Build Your Own Spectrometer ECE 198 Lab Manual Introduction/Overview: The purpose of this lab will be to build and characterize your own spectrometer out of standard laboratory optical components. Over the two weeks of this lab, you will build a spectrometer capable of detecting the intensity of light input into the spectrometer as a function of the wavelength (or frequency) of the light. You will calibrate your spectrometer to a known source, characterize your spectrometer to determine the resolution of the system and its response curve as a function of wavelength. Ultimately you will use the spectrometer to measure the optical properties of a variety of sources and materials. The first week of the lab will be focused on building the spectrometer and understanding the role of the various components in the spectrometer itself. You will gain familiarity with the standard opto-mechanical components we use in lab such as posts and post-holders, bases, rotational stages, and optical mounts. In addition, you will gain familiarity with many of the optical components you will be using throughout the semester; this may include gratings, lenses, slits, mirrors, and/or prisms. By the end of the first week you should have a working spectrometer and should begin testing the spectrometer’s performance, looking at resolution, spectral response, and ease of operation. In your second week, you will use your spectrometer to investigate the spectral output of a number of sources (lasers and LEDs), and also to determine the optical properties of a number of materials (filters, colored glass, etc.). In addition, you should spend some time thinking about, and testing out, potential additions to the spectrometer that could improve performance of the system. Your lab report will be in the form of a brochure for the spectroscopy system you have just built. You will want to include diagrams of the system’s optical layout and examples of the system’s performance, including some of the spectra taken with the system. In addition, you should figure out the cost of the spectrometer to assemble, build, test, and ship, and use this to determine a price point at which you can sell the system. What is the competition, and how does your system compare?

1

Basics of Spectroscopy: Spectroscopy, technically, is the action of determining a property of a material as a function of energy. We will be doing optical spectroscopy in this course, which means we will be concerned solely with the optical properties (absorption, scattering, reflection, transmission, luminescence) of our materials systems as a function wavelength (energy). As such, you will be building a system that can investigate the optical properties of materials (and light emitters). In future labs you will be using commercially available spectrometers to study nanostructures, but in this lab you will actually build a spectrometer yourself! Optical spectroscopy can cover a wide range of wavelengths (energies), from the deep Ultraviolet (UV) to the terahertz wavelength range. Our spectrometer will mostly cover the visible frequency range to the very short-wavelength IR. In Table 1 below you will find the optical spectrum divided into wavelength ranges, with wavelengths, frequencies, and energies listed, as well as a brief description of the wavelength range. Please bear in mind that what is shown below is a small part of the larger electromagnetic spectrum, which includes X-rays, gamma-rays, radio waves and microwaves, etc, all wavelengths we will not be investigating. Energy Frequency Wavelength -19 12 (1eV=1.6x10 (THz, 10 Hz) (1 nm= 10-9 m) Joules) (1 µm= 10-6 m) 4.1-1240µeV 1-300GHz 30cm-1mm Microwaves: can be generated by electronic sources, like your microwave oven, wireless phones, RFIDs. 1.24-41 meV 0.3-10 THz 1mm-30µm The terahertz: These are frequencies hard to generate optically, but also a little too fast to generate electronically. Used for body scanners at airports. 41-124 meV 10-30 THz 30-10 µm The Far-IR: Also associated with heat (but lower temperature), sensing applications. 124-775 meV 30-188 THz 10-1.6 µm The Mid-IR: Wavelength associated with heat, used for sensing, nightvision applications. 0.775-1.77 eV 188-430 THz 1600-700nm The Near-IR: These wavelengths are most frequently used for telecommunication applications. 1.77-3.1 eV 430-750 THz 700-400nm Visible: These are the wavelengths we can see with the naked eye. 3.1-124 eV 750THz400-10nm Ultraviolet (UV): Can be emitted 30PHz from the sun (responsible for sunburns), used for flouresence (black lights)

2

If we want to understand how a material or a device behaves as a function of energy, then we need some way to investigate the device properties as a function of energy. In other words, we need to look at the light coming from, passing through, or bouncing off our material or device at different energies. For simplicity, let’s call our material or device our device under test (DUT). Device Under Test (DUT): The material, structure, or device we wish to obtain spectroscopic information from. As we have already hinted, there are different ways to interrogate our DUT, depending on what information we want from the device. To understand this, we need to understand how materials interact with light. Later in the course we will spend a lot of time discussing the physics of light/matter interaction, but for now, we can take a very simplistic view. Below are three configurations where we might want to know about the interaction of light and our DUT.

