Date of Award




Document Type

Master's Thesis

Degree Name

Master of Science (MS)


Department of Chemistry

Content Description

1 online resource (iv, 33 pages) : color illustrations.

Dissertation/Thesis Chair

Mehmet Yigit

Committee Members

Alan Chen, Maksim Royzen


biomimetic sensor, biosensor, fluorescence, machine learning, nanoparticle, nucleic acid, Milk, Biosensors, Biomimetic materials, Nanobiotechnology, Fluorescence spectroscopy

Subject Categories

Analytical Chemistry | Chemistry | Nanoscience and Nanotechnology


Herein, we developed a novel artificial tongue using machine learning and 12 nanoassemblies (2D-NAs) to identify and analyzed different kinds of milk beverages for quality control. This biomimetic sensor array was trained to “taste” different milk types as an “artificial tongue” which is the first time we demonstrated that this sensor array can be implemented to complex systems. Two-dimensional nanoparticles (2D-nps) and nine fluorescently labeled single stranded oligonucleotides (ssDNA) with different length and nucleobases were assembled to create 12 2D-NAs. The artificial tongue was deployed to identify and analyze five milk types. All five milk types were discriminated with 95% confidence level. The sensor array was able to analyze different milk types for quality. Upon addition of different beverages, fluorescently labeled ssDNA that was quenched on 2D nanoparticles would desorb and generate a fluorescent recovery. Fluorescence recovery of each sensor was calculated by the equation, Δƒ=ƒ-ƒ0, to generate a fluorescence fingerprint for each milk type. For this study, there were no known interactions between the 2D-nps, ssDNA and the 5 milk beverages. Thus, this is a non-specific sensor array that utilized machine learning to process the data. Later, the fluorescence data collected was analyzed in partial least square discriminant analysis (PLSDA) and each milk type was successfully differentiated. This new approach of food quality control, complex system identification and chemical analysis, in addition to the rapid advancement of the artificial intelligence, shows great potential for broader application in the future.