Optical spectrometry is a powerful, non-destructive, high throughput technique, that can measure the quality and quantity of chemical composition. Commercially available spectrometers are expensive, bulky and are limited to either visible or infrared light. In the Spectrodot project, we have built a compact spectrometer prototype that covers the visible and SWIR range based on a low-cost graphene/quantum-dot detector.
The Spectrodot device is a portable, stand-alone unit with a built-in light source that enables sample characterization in reflective mode. The wavelength range in which is it is sensitive is from 300-1900 nm, and it has a spectral resolution of 2 nm. We developed a dedicated software that allows data read-out, but also the implementation of data interpretation algorithms.
For the Spectrodot device, we have developed two application showcases:
• Allergens: Nuts are for a large number of people with an allergy a matter of life and death. The Spectrodot device can detect if milk is nut-based or not. We collected a dataset and trained the algorithm to distinguish 3 different kinds of milk (almond, cow and rice) that all appear white to the human eye.
• Ingredients: Coffee is one of the most consumed beverages. There are two families of beans: Robusta and Arabica. We developed a dataset and algorithm to sense which of the type of beans a particular coffee ground was made of.
Further miniaturization of spectrometers will increase the size and reach of the current spectrometer markets and enable penetration into the commercial consumer space. Near-future advancements of the Spectrodot device will demonstrate a form factor that will allow its use in a mobile handset. We envision that coupling the development of this miniaturized spectrometer technology platform with suitable use cases will be a catalyst for the broad adoption of spectroscopy in everyday life.