Figure 1. Schematics for a) Emission, b) absorption, and c) reflection spectroscopy experiments. In the first configuration, which we will call Emission Spectroscopy, we wish to know the nature of the light emitted by our device under test. There are numerous ways to get a material or a device to emit light, but in general, the device must be “excited” in some way. Energy is put into the device, and this energy is converted into light emitted from the device. Typically, materials/devices are excited electrically (for instance a light emitting diode, or LED, or a laser) or optically (for instance, my glow-in-the-dark Spiderman pajamas). However the light is excited, the spectrometer collects this light, and tells us the wavelength(s) of the light coming off of our sample. Figure 2 shows what the emission from a red LED might look like. As you can see, the emission intensity is plotted as a Figure 2. Example emission spectrum for a function of wavelength. Each point on red LED. 3

the plot tells us the strength of the signal (or the amount of light emitted) at each wavelength. Because this is a red LED, the emission peaks at about 650nm, right in the middle of the red part of the visible wavelength range. IMPORTANT: When you plot ANYTHING on a graph, you MUST label as much of the plot as you can! This means titles for the x and y axes, units of measurement, and ideally, a legend telling the viewer any other important experimental parameters, such as sample name/information, excitation source, etc. In the second configuration, labeled Absorption Spectroscopy, we are interested in knowing at what energies, or energy ranges, a material absorbs light. Because we do not know, a priori, at what energies a sample will absorb, we must interrogate the sample with as many possible different wavelengths as possible, so for this, we use a white light source. Because perfectly white light contains all the colors of the optical spectrum, this allows us, again in an ideal situation, to measure absorption across the widest range of frequencies. In absorption spectroscopy, we shine the white light at our DUT, and the spectrometer collects all of the transmitted light. What we want to know is how much light at each wavelength is absorbed, but we cannot determine this unless we know how much light is incident on the sample. So to determine this, we first need to take a background spectrum. A background spectrum, B(), is basically the emission spectrum of the exciting source, and it tells us what kind of light we are interrogating our sample with. Once we have our background spectrum, we then put the sample in the optical path and take a second spectrum (the sample spectrum, S()). This should give us the background spectrum minus all of the light absorbed by the sample. Now, we take the sample spectrum and divide by the background spectrum and we get our transmission spectrum, T()=S()/B(). If we assume that all light on the sample is either transmitted or absorbed, then our absorption spectrum is simple A()=1-T(). However, in reality, light can be scattered or reflected from the sample, and never makes it to the spectrometer, but at the same time is not absorbed. For this reason, A() is more accurately desctribed as an extinction spectrum. Figure 3 gives an example of what the data sets for such an experiment might look like.

Figure 3. Data for (a) background and (b) sample spectra. absorption/extinction and transmission spectra from (a) and (b).

4

(c) Calculated

The third configuration shown in Figure 1 is that of reflection spectroscopy. We will not be performing reflection spectroscopy in this class, but the basic idea is very similar to that of transmission spectroscopy. In reflection spectroscopy we wish to understand how light reflects off of a surface, but again, we need to understand what light is incident upon the surface before our reflected spectrum has any meaning. We need a background spectrum. For a background spectrum in reflection spectroscopy, one generally tries to use a perfectly reflecting (so that all of the incident light is reflected) and smooth (so that no light is scattered) surface, such as a mirror. Once you have the reflection spectrum off of your mirror, you replace the mirror with the DUT and take your sample spectra. From here the data analysis is pretty much the same as for absorption spectroscopy. Spectrometer Basics Up until this point, we have simply taken it for granted that we have some magical machine that can measure light intensity as a function of wavelength. This is what the box labeled “spectrometer” does in Figure 1. But the goal of this lab is to build that magical box, so we should probably understand the basics underlying our magical box. In order to detect the intensity of light as a function of wavelength, we need to somehow split our light up as a function of wavelength. Figure 4 shows the basics of spectrometer design.

Figure 4. Basic configuration for a spectrometer. In all cases, light enters into the spectrometer and is dispersed by a “dispersive element”. In the top schematic, the dispersed light is simultaneously detected by an array of detectors, in the middle, a single element detector is used, and scanned across the dispersed light. Alternatively, the dispersive element could be rotated, and the detector remaining fixed (bottom).

