In many scientific and medical applications, it is essential to detect low intensity signals down to a single photon. For this purpose, PHOTONIS have developed a range of photon counters, including Hybrid Photo Diodes (HPD) and Multi-MCP photomultiplier tubes (PMT).
The photocathode can be regarded as the first stage of the amplifier, so its properties are highly important to understand the overall performance of the device. The quantum efficiency (QE), response time, and dark rate directly influence the properties of the device. For example, the photon detection probability is predominantly characterized by the QE of the photocathodes.
The dark electrons emitted by a photocathode are amplified in the same way as a photoelectron, so it is not possible to seperate dark pulse from the real single-photon event. In most cases the timing jitter of the devices is limited by the traveling time of a photoelectron in a cathode before escaping into the vacuum. This article discusses the photocathode properties and demonstrates the overall performance of a PHOTONIS dual-MCP PMT using new Hi-QE photocathodes.
Quantum Efficiency Spectra
The quantum efficiency spectra for traditional and Hi- QE series of S20 photocathodes within the spectral range of 200 - 700 nm is shown in Figure 1. The peak QE of about 25% is achieved at 270 nm for the std-S20 process.
These photocathodes have a wide sensitivity spectral range, with a QE to be still in the range of 3 - 4% at 700 nm (Figure 1). However, the wide sensitivity range is at the expense of lower QE at the peak and higher dark rates.
It should be noted that fast gating photocathodes can be offered with mesh underlay to allow the gating time to be reduced to 3 ns, but at the same time it also reduces the averaged QE by about 10% - 15%.
Figure 1. Spectrum of Quantum efficiency for newly developed Hi-QE photocathodes: Hi-QE-UV (dark blue), Hi-QEblue (light blue) and Hi-QE-green (green) in comparison to Standard S20 photocathode (dashed red). Image Credit: Photonis
PHOTONIS developed the Hi-QE series of photocathodes targeting specific spectral ranges based on customer requests. Hi-QE photocathodes were developed on a fused silica input window with a high-energy cutoff at 170 nm. These photocathodes demonstrate much higher QE values (>30%) in the spectral ranges of interest (Figure 1) as well as showing extremely low dark rates (Figure 2).
The response time is also much lower than 100 ps. Hi-QE-UV photocathodes are designed for UV range with an optimum QE at 270 nm typically in the range of 31% to 34%. It is also possible to grow these photocathodes on sapphire cathode substrates, which allow extending the sensitivity spectral range down to 150 nm.
PHOTONIS designed the Hi-QE-blue photocathode to provide the highest QE in the spectral range of 260 - 410 nm. In this range, the QE spectrum displays a plateau with a typical QE higher than 30%. Below 260 nm, the decrease of QE (compared to Hi-QE-UV) is at the expense of high sensitivity in the blue spectral range. Hi-QE-green photocathodes show very high QE value higher than 30% in the 390 - 480 nm range. However, the value of the QE is around 25% at 500 nm.
Hi-QE-green has much higher sensitivity at a longer wavelength of up to 700 nm when compared to other Hi-QE photocathodes. It should be noted that the dark rate of these photocathodes remains extremely low like other Hi-QE cathodes, despite their high sensitivity at longer wavelengths. This aspect makes Hi-QE-green a unique photon counting device in this spectral range.
Dark Rate of Hi-QE Photocathodes
Low dark currents are a typical characteristic of the high band gap photocathodes. However, lowering the dark rate to nearly zero is important for low-rate single photon detection.
The evolution of dark rate vs. time at room temperature for std-S20 and newly developed Hi-QE S20 photocathodes is shown in Figure 2. Here, the cathode is positioned at time zero in dark conditions.
The dark rates are the same for all Hi-QE S20 series photocathodes. The low-rate plateau is reached after two to three hours, as shown in Figure 2. The high dark rate measured initially seems to arise from the population by ambition light of long-living surface and bulk states, lying above Fermi level.
Figure 2. Evolution of dark rate vs. time at ~23°C for standard S20 photocathode (blue) and new developed Hi-QE S20 (red). Extremely low dark rate < 30 cts/cm2 can be achieved with Hi-QE series photocathodes. Image Credit: Photonis
More time is required for these states to be discharged, with a decay time to be also an important parameter of the detector performance. The growing photocathode process was adjusted for Hi-QE photocathodes to maintain low dark rate and rapidly discharge surface and bulk states. The dark rate is close to 1000 - 2000 cts/cm2 for traditional broad-range S20 photocathodes, whereas the Hi-QE photocathodes typically have a dark rate of only 20 - 30 cts/cm2. Here, the dark electron rate lowers below 50 cts/cm2 after 10 minutes in the dark.
