Pengembangan Kinerja Kolorimetri Multi Spektra Berbasis Arduino Melalui Optimasi Penggunaan Diffuser dan Lensa Fokus

Abstract

The use of synthetic azo dyes is highly hazardous to the environment because they can degrade into aromatic amines that are toxic, non-biodegradable, carcinogenic, and mutagenic. Spectrophotometry is commonly used to analyze these compounds, but such instruments rely on complex optical elements, making them precise to operate and relatively expensive. Therefore, this study developed an RGB LED-based multispectral colorimeter controlled by an Arduino as a simple, inexpensive, and easy-to-control alternative. The colorimeter was designed using an RGB LED module as the light source, a BH1750 sensor as the detector, and an Arduino Nano as the microcontroller. Optimization was performed on the LED diffuser thickness and the HPL focusing lens angle to produce 36 variations of homogeneous and stable light colors, representing wavelengths from 380 to 730 nm at 10-nm intervals. Testing was conducted by comparing the absorbance profiles and performance parameters of acid yellow 17, acid orange 7, and reactive orange 16 between the multispectral colorimeter and a visible light spectrophotometer. Optimal conditions were achieved with a 0.6 mm thick diffuser combined with a 30° HPL focusing lens. The absorbance profiles of the three compounds from both instruments were generally similar; however, shifts in maximum wavelength were observed. Acid yellow 17 shifted from 426 nm to 440 nm, acid orange 7 from 486 nm to 440 nm, and reactive orange 16 from 492 nm to 480 nm. The performance of the multispectral colorimeter is still inferior to that of the spectrophotometer; however, reactive orange 16 showed the closest performance with an R² value of 0,9944; slope of 0,0145; SD of 0,0068; %RSD of 3,8772%; LOD of 1,4089 ppm; LOQ = 4,6965 ppm; and r = 0,9619. These results indicate that as the maximum wavelength more closely approaches its optimal complementary color, the performance of the two instruments becomes increasingly similar.