Figure 34 – Layout of the experiment, the light source can be rotated around the cell. In Figure #34 I am showing the layout of my experiment. The fiber optic probes are embedded in a PVC plastic mount in holes pointed at the ferrofluid cell.
The idea behind the apparatus is that the light can only enter the fiber optical probes from a known angle. Because we know the angle of the probe in relation to the ferrofluid cell's mirror and the probe is shaded from direct light, then only light from known angles can be picked up by the spectrometer. Another benefit is that each channel can confirm the other channel's readings but at different times during the lab run because the two angles are complementary with a common light source.
The light source can be rotated around the ferrofluid cell at constant distance to the center of the cell. The light source is a consumer grade 20 watt white light LED with a flat emitting surface of about a square centimeter. This produces a cone of light of constant brightness which is good enough for doing a preliminary experiment. Ideally the fiber optic probes should be pointed at the dead center of the ferrofluid cell but my readings will only be taken at every five degrees which should mitigate any induced parallax error.
On top of the ferrofluid cell in Figure #34, you will see a polarization filter that looks like a square piece of plastic film. Notice that there’s a black dot on the right rear corner. The polarization filter is a commercial grade filter sold in sheets. The polarization direction of the plastic sheets is not marked therefore I will call the direction of the filter shown in the figure as Polarization One and if we rotate the black dot 90 degrees to another corner of the ferrofluid cell, that is defined as Polarization Two.
Figure 35 – Spectra readings of the light source, notice the 454nm signal. In Figure #35 I am showing the spectrometer readings of the light source, notice that spectrum is a bit uneven for a product sold as a white light and has pronounced shoulders around a 454nm peak. The blue and green lines of the figure show the light source readings after passing through the polarization filter.
Figure #36 is showing the readings for different incidence angles of the light source. One may wonder how I got readings at exactly 90 degrees? The answer is that I took a bunch of readings and the one with the highest amplitude was defined as being at 90 degrees of incidence. We can do this if we assume the aluminum surface mirror will produce a larger signal versus the scatter contributions of the ferrofluid layer.
Speaking of signal, where the heck did the 454nm signal go in Figure #36? We know the light source is producing a lot of 454nm photons but ferrofluid cell isn’t scattering them back to the fiber optic probe. When I saw this reading, I stopped the experiment and started checking all my connections and took different readings of the polarization filters and the light source , and verified everything I could verify. The readings were the same with both spectrometer channels.
Figure 36 – Scatter readings at different angles for Polarization One. During the process of looking for the reason for the missing 454nm photons, I tried the Polarization Two position and found my photons! Figure #37 shows that the Polarization Two filter allows 454nm photons to be scattered into the spectrometer.
I am only showing eighteen readings in this section but the second spectrograph channel showed the same effect; one polarization position produced a 454nm signal and the other polarization position did not produce the same signal, as shown in Figure 40.
Figure 37 – Scatter readings at different angles for Polarization Two.
When I am taking these readings, literally nothing moves other than moving the polarization filter. In Figure #38 I am showing the readings for 90 degree incident angle spectra with and without the polarization filter. Clearly with no filter and with Polarization Filter Two we have the 454nm signal but not with Polarization Filter One.
I notice that the blue line in the Figure #38 resembles the spectrum of the ferrofluid cell with an applied magnetic field producing yellow lines from previous sections, and the green line resembles the spectra of ferrofluid cell with an applied field producing the red lines. In other words, the polarization of the light inside the ferrofluid cell is linked to the different color features that we see in the ferrofluid photographs.
Figure 38 – Ferrofluid cell spectrum readings at a 90 degree incident angle. In Figure #39 light in divided by light out gives us the spectrum scattering response of the ferrofluid plus the mirror, at a 90 degree incident angle. Basically Figure #39 is stating that the ferrofluid cell is optically active at around 445nm, and the polarization of the light is affecting the activity rates.
Figure 40 – Scattering response of the ferrofluid cell at a 45 degree incident angle In Figure #40, we are confirming the spectrum reading differences between horizontal and vertical polarization in the second data channel at a 45 degree angle of incidence.
The reason for the slightly negative numbers is do the baseline subtraction procedure for processing spectrums, and is considered normal for this type of graph. The 20 watt white light source, seen in Figure #35, is not producing much signal strength at frequencies greater than 700nm which leads to the choppy appearance above 700nm.
In an effort to see all the spectrum, I repeated the experiment using an incandescent heat lamp and the PC1000 NIR spectrometer which is shown in Figure #41. Clearly there is an difference between horizontal and vertical polarizations in the ferrofluid cell and I have confirmed the readings on three different spectrometers.
Figure 41 – NIR Scattering response of a cell at a 45 degree incident angle To put these findings in context, imagine looking at an ocean while using a pair of sunglasses and seeing a green ocean, and then tilting your head and seeing a blue ocean.