Arduino Spectrophotometer for Food Dye Mixtures

ISEF Category: Chemistry

Subcategory: Analytical Chemistry  ·  Difficulty: Intermediate  ·  Setup: School Lab  ·  Time: 1 to 2 Months

The Hook

A tiny sensor can miss what your eye can still sort out. That is the whole challenge here. You will build a simple spectrophotometer, then use math to pull apart dye signals that overlap in messy mixtures. That turns a cheap device into a smarter analyzer.

What Is It?

A spectrophotometer measures how much light a sample absorbs. Think of it like checking which colors get blocked as light passes through a liquid. If you shine several LED colors through a dye mixture and measure the light that makes it through, each dye leaves a different pattern.

The catch is that cheap hardware has limited color resolution. Two dyes can look almost the same to the sensor, even if their absorbance curves are not identical. Non-negative matrix factorization, or NMF, helps by finding hidden patterns in the mixed signals. It only allows values that make physical sense, so it can separate a blended signal into parts that likely came from different dyes.

That makes this project more than a gadget build. You are pairing hardware, calibration, and data analysis to estimate what is inside a mixture from imperfect measurements.

Why This Is a Good Topic

This is a strong science fair topic because you can test it with real samples, clear numbers, and repeatable mixtures. It connects to food quality testing, environmental sensing, and low-cost chemistry tools. You can learn sensor calibration, absorbance, error analysis, and data decomposition without needing a full research lab. You also get room to improve the design, which gives you a real research angle.

Research Questions

• How does the number of LED wavelengths affect how accurately NMF separates overlapping food dye mixtures?
• What is the effect of using different LED wavelength sets on dye concentration prediction error?
• Does adding a reference blank improve the stability of the spectrophotometer calibration?
• To what extent can NMF recover the proportions of three-dye mixtures from low-resolution absorbance data?
• Which sensor geometry gives the most repeatable absorbance readings across replicate samples?
• How does sample color depth affect the accuracy of deconvolution at low and high absorbance levels?

Basic Materials

• Arduino board or compatible microcontroller.
• Photodiode sensor with resistor components or a photodiode module.
• Three to five LEDs at different wavelengths.
• Breadboard and jumper wires.
• Sample holder or small cuvette stand.
• Transparent cuvettes or clear test tubes with consistent dimensions.
• Distilled water for blanks and rinsing.
• Artificial food dyes in several colors.
• Digital kitchen scale with 0.1 g accuracy.
• Measuring spoons or graduated droppers.
• Dark box or opaque container to block room light.
• Notebook or spreadsheet for calibration data.

Advanced Materials

• Arduino board or compatible microcontroller.
• Photodiode with transimpedance amplifier circuit.
• Three to five narrow-band LEDs with known peak wavelengths.
• Optical filters for stray-light control.
• Fixed cuvette holder with repeatable alignment.
• Matched glass or plastic cuvettes.
• Laboratory pipettes and tips.
• Analytical balance.
• Volumetric flasks and Class A glassware.
• UV-Vis spectrophotometer for comparison data.
• Pure dye standards or certified dye solutions.
• Light shield enclosure with internal black surfaces.
• Stable power supply for LED current control.

Software & Tools

• Arduino IDE: Programs the microcontroller and reads sensor output from the photodiode circuit.
• Python: Cleans calibration data and runs the matrix factorization analysis.
• NumPy: Handles arrays, signal calculations, and concentration matrices.
• SciPy: Fits calibration curves and supports numerical analysis.
• ImageJ: Helps check LED spot size, alignment, and sample holder consistency if you document the setup with photos.

Experiment Steps

1. Define the dye pair or dye set you want to separate, then decide which mixture space you will test first.
2. Design the optical path so one LED, one sample, and one detector stay aligned the same way every time.
3. Build a calibration plan that converts raw sensor readings into absorbance-like values you can compare across wavelengths.
4. Collect single-dye standards and mixed samples so you can create the data matrix for NMF.
5. Choose the number of hidden components you expect, then test whether the math recovers the known mixture ratios.
6. Compare your model against a simple baseline, such as using one wavelength alone or a direct linear fit.

Common Pitfalls

• Letting room light leak into the setup, which adds noise and hides the real absorbance signal.
• Using LEDs with broad or poorly matched spectra, which makes different dyes harder to separate.
• Changing cuvette position between runs, which shifts the light path and ruins repeatability.
• Skipping blank correction, which leaves the solvent and container signal mixed into every reading.
• Treating NMF output as exact dye identity, which can overstate what the hardware can really resolve.

What Makes This Competitive

A stronger version of this project goes past a simple working prototype. You would compare several wavelength sets, test multiple mixture types, and show which design choices improve separation most. You would also report uncertainty, residual error, and how often the model fails on blind samples. That kind of analysis shows you understand both the chemistry and the data.

Project Variations

• Test whether the same setup can separate natural food colorings instead of synthetic dyes.
• Swap NMF for principal component analysis and compare how well each method recovers mixture ratios.
• Use smartphone camera readings instead of a photodiode and see whether image-based color data performs better or worse.

Learn More

• PubChem: Look up absorption-related information for common food dyes and find linked peer-reviewed papers by searching dye names.
• NIH PubMed: Search for review articles on spectrophotometry, chemometrics, and non-negative matrix factorization in analytical chemistry.
• NIST Chemistry WebBook: Check spectral and chemical reference information for compounds when available.
• NOAA Education Resources: Review light, optics, and sensor basics for building a controlled measurement setup.
• MIT OpenCourseWare: Search for analytical chemistry or linear algebra course notes that explain calibration, matrices, and data fitting.