攝譜儀

This 3d printed spectrograph is crudely based upon the spectrograph in the article Grove, et. al., American Journal of Physics, 86, p. 594 (2018). For those not familiar with the terminology, a spectrograph takes pictures of images that are representative of light inteity veus wavelength, and this spectrograph is a low-budget way of incorporating spectra and spectroscopy into secondary or undergraduate education. I have included pictures of spectra taken by this spectrograph in the photos (fluorescent lights and Helium). The fluorescent light spectra was used to calibrate the device and the accuracy of measuring unknown wavelengths is ~ 2nm. There are many activities you can do with these spectrographs (for example see, pp. 72 – 81), but I prefer to let my students explore and invent their own light investigatio whenever practical. The complete design uses a webcam with suggested low budget webcams listed in the requirements. If you are not sure you want to purchase a webcam, you can print out the spectrograph box, grating holder, and lid (not the webcam holde). Then (once you add a DVD fragment to the grating holder) you can take pictures or spectra using your cell phone, but this method makes calibration very difficult; the calibration will change any time your phone moves. Requirements: ? A print bed that is at least 220mm by 220mm. ? A low budget webcam. There are specific desig to incorporate either the Logitech c310 HD webcam (preferred) or the Creative Labs Live! Cam Chat HD (unless you remove the IR-filter that is part of the le assembly, the red part of the visible spectrum is too dim). If you want to use a different type of webcam, you will have to either design your own mount or plan to glue it (school glue) to the spectrograph. If you are going with the glue method and an uupported webcam, black cotruction paper may be useful to help block out light. ? A dark color filament for your 3d printer (to eure that light doesn’t pass through the devices walls). ? Flat black paint (to cut down stray reflectio). ? A blank DVD disk and strong shea to cut it (cutting directio are further down). ? School glue. ? General knowledge of how to use a spreadsheet program like Excel or Gnuplot. Printing Itructio: I printed mine using black PLA filament with support everywhere. One can, if one wishes to do it, print with no support (the top of the holes and slits will be slightly deformed). I used a layer height of 0.1mm because I was in no rush, but I believe a 0.3mm layer height will work. I used a 20% infill and 1.0mm thick walls. This spectrograph design assumes a well calibrated 3d printer (particularly for the webcam holde). Webcams are a fixed size and if your printer is printing shapes that are a little largemall, the webcam holder will not hold the webcam very well. To test your printer’s calibration, you will fit need calipe accurate to ~ 0.01mm. Then print out a shape with a side length of exactly 100mm. If the calipe confirm that the side’s length equals 100mm, your 3d printer is well calibrated. If not, you should scale the design up/down so that it prints the parallelepiped to the expected size. Make sure you do this test for the xyz directio. If you don’t have access to calipe, you could try and guess the needed scaling factor. Print out a webcam holder and then see if the webcam actually snaps into and is firmly held by the holder. If it doesn’t fit well enough, try changing the scaling by plus/minus 0.50% and printing another webcam holder. Within a few test prints and fits, you can get the pieces sized correctly. Assembly Itructio: Obviously start by printing the spectrograph box, lid, grating holder, and webcam holder (if you are using the Logitech c310 or Creative Labs Live! Cam Chat). Make sure you remove the support pieces on the spectrograph box (there should be a circular hole on one side and slit on another side). The slit and hole should allow light to pass directly through them. Paint the iide of the spectrograph box with flat black paint. Failure to do so will result in extra spectral features (the light through the slit reflects off the walls and these false lines move with changing alignments). Plastic tends to have glossy sides and this is unwanted iide the spectrograph box. Cut a blank DVD to fit into the grating holder (read all the cutting itructio before attempting). The DVD fragment must come from the outer edge of the DVD and the cut DVD fragment should be approximately 25mm by 25mm. The outer edge of the DVD fragment should be glued onto the grating holder (school glue) such that the curved edge of the DVD abuts the curved edge of the grating holder and the diffractive side outward. Imagine how the cut fragment will fit on the holder and use a felt tip marker to mark the DVD. Now to the actual cutting. DVD’s will crack, splinter, and/or delaminate if cut improperly. Fit, you really need a strong set of shea (I have used tin snips, sheet metal shea, and what my wife calls rose clippe). Attempting to saw through a DVD will probably cause serious cracks and delaminatio. Secondly, you need to heat the DVD before cutting. The DVD should be hot enough that holding it by hand would be painful, but not so hot that the DVD warps or deforms (if you smell hot plastic, remove the DVD from the heat source). I have used a heat gun, a hand held hair dryer, and even steam from a tea kettle (the spout directs the heat) for this purpose. Once the DVD is hot, cut it with the shea. If you see cracks, the DVD wasn’t hot enough. Once the DVD fragment is cut to size, glue it to the grating holder and wait until the glue dries before proceeding. If I am making multiple spectrographs, I can usually get 5 – 6 usable DVD fragments from a single DVD blank. The grating holder (with attached DVD fragment) should slide into the box so that triangular flares on the grating holder base slide into