23 October 2011
Cell Phone Spectrometer
Posted by Ryan Anderson
Here’s another guest post, this time by Seth Humphries. Seth helped me collect the data at Los Alamos that comprises the bulk of my thesis, so when he asked to write some guest posts for the blog, I was happy to oblige!
Working with light and spectrometers is a part of my job. I worked with lab-built spectrometers and tunable lasers as a grad student at Montana State. At Los Alamos I worked with a mock-up of the ChemCam spectrometers and laser system. I still work with spectrometers at Apogee Instruments. I am also a smart-phone nerd. Recently, I stumbled onto an article about research using a modified cell-phone to enable doctors to perform in-situ analysis by turning the phone into a microscope or spectrometer. They can look at cells or capture a fluorescence spectrum after exciting a tracer in a sample. As typically happens now, doctors will have a nurse draw a sample of blood, label it and send it to a lab for analysis. This takes coordination, correct sample prep, good labeling, careful analysis at the lab to not contaminate the sample, etc all of which cost money and a large amount of time. Instead, this new cell phone device would allow the doctor to smear the sample of blood on a glass slide, shine a light on it and take a picture with his/her cell phone. The doctor can then do the analysis or send the image, not the sample, to someone else for analysis or second opinion. The results will be as accurate as sending the sample to a lab, or so this article claims. I was so excited when I read the blurb that I decided to track down the original article to learn more about it.
The setup is rather simple but yields excellent results and it doesn’t even have to be a “smart phone” just a phone with a good camera. I will focus on the spectrometer portion of the device but they did create and test a ball lens to turn a cell-phone camera into a microscope. You can read the article for more about that.
To create the spectrometer from the cell-phone camera, the researchers used two narrow slits made by spacing pieces of electrical tape on each end of a short piece of ½ PVC pipe and a transmission grating between the slit and the camera lens. The grating splits the light into separate colors, much like a prism, while the original camera lens does all the focusing. This results in the spectrometer having a 300 nm range, meaning the range of colors it detects. For reference, human vision is defined as between 400 and 700 nm. So our eyes have about the same optical range as this camera spectrometer whereas the spectrometers on the next Mars rover as part of ChemCam range from about 230 nm to 800 nm.
This new instrument achieves about ~5 nm resolution. For a spectrometer, resolution indicates how well it can distinguish one color from an adjacent color. Or, more simply put, whether it can, unlike me, figure out the difference between azure and beige. In order to distinguish between several different atoms emitting light, this is not enough but for many applications this is great, ie printers could use an intensity calibrated version of this to do color matching of photos or prints.
In a spectrometer, the image plane (which used to be the film but is now pixels) is a direct image of the input aperture (inlet hole). So, the smaller the aperture, the less fuzzy the image will be. This idea is very analogous to the pinhole camera. Less fuzzy, smearing of colors, means better the resolution. However, the smaller the aperture (or slit for a spectrometer), the less light gets in and the harder it is to detect. This spectrometer uses electrical tape to form about a 1 mm slit. This could be improved. To give you an idea, the spectrometers on ChemCam use a 0.6 mm slit to achieve better than 0.2 nm resolution which is 25x better than the cell-phone spectrometer.
There are some shortcomings to this spectrometer developed by Smith et al. The camera in your phone likely has 8 bit resolution, instead of scientific cameras which have 12 to 24 bits. This means that the intensities recorded by the camera are not as accurate. The camera in your phone is also not optimized to reduce noise from cross talk between pixels and also from heat. However, it performs real time and without collecting samples and sending them away for analysis. I can’t stop thinking of all the possibilities. Here are just a few off the top of my head:
Plants exhibit stress by turning brown. By the time this is evident to the human eye it is too late, but even cameras with just 3 filters can pick up on the signal long before we can with our eyes. With the smartphone spectrometer that Smith and colleagues have developed, you could walk out into your garden every few days, take a picture, and determine if your plants need to be watered! You could even adapt the system to a webcam connected to a PC with some analysis software and a connection directly to your sprinklers. No more brown lawns, no more wasted water. I’m already counting the savings!
What about inside the house? Well, you could take a picture of the paint on your walls and by looking at the relative positions and intensities of the peaks in the spectrum get a great match to your paint from your local paint supplier.
Amateur astronomers would have a field day with their own personal phone spectrometers! From my backyard, I could point my telescope at a star, attach my phone/spectrometer at the eye piece and measure the spectrum of the star. This is not high enough resolution to measure something like Fraunhofer lines but I should get enough detail to estimate the star’s temperature. (Warning: Don’t point your telescope or phone at the sun, I am sure you will melt something you didnt intend to melt!)
I don’t think this spectrometer is quite ready for mass production yet but I am sure it will be soon. I can’t wait! My head still spins thinking about it.
Seth Humphries is the Product Development Scientist at Apogee
Instruments. Previously he was a Post-Doc at Los Alamos National Lab
working on a laser based carbon sequestion monitoring system, ChemCam
validation experiments, etc. He has a long history of involvement in
space-based research and continues to hope to become an astronaut.
Smith ZJ, Chu K, Espenson AR, Rahimzadeh M, Gryshuk A, Molinaro M, Dwyre DM, Lane S, Matthews D, & Wachsmann-Hogiu S (2011). Cell-phone-based platform for biomedical device development and education applications. PloS one, 6 (3) PMID: 21399693
Indeed a wonderful post. I have been studying the chromatography, spectroscopy etc. and do look over the web for latest and informative stuff and just landed on your blog. Thanks for writing such good posts and as I have subscribed to your blog, I do expect that you will be posting nice stuff like this on a regular basis.
Very interesting, we came through and we now offer a spectrometer dongle compatible with Iphone, Ipad and Ipod, covering the visible from 350nm to 700nm…you can find us on http://www.lightingpassport.com
how to obtain that spectrum distribution graph in figure 8 from just a image of the spectrum.suppose i have the spectrum image and i want to convert it to a graph like that…can you tell me the name of any software which can do that?
wanted to ask for more information as much as possible the site, articles that talk, site to compare with molecular absorption spectrophotometer in the visible region, some slides for me to study and which application.
Thanks in advance