5

Here the light from the external optics is incident upon a dispersive element inside the spectrometer. The dispersive element ‘disperses’ or spreads the light by wavelength. The light intensity at each wavelength is detected either simultaneously by an array of detectors (usually this array is a charge couple device, or a CCD, like the camera in your cell phone), or by a single element detector that moves and measures intensity as a function of position, or finally, one can rotate the dispersive element, and hold the detector fixed. Dispersive Elements In this lab you will get to choose the dispersive element you are going to work with. We have three elements you can choose from. The first is a prism. A prism disperses light by refracting light of different wavelengths at different angles. We have discussed the concept of refraction in lecture, and we have discussed the fact that in all real materials, the refractive index varies as a function of the wavelength of the incident light, an effect known as dispersion. Prisms make use of the dispersion of glass to spread different wavelengths of light at different angles. Figure 5 shows a basic schematic of prism dispersion.

Figure 5. Schematic of dispersion through a prism, from http://educationalelectronicsusa.com/l/light-XV.htm . The second and third dispersive elements you can use are both diffractive elements, these are diffraction gratings, which operate by the phenomenon of diffraction. The gratings can be transmissive or reflective diffraction gratings. Let’s start by looking at a transmission grating. The transmission grating consists of a series of slits in an opaque film. Light transmitted through each grating slit diffracts, as we have discussed in class, emanating in all directions. For a single slit, transmitted light is simply diffused in all directions. However, a series of slits separated by a distance d acts like an array of point sources. The light from each of these sources interferes with the light from the other sources, creating a pattern of constructive (intensity peaks) and destructive (intensity minima) interference. The position of the constructive interference peaks depends on the wavelength of the light emitted from the slits, so that different wavelengths are ‘bent’ at different angles. The angles at which the light is diffracted can be calculated using simple geometry, knowing that a path-length difference of m, where m is an integer, gives constructive interference, as seen in Figure 6. Reflective diffraction gratings operate with very similar physics to transmissive diffraction gratings, except light is now reflected from a corrugated or periodic metal surface, instead of transmitted through an array of metal slits. Imagine we take the same 6

grating we were using as a transmission grating, but now we look at the light reflected from the grating. If the material between the slits is reflective, it will act as a series of point source (reflecting the incident light from the narrow metal stripes). Thus, we can use the same equations as for the transmission grating to determine the dispersion of the reflection grating!

Figure 6. Top: calculation of diffraction position as a function of wavelength for a diffraction grating with periodicity d, and a screen distance L. Bottom: Schematic of light diffraction through a transmissive diffraction grating for L=10cm, d=1µm, for light of wavelengths 400, 500, 600, and 700nm.

7

It should be noted that actual commercial gratings, and in particular the ones you will be able to use in class, are not quite as simple as what we have described above. These grating can be engineered to preferentially diffract into certain orders or to work preferentially for certain wavelengths (by using something called ‘blazing’). However, the basic physics of the grating remains basically unchanged, and for the purposes of this class, we will still to the simple description above. Spectrometer Layout Now that we have a way of separating our wavelengths, we are almost there! We still need to get our light into the spectrometer, to the grating, and then to the detector. In most cases, the light going through your DUT (in absorption spectroscopy) or coming from your sample (in emission spectroscopy) is not collimated. Collimated light is light which is neither converging nor diverging. In practice, for absorption spectroscopy, you will usually want to focus your light on your DUT, which means the light will be diverging upon leaving the sample. For emission spectroscopy, your light is often coming from a small sample area, and diverges as it leaves the sample. In both cases, you will want to 1) collimate your light and 2) focus it into the spectrometer.

Figure 7: Schematic of external optics and internal spectrometer configuration. Once you have your light in the spectrometer, it is essential that you collimate the light before it hits the dispersive element. This is because the diffraction or refraction of incident light on the grating depends on the angle of incidence. If your light is hitting the dispersive element with a broad range of angles, it will diffract across a broad range of angles as well, and your wavelengths will overlap in space, and separating them, in order to detect the intensity of each wavelength, becomes exceedingly difficult. Figure 7 shows the schematic for a typical spectrometer configuration. Note that in typical commercial spectrometers, the lens are replaced by curved mirrors to reduce chromatic aberration and transmission losses in the lens itself. Let’s walk through the set-up in Figure 7: -Light leaving your DUT (or source) is diverging, which means you will want to collimate this light. Do this by placing a lens with focal length f a distance of approximately f from the DUT. -A second lens focuses the light into the entrance of the spectrometer.

