A miniature near-infrared sensor that can be plugged into a smartphone can analyze the chemical composition of the materials of milk and plastic.
A TU/e analysis team has developed a brand new near-infrared sensor that is easy to fabricate, measures comparable to a smartphone sensor, and is prepared for rapid use in the process. industrial and agricultural supervision. The breakthrough has just been revealed in the journal Nature Communications.
The human eye is an amazing sensor. By using three different types of light-sensing cones, which convert visible light into indicators for different colors, attention provides important details about the world around us. we.
“When our mind sets the indicators in a general way, it makes predictions about what the indicators imply largely based on our experience. For example, a pink strawberry is beautiful, but an inexperienced strawberry is not,” says Kaylee Hakkel, a doctoral researcher in the Photonics and Semiconductor Nanophysics group in the Department of Utilization Physics and is first co-author of the survey, explains.
Mantis shrimp.
Measurements don’t matter
While the human eye is impressive, it is not considered the most advanced pure light sensor on the market. “The eyes of the Mantis shrimp have 16 different types of cells that are sensitive to mild ultraviolet, visible and near-infrared (NIR) rays,” Hakkel said. “And spectroscopy inside the infrared is the most attractive thing for commercial and agricultural functions, but there is one key difficulty – current near-infrared spectrometers are simply too large and costly.”
Hakkel and her colleagues solved this problem by creating a near-infrared sensor that fits on a small chip. And similar to the Mantis shrimp’s attention, it has 16 completely different sensors – yet all of them are sophisticated in the near-infrared region. “The miniaturization of sensors while preserving the low cost is a serious problem. So we designed a whole new wafer making course to achieve this. “
“The cost is low because we will be manufacturing several sensors by the same time and it is already ready for use in reasonable real-world functions,” Hakkel offers. “The sensor chip is small and will even be embedded in smartphones in the future.”
Fix the spectral sensor
Andrea Fiore, head of analysis from the Department of Physics of Use and the Eindhoven Hendrik Casimir Institute, is delighted with the work of their analytical team. “We have been investigating this expertise for many years. And now we have efficiently integrated spectral sensors on a single chip, while also dealing with another important difficulty – environmentally friendly use of information.”
Normally, when the sensor measures a light level, the resulting signature is used to reconstruct the spectrum – or optical fingerprint – of the fabric. Sensor algorithms are then used to investigate the information.
On this new method, the researchers demonstrate that the spectral reconstruction step is unnecessary. According to different phrases, the metrics generated by the sensor are sent directly to the evaluation algorithms. “This greatly simplifies the design needs for the system,” Fiore notes.
Milk and plastic analysis
With the sensor in hand, the researchers then tested the sensor in multiple tests, as defined by co-author Maurangelo Petruzzella, who may be working for the startup MantiSpectra. “We used the sensor to measure the dietary characteristics of a variety of sources along with milk. Our sensor offers comparable accuracy within the predictive range of milk fat content materials such as a typical spectrophotometer. We then use the sensor to classify many different types of plastic.”
The dietary properties of milk determine its financial value, and the sensor has been confirmed to be able to accurately measure these properties. In addition, these measurements can most likely be used to observe the overall condition of the cow. Sorting plastics using sensors can assist in optimizing waste sorting processes.
“In addition to these functions, we predict that the sensor could very well be used for personalized healthcare, precision agriculture (e.g. monitoring fruit and vegetable ripeness), Process management and testing in the on-chip lab. Now we have a fully innovative device built primarily on this expertise, the SpectraPod™, which corporations and institutions are using to build their functions. And the nice factor is that this sensor could even be ubiquitous in smartphones in the long run, which means people could use it at their residence to verify their meal standards or verify aspects of their health,” provides Petruzzella.
The future of MantiSpectra
And matters are simply starting to get thrilling for Hakkel, who defended his thesis on January 14 at TU/e. She will then become a member of Petruzzella in the startup MantiSpectra, where they will try to improve the sensor for more streamlined functions. “I am really excited to begin my involvement in the next part of sensor innovation with MantiSpectra. This sensor can contribute to cleaner setups and meal waste disposal, functions that can be very important to everyone. “
Reference: “Integrated Near Infrared Spectroscopy Sensor” by Kaylee D. Hakkel, Maurangelo Petruzzella, Fang Ou, Anne van Klinken, Francesco Pagliano, Tianran Liu, Rene PJ van Veldhoven and Andrea Fiore, 10 January 2022 , Nature Communications.
DOI: 10.1038 / s41467-021-27662-1
