I tried to use the photoelectric effect as a paradigm lab for our unit on atomic, nuclear, and particle physics. Ultimately, it was a failure. This post will do two things: (a) explain how to create and use apparatus to make this lab work, and (b) analyze why it didn’t work for me.
Two sources need to be acknowledged first. As always, Arons did a fantastic job of explaining why the photoelectric effect is difficult for students, and he also provides a good plan if you wish to introduce it. Secondly, the design of my apparatus was inspired by one designed by Rice, Garver, and Schober.
The photoelectric effect is a simple idea with devastating consequences. Essentially, light shining on a metal causes electrons to be knocked off the surface of the metal. If two electrodes are placed close together and connected with a metal wire, these electrons cause a current to flow in the wire, which can be detected with a microammeter.
Thus, we should be able to detect the photoelectric effect with just four pieces of equipment: a pair of closely-spaced electrodes, a microammeter, some wire to connect them, and a light source.
The first of these is the toughest: the electrodes should be in a vacuum, so the electrons are not absorbed or scattered by air molecules. The 1P39 vacuum tube is perfect. There are usually a few such “new old stock” on ebay, at a price of about USD 40 each. I’d suggest getting stands for the tubes, too: the part number is 27E122, and the active sockets are 4 and 8.
A decent multimeter can accurately read down to 0.1 microamperes, so that’s no problem. I put little wire loops on my stands so that we can use wires with alligator clips. Ambient light is enough to produce about half a microamp of current.
I demoed this with the students, and they were able to draw the relevant simple circuit diagram and construct their own simple devices pretty easily. I had them spend about 15 minutes investigating the impacts of light intensity and light colour (frequency) on the current. We didn’t spent much time analyzing this, but the results were clear and straightforward: more light means more current, and different colours seemed to have different currents as well, but in a less-obvious way.
Okay great, we’ve got light creating electricity. So what? The clever trick is to apply a (variable) potential difference across the electrodes. If you adjust the potential difference so that a large positive charge is on the “receiving” electrode, the electrons will accelerate to that electrode, and the current should increase to a certain saturation point. This is where all the electrons are reaching the “receiving” electrode.
On the other hand, if you reverse the potential difference and apply a negative charge to the “receiving” electrode, the electrons will be repelled. However, at a very low potential, some of the electrons will have enough kinetic energy to overcome the electrostatic repulsion and enter the “receiving” electrode. Thus, the game is to slowly increase the potential until all the electrons are stopped from reaching the receiving electrode.
At this point, the kinetic energy of the electrons is equal to the electric potential energy provided by the potential difference (ie: KE = q * V). Thus, given the charge of the electron, we can use the Stopping Potential to calculate the amount of kinetic energy the electrons have when they exit the metal.
The range of potentials needed is about 0 to 1.5 V. I used some 10 kOhm potentiometers and 56 kOhm resistors to make potential dividers that would give appropriate voltages from a 9V battery. I like working with 9V batteries, but there’s no reason not to use a AA directly on a potentiometer instead.
When you add the potential difference, you also need a voltmeter and another pair of wires — things start to get a bit messy on the lab table. This was where my students got caught. Although they were able to digest the idea of a stopping potential (with a bit of prompting, the students put together the idea without my help), actually constructing the circuit on the lab bench proved to be too much.
A few (top) students were able to recall what they knew about voltmeters, and were able to draw appropriate circuit diagrams. Most were not, which means that (a) our study of that topic last year did not have lasting outcomes, and (b) I wasn’t scaffolding the circuit-building well enough.
Worse, though, a design flaw popped up that took me a while to diagnose and repair. If you apply the potential difference to the vacuum tube backwards, and turn the potentiometer, you’ll get a variety of currents. When the potential difference is zero, the current is nearly zero, and when the potential difference is increased, the current increases as well. To the students, this behaviour didn’t seem abnormal. Thus, I had to check with the groups one at a time, check that they’d wired up correctly, and do a bunch of plugging/unplugging if they had it backwards. Murphy’s Law stepped in here, and so I ended up spending a couple minutes fixing all of the apparatuses before we could even begin experimenting.
I ought, thus, to have labelled the tube stands with + and – signs, and asked students to ensure they were connecting things correctly.