8

-Next, collimate the light entering the spectrometer with a lens, and send this light to your diffractive element. -The diffractive element should then spread the light. -A lens placed some distance from the diffractive element will capture the dispersed light and focus the light down to the spectrometer exit. -A detector at the exit records the intensity of the focused light. A couple of things are worth noting. The performance of your spectrometer depends entirely on your ability to isolate different wavelengths of light. The combination of the diffractive element and the last focusing lens are designed to focus only one frequency on the detector. Getting the positioning and spacing of these components right will be key to your spectrometer performance. One way to prevent unwanted frequencies from leaking onto your detector is by using a slit. You will see, in Figure 7, that I have placed a slit and the entrance and exit of the spectrometer. The exit slit prevents unwanted frequencies from reaching the detector, while the entrance slit prevents light rays from taking unwanted paths from the input to the detector…also hurting your resolution. The smaller the slits, the higher the resolution of your system, but there is a trade-off, because the smaller you make the slits, the less light you get through. It’s hard to have a working spectrometer that lets no light in! Characterizing Your Spectrometer As part of your lab report for this lab, you are asked to discuss the performance of your spectrometer. As you work with your spectrometer, keep this in mind. Some benchmarks of spectrometer performance do not require any real expertise in optics or engineering. Since you are the users and the builders, you can give yourselves feedback on the performance of your system. I am sure you will be able to come up with a number of benchmarks for your system, but to give you an idea of what we are looking for, think about the following: how easy or intuitive is your system to use? Is it robust? Would a tiny bump to the system throw it out of whack? How reproducible are results given by the spectrometer? If you do the same experiment twice in a row, do you get the same results? Other aspects of spectrometer performance are a little trickier to determine. Key among these is your spectrometer’s resolution. In spectroscopy, resolution means the following: it is the smallest wavelength difference between two signals that can be differentiated with your spectroscopy system. In other words, if you have a narrow emitter at wavelength 1=650nm, and you add to this a second emitter at a wavelength 2=650+nm, what is the smallest , in wavelength, that gives you two distinct peaks in your spectrum. This is the resolution of your system. That having been said, finding two narrow linewidth sources, one of which can be tuned, is very difficult, and we don’t have this in lab. So the best we can do is to use a single wavelength, narrow linewidth source (a gas laser) and measure the linewidth of the laser as it appears on our spectrometer. Because the laser linewidth will almost certainly be smaller than that of our system, we can assume that our resolution is basically what we measure the linewidth of the laser to be.

9

Equipment Before we start the lab itself, let’s familiarize ourselves with the equipment we will be using. This lab provides an excellent introduction to much of the optical equipment we will be using throughout the semester. You will work with lenses, mirrors, diffraction gratings, prisms, apertures/slits and all of the optomechanical components required to mount these optical components. In the first week of lab, you should have had the opportunity to play around with this equipment and gained a familiarity with how we build our optical set-ups in this lab. NOTE: Never touch any optical components on their faces. Your fingers have natural oils on them and these oils will remain on the lenses, mirrors, gratings and filters. Sometimes the oil is simply difficult to clean, but other times it can actually damage the surface of the optical component. Wear Nitrile gloves when handling optics. If this is not possible, always hold optics by the outer edge. These optics are expensive, and we cannot afford to replace them after every lab!!

10

Procedure Week 1 The goal for this week is to get a feel for how the dispersive elements (prism, transmissive grating, and reflective grating) work and to begin constructing a spectrometer. By the end of the lab session, you will have the framework of a spectrometer constructed around the reflective grating. The first part of the lab is to experiment with the three dispersive elements to understand how they work and how they can be exploited for our purposes. But first, we have to set up the light source and collimating lens. 1) We first need to set up the white light source. Use a 1” diameter lens mount (LMR1) mounted using a post and post holder. Use the white light fiber adapter to mount the fiber in the lens mount. (Note: the height of the white light source is the height the light will be throughout the spectrometer path.) 2) In order to make our setup as simple as possible (and since we aren’t trying to measure anything yet), we will place the white light source just before the entrance slit shown in Figure 7. Place a pinhole (mounted using a post and post holder) just in front of the light source. 3) Next, we need to use a lens to collimate the light. The lenses are already mounted in LMR1 lens mounts, but you need to set it up using a post and post holder. Adjust the placement of the lens until the light is collimated. At this point, we have a beam of light approximately 1” in diameter. Next we want to insert a dispersive element, so we can see how the light is separated. Try each of the three available dispersive elements, and think about the benefits and drawbacks of each for our purpose. 4) In order to easily see the effects of the dispersive elements, we need to be able to change the incident angle. Since we can’t move the optics we’ve already set up, we want to be able to rotate the dispersive element itself. Attach a rotating mount to your breadboard in the line of the beam path. 5) Each element is already mounted in a post-ready holder. Use a post and post holder to mount the element on the rotating mount. 6) Hold a spare piece of paper or business card in the dispersed light path to see the effect of the element. 7) Play around with the incident angle to see what happens to the dispersed light. Record your observations in your lab notebook. Some things to think about: how well the light is dispersed, spacing between wavelengths, and (for the gratings) the presence of higher order modes. Next we are going to start building a spectrometer around the reflective grating, which is the dispersive element most commonly used in commercial spectrometers.