the triangular slots on the box. It is supposed to be a tight fit. This will require some pushing from one side of the grating holder to the other and squeaking and the occasional snapping sound are normal. Make sure that the grating holder’s base is flush with the bottom of the spectrograph box. If for any reason the grating holder seems loose, use school glue to fasten it to one position iide the spectrograph box. Now you are ready to verify that the spectrograph forms spectra. Place the lid on the spectrograph box. Aim the slit (at the front of the spectrograph box) toward a light source (sun light, light bulb, or fluorescent lights are fine) and look into the box through the circular hole. Small alignment changes can make big differences in what you see. At optimal alignment you should see very bright features. Once you see these bright features, you can glue the lid to the spectrograph box. Finally, you need to attach the webcam to the spectrograph box in order to capture the spectral image. If you are using the Logitech c310 HD webcam or the Creative Labs Live! Cam Chat HD webcam, iert the webcam into the webcam holder. Then connect the webcam to a computer and verify that the webcam is working (selfies are fine). Before gluing the webcam holder to the spectrograph, I like to verify that all spectrograph components are aligned. Holding the webcam holder in position (it should be a somewhat tight fit in the marked recess near the circular hole on the spectrograph box), aim the slit at a light source and see if the spectra is seen on the computer screen. Once this is achieved glue the webcam holder to the spectrograph box. Congratulatio, you have completed assembly. Calibration Itructio: Seeing spectra is great and to some extent interesting, but if you want to know the wavelength of spectral lines, you have to calibrate the device. A low budget way of accomplishing this is to use fluorescent lights (most schools and univeities use fluorescent light tubes so you probably have these in abundance) and an open access program called imagej ( or if that fails do a google search for imagej download). To calibrate we are relating the horizontal location of a pixel with a bright spectral feature to the wavelength of that spectral feature. Step 1: take a picture of your fluorescent light spectra. Step 2: Using imagej’s rectangle tool, draw a rectangle around your spectra so that it extends horizontally across the entire picture and the vertical part of the rectangle only cove the region where the spectral lines appear. In this way, the pixel position on the far left side of the photograph is always 0. Failing to do this will result in a very poorly calibrated device. Step 3: In imagej go to Analyze ? Plot Profile. This will give an inteity vs. horiznontal pixel location graph. Step 4: You now have to find the horizontal pixel location for every peak and match these peak locatio to known wavelengths. Most fluorescent light tubes (see ), have the following prominent spectral lines: 405.4nm, 435.8nm, 487.7nm, 546.5nm, 577.7nm, 580.2nm, 584.0nm, 587.6nm, 593.4nm, 599.7nm, 611.6nm, and 625.7nm Guess which wavelength corresponds to which horizontal pixel location. If you guessed correctly, a wavelength vs. horizontal pixel location graph will have a slightly curved shape. The photo has dashed vertical lines indicating the correspondence between some of the spectral features and wavelength (it will greatly help if this is the fit time you have done something like this). Keep making guesses until this graph forms a smooth curve. Use a spreadsheet type program (for example, Excel) to fit a quadratic (2nd order polynomial) to the data. In the example shown in , the relatiohip between horizontal pixel location (x) and wavelength (lambda) is given by: lambda = (1.934E-05 nm/pixel^2)x^2 (0.1966 nm/pixel)x 239.7nm. Note: you will very likely find a different calibration curve for your spectrograph; the above equation was used as an example. As an itructor you can decide for youelf whether you want your students to do the calibration or whether you give the students the calibration curve (it can be done as simply as writing it down on a strip of masking tape and attaching it to the device). Using a calibrated spectrograph: When looking at unknown spectra, you can now determine wavelength using the previously found calibration curve. For example, if we use the calibration curve shown in , the pixel at the far left (pixel location = 0) would have a wavelength of 239.7 nm and the pixel at the far right (pixel location = 2591) would have a wavelength of 878.9 nm. The photograph shows how to make measurements for a spectra. Fit take a photograph of the “unknown source” – helium in this case. Then use imagej to find the plot profile. I typically copy this data to a spreadsheet program (Excel) and then use the spread sheet to find the wavelength corresponding to each horizontal pixel location). Finally, I plot the pixel inteity vs. wavelength and use this to find the wavelength for every bright line. This plot is shown in the photo . The table at the bottom of shows the correspondence between the measured features and the known wavelengths (via the United States National Ititute of Standards and Technology - NIST). Final comments: This spectrograph was designed with low cost and general ruggedness in mind. Besides looking at just atomic sources, students can also examine light from just about anything without much worry about potential damage to the device. My students have already looked at the Fraunhofer lines in sunlight, reflection spectroscopy (i.e., white light reflected from differently colored object into the spectrograph) including dandelion flower spectroscopy, absorption spectroscopy (i.e., white light passing through water food coloring), and even fluorescence. I hope you like this design and find use for it.