Various Monochromator Measurements

Five different LEDs are measured using an Oriel 74000 monochromator. Each LED is aligned such that the monochromator input lies inside the viewing angle of the LED. The LED model numbers are given in table 1 along with their expected and observed peak emission wavelengths. Every time the LED was switched, the monochromator was power-cycled in order to maintain calibration. Data was collected in 2 nm steps from 200 nm to 800 nm.

Figure 1: Monochromator setup

Figure 2: UV LED power spectrum
Observed	Expected
	[none found]

Figure 3: Blue LED power spectrum
Observed	Expected

Figure 4: Green LED power spectrum
Observed	Expected

Figure 5: Yellow LED power spectrum
Observed	Expected

Figure 6: Red LED power spectrum
Observed	Expected

A xenon filament lamp is used as a broad spectrum light source and is focused at the Oriel monochromator. A calibrated Newport photodiode (model #818-ST2-UV-DB) is used to find the emission spectrum of the lamp, as seen in figure 8. The photodiode is then replaced with an R1332 PMT supplied with a -2000 V bias. The current of this PMT is then divided by the photodiode wattage measurement to obtain the responsivity in A/W for a given wavelength, as seen in figure 9.

Figure 7: Responsivity curve (blue) for calibrated photodiode

Figure 8: Spectral density of white light source

Figure 9: Responsivity curve for R1332 PMT with -2000 V bias

The UV LED whose spectra is shown in figure 2 is used to illuminate a polystyrene sample (JPB 1-72-1) doped with wavelength shifters PPO and POPOP. The sample is placed just outside the input to the monochromator so its emission spectrum can be measured. The resulting spectrum is seen in figure 10. Compared to the LED spectra, the scintillator is dimmer and its light is emitted isotropically, so the measured spectrum has a much lower total power. A possible peak from POPOP emission can be seen between 400 and 450 nm, whereas no peak can be seen in PPO's emission band between 325 and 410 nm. The UV LED's emission spectrum falls within the absorption region of POPOP but not PPO, perhaps explaining the missing PPO peak.

Figure 10: JPB 1-72-1 emission spectrum

A different sample with a blue-to-green wavelength shifting fiber in its center is machined so that a UV LED can be embedded in the scintillating material. Two spectra are measured, one shown in figure 11 with both the WLS fiber and scintillator light entering the monochromator, and another shown in figure 12 with only the WLS fiber light entering the monochromator.

Figure 11

Figure 12

The WLS fiber has a peak emission wavelength of 494 nm, and this matches the peak seen in figure 12. The expected POPOP emission spectrum is more complicated, but it is concentrated around 375-450 nm, which generally agrees with our observation.

The same scintillator sample with embedded WLS fiber is held in a fixed position in order to directly compare unmasked and unmasked sample spectra. The sample position between runs is maintained by aligning the WLS fiber with two crossed laser beams held by a ring stand. The results are given in figure 13 along with the difference between the masked (fiber-only) and unmasked (entire sample) spectra.

Figure 13

If we approximate the area under the “masked” and “difference” curves in figure 13 with Riemann sums, we find that the masked-difference ratio of the areas is 0.0667, i.e. an energy loss of 93.33%. Previous estimates put the energy loss from scintillator to fiber at 68.3 ± 5.0%, but the previous method relied on a PMT with unknown window material / quantum efficiency.