11

8) Place the reflective grating in the dispersive element position. 9) Using the white light source and subsequent dispersed light as a guide, choose a position to place the focusing lens. Remember, you have to be able to see all of the dispersed wavelengths through the lens as you rotate the grating. 10) Using Figure 7 as a reference, move the white light source to the DUT position and the collimating lenses to focus on the pinhole. Aligning the lenses is a tricky process. If you are having difficulty, ask the TA for help. At this point, we have a nearly completed spectrometer. The white light source should be split by the reflective grating, and the dispersed light should be focused to a point by a focusing lens. If you place a piece of paper at that focus point, the color of the light should change as the dispersive element is rotated. This concludes the first week of lab. If you finish early, you can start next week’s lab by adding a detector to the setup. Week 2 In second week of the lab, you will add a detector to your spectrometer, and use a Labview program to record data points. At the conclusion of the lab, you should have a white light background spectrum as well as various other spectra to calibrate the spectrometer and measure its performance. 1) We will use a ThorLabs DET36A Silicon-based detector to measure the light intensity at a given wavelength. Mount it at the focus point of the focusing lens as shown in Figure 7. Use a pinhole to block stray light from reaching the detector. 2) Use a BNC to alligator wire to connect the DET36A detector to pins 1 (GND/black) and 2 (red) of the National Instruments USB-6008 DAQ. Use a USB cable to plug the DAQ into your computer. 3) Use the provided Labview program to record the intensity of light collected by the detector as a function of angle. We will learn more about how Labview works in future labs, but for now follow the instructions below. Feel free to ask the TA for help. a. Open “Spectrometer _Lab.vi” in Labview. b. Press the box with the arrow in the upper left hand corner just below the “Edit” menu. This will start the program running. c. There is a box labeled “Save?” To record a spectrum, select the “Save” radio button. If you just want to play around with the software, select “Don’t Save.” d. In the middle, there are two boxes labeled “Angle” and “Current.” The “Current” box actively displays the intensity detected by the detector. Record the angle on the rotating stage in the “Angle” box, and press the “Read” button to record the value e. To the right, there a structure labeled “Array.” After pressing the “Read” button, the angle and corresponding intensity will be stored in the array. This array is what is saved to file as your spectrum.

12

f. The graph will update in real time to show the spectrum you have recorded. g. Press the “STOP” button in the upper left hand corner to end the program. You must stop and restart to take a new spectrum. If you chose the “Save” option, it will ask you to select a file path to save your file. 4) Familiarize yourself with the Labview program, including how to store data. It will be stored in text format, but can be opened in Excel, where the data can be plotted. Using the Labview program, we have a graph of intensity vs. grating angle. Since the changing the grating angle proportionally changes the wavelength incident on the detector, the shape of our graph matches the spectrum. However, we ultimately want to plot intensity vs. wavelength (not angle), and therefore need to match grating angles to the appropriate wavelength. 5) In the lab there will be a box containing a variety of filters. These filters will block all wavelengths of light below a certain value, depending on the filter. If you take the spectrums of white light passing through a few different filters, you can match the known cutoff wavelength with a given angle of your reflective grating. With a few data points, an approximate translation between angle and wavelength can be determined. (Note: once this calibration step has occurred, it is essential that the placement of every optical element is left unchanged.) 6) Take a white light spectrum. Use the calibration measurements you took previously to create a plot of intensity vs. wavelength. Congrats! You now have a working spectrometer. Play around with different DUTs, and test the performance of your spectrometer. Can you insert a random filter and identify the cutoff wavelength using your spectrometer? Experiment with an eye toward the lab write-up. Make sure you take whatever data will be necessary. For example, use a provided laser diode (driven by 3V off the power supply) to determine the resolution of the spectrometer as discussed above. Remember to be careful when using a laser, and feel free to ask the TA for help.

13