Although silicon photodiodes are widely employed in a host of light-detection technologies, scaling them up is difficult and expensive. Researchers at the Georgia Institute of Technology (Georgia Tech) in the US have now compared the performance of these diodes with that of organic polymer-based diodes, which are easy to fabricate over large areas. Somewhat to their surprise, the researchers found that the organic devices match their inorganic counterparts in all areas apart from one: response time. “The result goes against conventional wisdom that switching to organic materials that can lead to scalable devices would mean giving up on performance,” says team member Bernard Kippelen.
Silicon photodiodes (SiPDs) are very efficient detectors of ultraviolet, visible and near-infrared light. One of the metrics that quantifies their performance is noise equivalent power (NEP), which is defined as the optical power that produces a signal-to-noise ratio (SNR) of one. Since a photodiode’s performance varies with its area and the bandwidth over which measurements take place, researchers also use another parameter, the specific detectivity, to compare the performance of different devices. Higher values of specific detectivity mean that the photodiode can detect fainter levels of light.
Small-area, low-noise SiPDS fare well against these metrics. They boast specific detectivities of around 1012 cm Hz1/2W-1 in the visible and infrared parts of the electromagnetic spectrum when evaluated at a low bandwidth. However, maintaining this performance when the devices are fabricated over larger areas requires stringent control of crystal defects in the photodiode material. This can be difficult to achieve, and team leader Canek Fuentes-Hernandez notes that knowledge of how well SiPDs actually perform can be patchy. “Unfortunately, these metrics are seldom measured, and unverified approximations can lead to large errors when estimating their values,” he tells Physics World.
By directly measuring these key metrics, Fuentes-Hernandez, Kippelen and colleagues found that low-noise, large-area solution-processed flexible organic photodiodes are just as efficient as small-area SiPDs at detecting faint light in the visible range. The organic devices also show electronic noise current values in the range of tens of femtoamperes and noise equivalent power values of a few hundred femtowatts. Both values compare well with silicon when measured at a low bandwidth.
Organic electronic devices
In their work, members of the Georgia Tech team studied P3HT:ICBA organic photodiodes on indium tin oxide/polyethylenimine ethoxylated and MoOx/Ag electrodes. The polyethylenimine electrodes are stable in air and also allowed the researchers to produce photovoltaic devices that exhibit low levels of dark current (the electrical current that flows through a device even when no light shines on it). These low dark currents mean that the material can be used in photodetectors designed to capture faint signals of visible light.
Like other organic polymer-based electronic devices, these photodiodes can be made using simple solution-processing and inkjet printing techniques. That makes it possible to coat them onto a variety of surfaces, including flexible ones like those employed in displays and solar cells. Organic thin films also absorb more efficiently than silicon, so the overall thickness of the active light-absorbing layer in organic photodiodes is very small. Indeed, the active layer of the Georgia Tech team’s photodiodes is just 500 nm thick. “A gram of the material could coat the surface of an office desk,” Fuentes-Hernandez says. “Even if you scale their area up, the overall volume of your detector remains small with organics. If you increase the area of a silicon detector, you have a larger volume of materials that at room temperature will generate a lot of electronic noise.”
2D MXenes make photonic diodes
Direct measurements revealed that a device based on these materials can detect as little as a few hundred thousand photons of visible light every second – equivalent to the magnitude of light that reaches our eye from a single star in a dark sky, explains Fuentes-Hernandez. This sensitivity, combined with their ability to be coated onto large, arbitrarily-shaped substrates, means that organic photodiodes “now offer some clear advantages over state-of-the-art SiPDs in applications requiring response times in the range of tens of microseconds,” he adds.
Expanding the range of applications
According to the team, organic photodiodes could be used in medical applications like pulse oximeters, which use light to measure heart rate and blood oxygen levels. The flexibility of these photodiodes might also allow multiple such devices to be placed on different areas of the body, and the researchers say they could detect a tenth of the light that conventional devices require. This would make it possible to build wearable health monitors that yield better physiological information.
There is just one snag: at 35 microseconds, the organic devices’ response times are significantly longer than those of SiPDS, which typically have response times of picoseconds or nanoseconds. The researchers, who report their work in Science, say they are working to improve response times to expand the range of possible applications for the devices. “The slower response time of our current devices comes from the fact that we use materials that are processed from inks using printing or coating techniques that are not as ordered as crystalline materials,” Kippelen explains. “As a result, the carrier mobility and the velocity of the carriers that can move through these materials are lower, so you can’t get the same fast signals you get with silicon. But for many applications you don’t need picosecond or nanosecond response time.”