Here is an idea that you might try. You will need to research it yourself since it is your project. A spectrometer is a device that analyzes a sample's interaction with a spectrum of energies. An infrared spectrometer will analyze the sample's absorption of infrared energy of varying wavelengths. A 'visible' spectrometer will analyze the sample's absorption of visible light. This is accomplished using a rather expensive diffraction grating on commercial instruments. The source radiation (light bulb) contacts the diffraction grating and a rainbow is reflected off the surface. If either the grating's angle from the source to the sample is changed or
the position of the sample and detector is moved, the radiation of a narrow band of colored visible light is allowed to strike the sample. The radiation passes through the sample and a fraction of it is absorbed. That fraction of the narrow band of colored visible light that is not absorbed passes through the sample and strikes the detector which converts the energy into either a voltage or a small current. The signal of the detector is recorded continuously as the narrow band of colored light is swept through the visible spectrum (violet to far red). The resulting absorption spectrum is somewhat
unique for each sample that has a visible absorption... another way of saying it's colored.
Like I said, the diffraction grating is rather expensive but you may find that a prism would work. Fix the prism on a bearig of some sort that has a pointer that points to a graduated scale... a good application for that (almost) useless protractor you aren't using anymore. Alternatively you could use a series of colored LED lamps. I would use three at evenly spaced points in the visible spectrum. One blue, one yellow and one red. You will need to research the wavelengths given off by each of these. You might want to throw in an IR LED as well (that is 'near infrared' to the spectroscopist). It is up to you to determine how to wire and power the thing. You will need to think about a sample holder and a sample 'cuvette'. Choose one that doesn't absorb in the frequency range you will span (no red or blue test tubes!). Alternatively you could use a series of filters... one red, yellow and blue, with a white LED. An array of photodiodes with a built in filter over each of the tiny pixel elements is termed a 'photodiode array'. (http://elchem.kaist.ac.kr/vt/chem-ed...tector/pda.htm
You could also use your computer monitor. Write a program (you choose the language) to change the color of the screen from blue to deep red. In this case, the spectrometer will consist of a sample holder with an integrated detector (photoresistor). Place a sample of the solvent into the instrument and record the voltage produced when the screen if blue, yellow and red (or choose as many frequencies as you have the energy/stamina). Did I mention that you will need to hold this thing up against your monitor? Replace the blank sample with the actual sample and repeat the process. Record the voltages in an Excel spreadsheet. Subtract the voltage from the sample from the voltage of the blank and plot the result Voila! A crude visible spectrum.
Or you could just use something like this:
Search "Photodiode array spectrometer"