Microplastic Detection Using Darkfield Color-Spectral Imaging (DCSI)
A rapid, optical method to detect microplastics ranging in size from 2-micron to millimeters without Nile Red or Fluorescence.
ColSpec™ A Unique Optical Approach To Microplastic Detection
A non-destructive test: Detects and reveals microplastics as small as 2-micron.
Detects transparent or colored plastics: Visualize transparent or colored microplastics in high contrast*.
Reveals microplastics: A 40,000 lumen light source in critical-angle darkfield light illumination “CADFLI”. Visit the image gallery.
Eliminates staining and fluorescence: A non-destructive, budget friendly, instrumental alternative to nile red staining or spectroscopy.
Fast acquisition time: One field of view typically takes less than 100-milliseconds.
Minimal sample preparation: In many cases, no sample prep is required.
High sample throughput: Average 50-80 per hour manually for a series of similar samples.
Morphological characteristics of plastic fragments: High-resolution imaging reveals plastic “shreds” or “chips”. See the gallery images for examples of each.
Detects cosmetic “glitter”: Cosmetic glitter often made with PET contaminates oceans, waterways, and landfills.
ColSpec hardware: Includes: high numerical-aperture, long working distance 12:1 zoom magnification with critical angle illumination, and xyz sample translation,
ColSpec software: Written in Python, provides classification and recognition in color-spectral space, object sizing, areas occupied, and abundance. See here.
How ColSpec™ with Critical Angle Darkfield Illumination (CADFLI) works
Microplastics: are typically plastic particles less than 5-mm in size. However, microplastic particles in the 1-micron to 1-mm range are of increasing environmental interest but very challenging to detect.
ColSpec overview: ColSpec is a portable optical instrument that non-destructively detects microplastic candidates (1-2-micron or greater) with minimal sample preparation.
- Designed for both the laboratory and the field, ColSpec reduces the number of samples that require time-consuming, expensive, analytical polymer identification.
- The secret is in the illumination: : ColSpec works with very high-intensity (40,000 lumen) illumination striking objects at angles greater than 75-degrees.
In this configuration, microplastics scatter light to reveal themselves as bright, high contrast, structurally detailed objects on a black or dark background. In contrast, weakly scattering non-plastic materials contribute to an overall dark background.
We refer to this reflection illumination technique as “Critical Angle Darkfield Illumination” (CADFLI.)
Color Spectral Imaging: Whereas, Hyperspectral Imaging uses all wavelengths in a given wavelength range, color-spectral imaging uses all colors defined in HSV color-space. Therefore, when coupled with CADFLI darkfield illumination, color-spectral Imaging becomes Darkfield Color Spectral Imaging (DCSI).
HSV stands for Hue, Saturation, and Value, where “Value = Brightness.” This format includes the critical brightness component missing in RGB.
Technical overview of HSV color-space: Quantified color space was formalized in 1931 by the “Commission Internationale de l’éclairage” (CIE), known in English as the International Commission on Illumination. Wikipedia describes CIE color space like this:
“The CIE 1931 color spaces are the first defined quantitative links between distributions of wavelengths in the electromagnetic visible spectrum, and physiologically perceived colors in human color vision…” (https://en.wikipedia.org/wiki/CIE_1931_color_space).
A simple internet search provides comprehensive technical information describing quantitative color measurement.
Darkfield Color Spectral Imaging DCSI-HSV: HSV color values measured at normal incidence (0-degrees or straight-on) differ in value depending on the illumination angle.
At non-zero angles of incidence, reflectivity and surface texture impact HSV values and often reveals structural detail not evident at normal incidence (0 degrees). ColSpec utilizes this information to deliver highly sensitive microplastic object classification and characterization.
ColSpec software creates DCSI-HSV libraries: ColSpec training uses supervised learning based on the HSV values of either known MP objects or microplastics candidates. Each HSV reference library signature becomes a “class” that correlates with a particular object or plastic material.
Plastics detected by ColSpec: ColSpec detects PET (1), HDPE (2), PVC (3), LDPE (4), PP (5) PS (6), and others (7), where the numbers in parentheses are recycling codes. Note: ColSpec detects but does not identify the polymers present in microplastic particles.
Sample mounting: Most commonly, up to 47-mm filter membranes are adhered to microscope slides. ColSpec allows X, Y, and Z sample translation with reference scales.
Speed of analysis: Typically 100-msec per FOV with a sample throughput between 50 to 80 samples per hour. Each FOV can contain an unlimited number of particles.
Basic software features: ColSpec software classifies particles and determines their abundance, area, and size distribution. The software also detects objects that may be of interest that are not in a library.
Applications that use color space: Color measurement technology quantifies, verifies, and confirms the colors present in forensics samples, coatings, paints, pharmaceuticals, medical diagnostics, remote sensing, and agriculture. Paint color swatches and paint color matching illustrates an everyday use of color measurement technology.
Related technology – darkfield microscopy: Darkfield imaging in transmission microscopy was first formally reviewed over one hundred years ago in the “Transactions of the American Microscopical Society Vol. 39, No. 2, Apr. 1920” For a detailed description of darkfield microscopy, see https://micro.magnet.fsu.edu/primer/techniques/darkfield.html hosted by Florida State University.
ColSpec applies the same principles of microscope-based darkfield transmission into darkfield reflection.
Pricing and usage: ColSpec is a portable, modular system right-priced for tight budgets.
Bottom line: CADFLI measures color spectra in HSV color-space. Illuminating the FOV at grazing angles reveals bright, high-contrast objects such as microplastics in a dark background.
How The ColSpec™ Microplastic Detection System Works With Critical Angle Darkfield Illumination (“CADFLI”)
Rain water run off after filtering
Mask showing the location of candidate microparticles
Histogram showing the distribution of candidate microparticles in “equivalent diameter”
Histogram showing the distribution of candidate microparticles area occupied by particle size
MicroPlastic detection: Sub-millimeter blue polypropylene (PET) bottle cap fragments. 280-micron glitter added as a size reference
MicroPlastic Detection Image Galleries
All the materials used below were ablated from domestic plastic items such as screw caps and plastic food containers.
Ablated material was sieved to pass fragments less than 1-mm in size.
MicroPlastic fragments included both colored and transparent materials.
Glitter produced with PET is ubiquitous in the environment stemming from extensive use in cosmetics and advertising. Particles average 280-micron point-to-point in a hexagonal format.
Glitter was added to each microplastic sample to provide an internal size reference.
If you are an Institutional researcher ask for a complimentary sample kit of the microplastics shown in the galleries. Contact LightForm for details.
Gallery microplastics shown below include: PET/PETE: Polyethylene terephthalate; HDPE: High-density polyethylene terephthalate; LDPE: Low-density polyethylene terephthalate; PVC: Polyvinyl chloride; PP: Polypropylene; PS: Polystyrene; PA/PC: Polycarbonate/polyacrylic.
All images were acquired in CADFLI illumination thereby eliminating the need for nile red and fluorescence
If you are an institutional researcher ask for a complimentary sample kit of the reference microplastics shown in the galleries. Contact Lightform for details.
* ColSpec does NOT replace Raman or IR: ColSpec does not report the polymer present in objects.