By this point, of course, the lab had come off the rails. The experiment was no longer theirs, and after asking the students to sit patiently for about 20 minutes while I troubleshot their circuits, behaviour issues were cropping up too.
The goal of the lab is to get a graph of electron energy (ie: 1.6e-19 C times the stopping potential) over the frequency of light that is causing the photoelectric effect. This graph will have a negative vertical axis intercept representing the work function of the metal (ie: the amount of energy required to liberate the valence electron) and a slope representing the amount of energy a quantum of light will have, for each Hz of frequency (ie: Planck’s constant).
LEDs work quite well for this, since they emit a fairly narrow range of wavelengths. I made some multi-LED sources using surface-mount super-bright LEDs. I first laid down some adhesive copper tape, then super-glued the LEDs in place. Soldering wasn’t too bad, but problematically the plastic base started to melt if I wasn’t really quick with the soldering gun. A 6-position rotary switch allowed the students to choose between LEDs, and I used 5V wall-wart adapters for this (my love affair with 9V batteries aside, I really don’t like using batteries in the lab).
Easier is to get a selection of through-hole LEDs and 3V coin-size batteries. You can hold the battery between the leads of the LED and make a complete circuit by holding the LED leads against the battery faces with thumb and forefinger.
I had some blue and near-UV LEDs that my students used in this manner, and they didn’t have much trouble with it. I’d suggest making a small hole in the side of a cardboard box for this approach. The cardboard box covers the vacuum tube, blocking out stray light.
The weakest students had trouble figuring out the frequencies of light from the wavelengths, so I sat down with the worst offender and we worked it out; I asked him to write his answers on the whiteboard at the front.
Putting It Together
We had some trouble with LoggerPro. The students typed in their numbers in the format 1.3*10^(-19), but LoggerPro interpreted those as text labels rather than numbers. Worse, when they tried to correct the problem, by typing 1.3e-19, the cell was replaced with 0. Surely that cannot be correct! (It is.) I did a second tour of the room, showing each group about this strange notation, and the strange behaviour from LoggerPro.
At this point, it became clear that a lot of the groups had been quite careless with their data collection. Their points had a lot of scatter, and some had points that couldn’t have been correct. I noticed two differences between how I conducted the experiment, and how they did:
- I was careful to zero out the current with the light source off, then turn the LED on, and only then begin to increase the potential difference. The students usually didn’t check their light shields, and didn’t return the potential difference to zero between trials.
- I increased the potential difference slowly, carefully, and deliberately. The potentiometers are sensitive enough to get 0.01 V accuracy, which I could justify in my own trials of the experiment, but the students were not careful enough to get consistent and reliable stopping potentials.
We whiteboarded our results. Three groups had unusable data: two because of spurious results, and the third for reasons I cannot fathom but surely related to their being “done” with data collection after about 3 minutes (my suggestion of multiple trials was not adopted — this is probably related to the breakdown of discipline I mentioned earlier). Of the others, all but one made fairly-obvious calculation errors.
The single group with decent data and error-free analysis came up with a result of about 4e-34 J s, which isn’t bad — but not nearly as good as the robust result I was able to get when I did the lab myself, using the same equipment, of 6.3e-34 J s.
Summer rustiness? Surely. A tricky lab that needed more scaffolding and support. Yep. A lab that allowed us to see that E = hf and that light comes in quanta, without needing a teacher to point and explain? Nope.
Overall, my attempt to use this lab as a paradigm to anchor our unit on modern physics was not successful. We spent about 3 hours developing it, and yet came away with only a sketchy model. After our whiteboard session, I needed to decide between two alternatives: re-do the lab, hopefully getting better results and accept a week-long setback, or move on and try to use the Bohr atom as a model instead. There was so much frustration about the lab, and our IB curriculum is so unforgiving, that we could only move on.
This lab could work. It could be really good. It brings together ideas from earlier studies of electricity and waves, and it provides a clear basis for further studies of atomic physics. However, if you decide to use it, go slow, be deliberate, and ensure the students do the same.
Edit: check out the good comment by Andy, below. On Twitter, Frank asks, “Have you tried using the PhET simulation on the photoelectric effect? Could possibly using in tandem with hand-on lab, similar to how folks use the circuit sim in lab.”