Single/Multi Photon PHD Measurements with Hi-QE Dual MCP-PMT
Pulse Height Distribution (PHD) obtained with dual MCP-PMT and Hi-QE photocathode is shown in Figure 3, demonstrating the ability of single photon detection and also explaining the method of measuring the dark rate.
A Charge Sensitive Preamplifier CSP10 (1.4 V/pC), a multi-channel-analyzer MCA3 (scale = 0.89 mV/chn), and a shaping amplifier CSA4 (gain = 10, shaping time of 250 ns) were used to record the PHD. The threshold value set was 32 chn. The dark-rate measurements were carried out by placing the tubes into dark conditions and setting the MCP-voltages to obtain a gain of 1-2E05. The next step was to measure the count rates vs. time.
Figure 3. Pulse Height Distribution recorded with PHOTONIS’ dual MCP-PMT and Hi-QE S20 photocathode with single photon illumination (blue); Gaussian curve (black) is a fit of experimental results. Image Credit: Photonis
Figure 3 shows the PHD measured with low input background light illumination by maintaining the count rate down to a few hundred Hz. This example applied the MCP voltage of 1625 V (for dual set) to obtain an electron gain of 1.07E5. For the threshold used, the photocathode dark rate was found to be 30 cps/cm2 and the MCP dark rate was below 0.2 cps.
The peak is well separated in the PHD shown in blue curve in Figure 3, with a low-energy valley and noise to be less than the threshold. The peak agrees well with Gaussian (solid black curve).
The gain point (“G”) signifies the mean energy of the PHD, and it is slightly above the PHD peak position. Photon counting tubes are typically characterized using the peak/valley (P/V) ratio and the full width half maximum/gain (W/G) ratio. The measured values P/V≈6 and W/G≈0.86 are ideal for dual MCP PMTs.
Figure 4 shows another example illustrating the superior performance of PHOTONIS’ MCP-PMT. Here, a 100 kHz short-pulse (100 ps) defocused laser beam was used to illuminate the photocathode. The significant reduction of the laser intensity allowed the bunch of photoelectrons that escaped into the vacuum to have only a few electrons.
Image Credit: Photonis
Figure 4. Pulse Height Distribution recorded with PHOTONIS’ dual MCP-PMT and Hi-QE S20 photocathode (the same tube and settings as in Figure 3) with few photons illumination. The fitting curve is a sum (1->7) of Gaussian curves, corresponding multi (1->7) photoelectron amplification. Inset: the normalized area of each (1-7) Gaussian curve (blue dots). The red curve is calculated Poisson distribution with a λ≈2.52 to be expected value for the average number of electrons emitted from the photocathode.
The same tube and settings as shown in Figure 3 were used to perform the measurements. The black curve represents the fit, and the blue curve represents the measured PHD. The fit is a sum of seven Gaussian curves (Figure 4). The position (G1) and width (W1) of the first fitting Gaussian curve, relating to a single-photon illumination, were obtained from Figure 3.
Other Gaussian curves relative to amplification of 2,3…7 photoelectrons released by the photocathode. The positions of these curves were set as GN=G1*N, followed by scaling up the width in accordance with the statistical rule as WN=W1*N1/2. The fitting parameters were merely the amplitudes of the Gaussian peaks.
A very good fit of the measured curve can be found with clear separation of the first and second peaks, relating to single and two-electron photoemission. It is also possible to observe a peak for 3 electrons photoemission. However, the resolution of further peaks is not good enough because of the statistical increase of the PHD width for short-pulse photoemission of multi electrons.
The insertion (blue points) in Figure 4 shows the intensity (corresponding to the area) of each peak. The red curve is a Poisson distribution fit with expected value λ to be 2.52. The good agreement of the data corroborates that the average number of photoelectrons in the bunch emitted by the photocathode is around 2.5 for these measurements.
PHOTONIS developed high quantum efficiency UV, blue and green S20 photocathodes with QE higher than 30% in specified spectral ranges. With rapid response time (well below 100 ps) and very low dark count rates (down to 30 cts/cm2), these Hi-QE photocathodes are ideally suited for photon counting devices. The high quality of MCP-PMT PHOTONIS’ photon counters has been demonstrated by measuring PHD with a single photon and with a few photon illuminations.
This information has been sourced, reviewed and adapted from materials provided by PHOTONIS Technologies S.A.S.
For more information on this source, please visit PHOTONIS Technologies S.A.S.