Detection of micro/nano-particle concentration using modulated light-emitting diode white light source

Wenxiu Zhao, Haibo Yu, Yangdong Wen, Feifei Wang, Yang Yang, Zhu Liu, Lianqing Liu, Wen J. LiSensors and Actuators A Physical 2018 · DOI

TL;DR

This paper presents a novel method for measuring micro/nano-particle concentrations in solutions by integrating a visible light communication (VLC) system with a microscope. The system leverages the change in light intensity (absorbance) and bit error rate (BER) of the VLC signal as it passes through the particle-laden microfluidic chip, offering a potentially cost-effective and rapid alternative to existing methods.

What Problem Does This Paper Address?

Accurately determining the concentration of micro and nanoparticles in various solutions is crucial for many scientific and technological applications, from drug delivery and diagnostics to materials science. However, existing methods like hemocytometers, optical microscopes, and even flow cytometers have limitations in terms of accuracy, sample volume, or cost. Spectrometers, while more accurate, require significant sample volumes and complex optical setups. This paper addresses the need for a more convenient, rapid, and potentially lower-cost method for real-time particle concentration measurement, particularly for particles in the size range relevant to visible light wavelengths.

Methods

The core of the proposed method is a Visible Light Communication (VLC) system built around a white light-emitting diode (LED) and a microscope. The VLC system operates using On-Off Keying (OOK) modulation at a data rate of 156.25 kbps. The setup involves:

  • Microfluidic Chip: A custom-designed microfluidic chip (PDMS and glass) with microchannels of varying heights (130-340 µm) and widths (300 µm) to hold the particle solutions.
  • Light Source: A white LED (LUXEON Rebel ES) modulated by an arbitrary waveform generator.
  • Optical Path: An aspheric lens and a 20x objective lens condense the LED light, focusing it through the microfluidic chip.
  • Detection: A PIN sensor (OSRAM BPX 61) captures the transmitted light, which is then processed by an oscilloscope and sent to a computer for demodulation.
  • Measurement Parameters:
    • Absorbance: Calculated from the ratio of incident light intensity (when the sample is deionized water) to the transmitted light intensity through the particle solution. This is based on the Beer-Lambert Law.
  • Bit Error Rate (BER): The ratio of error bits to total transmitted bits in the VLC system. Changes in particle concentration affect the signal-to-noise ratio, thus influencing the BER.

The paper also utilizes Mie scattering theory to simulate the interaction of light with spherical particles, providing a theoretical basis for understanding the observed absorbance changes.

Datasets

The paper does not explicitly mention training or evaluation datasets in the traditional machine learning sense. Instead, it uses:

  • Polystyrene micro/nano-particles: Various sizes (48 nm to 19.2 µm) and concentrations were prepared and tested.
  • Deionized water: Used as a reference for measuring absorbance.

The experiments involved injecting these particles into the microfluidic chip at different concentrations and measuring the resulting absorbance and BER.

Key Insights and Ideas

  • Dual Functionality of VLC: The white LED in the VLC system acts as both an illumination source for the particles and a detector for their concentration, demonstrating a novel application of VLC technology beyond communication.
  • Absorbance as a Concentration Indicator: The study confirms a generally linear relationship between particle concentration and absorbance, especially at lower concentrations and for smaller particle sizes, aligning with Mie scattering theory. This provides a direct optical method for concentration estimation.
  • BER as a Complementary Indicator: The bit error rate (BER) of the VLC signal is shown to be sensitive to particle concentration. Higher concentrations lead to increased BER, offering a second, independent parameter for concentration measurement.
  • Particle Size Matters: The interaction of light with particles is highly dependent on their size relative to the wavelength of light. Larger particles (e.g., 600 nm and 1 µm) showed more significant effects on both absorbance and BER, likely due to their size being closer to the peak wavelengths of the white LED's spectrum.
  • Potential for High Resolution: While current BER resolution is limited, the authors suggest that with advancements in VLC data rates (e.g., to Gb/s), BER could offer even finer resolution for particle concentration detection.

Key Results

  • Absorbance Linearity: For micro/nano-particles of various sizes (48 nm to 19.2 µm), the absorbance showed an approximately linear relationship with concentration when tested in the microfluidic chip. This trend was validated by Mie scattering simulations.
  • Microchannel Height Influence: The absorbance measurements were consistent across different microchannel heights (130-340 µm), indicating that the optical path length within this range did not significantly alter the linear trend, though it did affect the absolute absorbance values.
  • BER Correlation: Higher particle concentrations consistently resulted in higher BER values, demonstrating the utility of BER as a concentration indicator.
  • Simulation vs. Experiment: Mie scattering simulations provided a good approximation of the experimental trends for absorbance versus concentration and particle diameter, though some discrepancies were noted, likely due to simplifications in the simulation (e.g., perfect spheres, no multiple scattering) and experimental uncertainties.

Limitations and Future Work

  • Simulation Simplifications: The Mie scattering simulations assumed perfect spheres and uniform distribution, which may not fully represent real-world particle behavior. The complexity of calculating extinction coefficients also led to approximations.
  • Experimental Uncertainties: Factors like surface roughness of the microchannel, inevitable errors in optical path length, and the distribution of particles after drying (observed in SEM images) can affect experimental accuracy.
  • BER Resolution: While promising, the current BER-based method has lower resolution compared to the absorbance method.
  • Future Work:
    • Higher Data Rate VLC: The authors emphasize that future VLC systems with much higher data rates (Gb/s) could significantly improve the resolution of BER-based concentration detection.
  • Synchronous Measurement: Combining both absorbance and BER measurements could lead to a more reliable and robust particle concentration detection system.
  • Broader Particle Types: Extending the method to a wider range of particle materials and shapes.
  • Real-time Applications: Further development for practical, real-time monitoring in various industrial and scientific contexts.

Summary updated

BibTeX
@inproceedings{zhao2018detection,
  author = {Wenxiu Zhao and Haibo Yu and Yangdong Wen and Feifei Wang and Yang Yang and Zhu Liu and Lianqing Liu and Wen J. Li},
  title = {{Detection of micro/nano-particle concentration using modulated light-emitting diode white light source}},
  booktitle = {Sensors and Actuators A Physical},
  year = {2018},
  doi = {10.1016/j.sna.2018.10.044},
}
Figure