However, ITO is brittle, and therefore not an ideal material for a flexible device meant to be worn flush with the human body. In this work an ITO-free near-infrared (NIR) OLED was fabricated on a flexible polyethylene napthalate (PEN) substrate. In place of ITO, the anode was thermally evaporated WO3/Ag/WO3. To improve scalability and increase through-put (compared to spin-coating), all polymer layers were printed via blade coating. These OLEDs exhibited a 735 nm emission peak and 2.17% external quantum efficiency (EQE) at 100 mA/cm2.

To maximize OLED scalability and reduce manufacturing costs for inexpensive optical biomedical devices, the entire device structure should be printed and high vacuum evaporation processes should be eliminated. Here, efforts were made to replicate the thermally evaporated WO3/Ag/WO3 anode with blade coated layers. Partially printed WO3/Ag/WO3 anode OLEDs with WO3 layers printed from a WO3 nanoparticle ink and thermally evaporated Ag were achieved. The OLEDs had a red, 617 nm emission peak and exhibited 1-10 mW/cm2 forward flux per pixel area, values compatible with the wavelength and irradiance requirements of optical biomedical devices. Further Ag ink development out of the scope of this work is necessary in order to achieve a fully printed WO3/Ag/WO3 electrode, however a step towards fully printed OLEDs suitable for biomedical applications has been made.

In addition to scalability and flexibility, OLEDs for biomedical applications should have optimzied emission spectra. NIR light has the highest penetration depth in human tissue compared to visible light wavelengths. However, the development of solution-processable organic emitters with emission peaks beyond 800 nm has been slow. Here, a printed ITO-free 745 nm NIR OLED with a secondary emission peak at 815 nm is achieved by increasing the thickness of the emissive layer from 150 to 250 nm, affecting the optical interference within the device to alter the emission spectrum.

To show the importance and impact of deeper NIR emission in an optical biomedical device, PPG signals acquired from a subject's forefinger using a NIR OLED with the secondary 815 nm peak and an NIR OLED with only a 745 nm emission peak were compared. The calibrated strength (magnitude) of the PPG acquired with the NIR-shifted OLED was 32.87% higher than the PPG acquired with the 745 nm OLED. What's more, the better PPG was acquired with less light: the 745 nm OLEDs had a maximum 1 mW/cm2 irradiance on the subject's forefinger, while the NIR-shifted OLED's irradiance was only .33 $mWcm^{-2}$. This underscores the value of developing NIR OLEDs with deeper NIR emission for biomedical applications.

For printed and flexible OLEDs to catch on in biomedicine on a wide scale, their improved utility over currently available devices needs to be demonstrated. Flexible organic optoelectronic pulse oximeter patch prototypes developed and fabricated by Cambridge Display Technology and the Arias Group were shown to physicians from a variety of specialties. The physicians provided feedback on how the organic optoelectronic oximeters could aide their practice in ambulatory emergency care, chronic respiratory disease management, intraoperative monitoring, and pediatrics. Strategies for where and how to best interface the sensors with the human body were also developed from the physicians' input. PPG signals were acquired from flexible organic optoelectronic oximeters secured to the abdomen and back, both novel locations for pulse oximetry.