Use of Machine Learning with Temporal Photoluminescence Signals from CdTe Quantum Dots for Temperature Measurement in...

Area of Science:

• Nanotechnology and Materials Science
• Biomedical Engineering
• Data Science

Background:

• Accurate temperature monitoring is crucial for biological processes in microfluidic devices.
• Existing quantum dot photoluminescence (PL) temperature sensing in microfluidics has limitations in accuracy (around 1 K over tens of degrees).

Purpose of the Study:

• To develop an advanced machine learning algorithm for improved temperature sensing in microfluidic devices using quantum dot (QD) photoluminescence.
• To enhance the accuracy and reliability of temperature measurements within microfluidic systems.

Main Methods:

• A seven-layer fully-connected neural network was designed and trained.
• The algorithm processed normalized spectral and time-resolved photoluminescence data from CdTe quantum dots.
• Data was collected across two temperature ranges: 10 K to 300 K and 298 K to 319 K.

Main Results:

• The machine learning model achieved high accuracy in temperature reconstruction.
• In the high-temperature regime (298 K to 319 K), accuracy was 0.1 K.
• In the low-temperature regime (10 K to 300 K), accuracy was 7.7 K, improving to 0.4 K above 100 K.

Conclusions:

• The developed machine learning approach significantly enhances temperature sensing accuracy in microfluidic devices.
• This method shows potential for precise temperature monitoring in microfluidic and nanofluidic applications, especially with stable normalized PL data.
• It offers a robust data analysis strategy for advanced microfluidic temperature sensing.