Microchannel plate (MCP) PMTs are very attractive photon sensors for low light level applications in strong magnetic fields. However, until recently the main drawback of MCP-PMTs was their aging behavior which manifests itself in a limited lifetime due to a rapidly decreasing quantum effciency (QE) of the photo cathode (PC) as the integrated anode charge (IAC) increases. In the latest models of PHOTONIS, Hamamatsu, and BINP novel techniques are applied to avoid these aging effects which are supposed to be mainly caused by feedback ion impinging on the PC and damaging it. Since more than four years we are running a long-term aging test with new lifetime-enhanced MCP-PMT models by simultaneously illuminating various PMTs with roughly the same photon rate. This allows a fair comparison of the lifetime of all investigated MCP-PMTs and will give some insight in the best techniques to be applied for a lifetime enhancement. In this paper the results of comprehensive aging tests will be discussed. Gain, dark count rate and QE were investigated for their dependence on the IAC. The QE was measured spectrally resolved and as a function of the position across the PC to identify regions where the damage develops first. For the best performing tubes the lifetime improvement compared to former MCP-PMTs is a factor of ~50 based on an IAC of meanwhile >10 C/cm2. This breakthrough in the lifetime of MCP-PMTs was achieved by coating the MCP pores with an atomic layer deposition (ALD) technique.
For the identification of low momentum charged particles and for event timing purposes a barrel Time-of-Flight (TOF) detector surrounding the interaction point is planned for the PANDA experiment at FAIR. Since the boundary conditions in terms of available radial space and radiation length are quite strict the favored layout is a hodoscope composed of several thousand small scintillating tiles (SciTils) read out by silicon photomultipliers (SiPMs). A time resolution of well below 100 ps is aimed for. With the originally proposed 30 x 30 x 5 mm3 SciTils read out by two single 3 x 3 mm2 SiPMs at the rims of the scintillator the targeted time resolution can be just reached, but with a considerable position dependence across the scintillator surface. In this paper we discuss other design options to further improve the time resolution and its homogeneity. It will be shown that wide scintillating rods (SciRods) with a size of, e.g., 50 x 30 x 5 mm3 or longer and read out at opposite sides by a chain of four serially connected SiPMs a time resolution down to 50 ps can be reached without problems. In addition, the position dependence of the time resolution is negligible. These SciRods were tested in the laboratory with electrons of a 90Sr source and under real experimental conditions in a particle beam at CERN. The measured time resolutions using fast BC418 or BC420 plastic scintillators wrapped in aluminum foil were consistently between 45 and 75 ps dependent on the SciRod design. This is a significant improvement compared to the original SciTil layout.
The hadron identification in the PANDA experiment at FAIR will be done with DIRC detectors. Because of design and space reasons the sensors of the DIRCs have to be placed inside the strong magnetic field of the solenoid. As the favored photon sensors microchannel-plate photomultipliers (MCP-PMTs) were identified. However, these devices showed serious aging problems until very recently, which manifest themselves by a fast degrading quantum efficiency (QE) of the photo cathode (PC). This is mainly due to feedback ions from the residual gas. In this paper we discuss the recently accomplished huge improvements in the lifetime of MCP-PMTs. With innovative countermeasures applied to the MCP-PMTs in the attempt to reduce the aging effects the manufacturers were able to increase the lifetime of MCP-PMT prototypes by almost two orders of magnitude compared to the former commercially available devices. Our group has studied the aging of MCP-PMTs for more than four years by simultaneously illuminating different types of lifetime-enhanced MCP-PMTs at the same photon rate. Gain, dark count rate, and QE as a function of the wavelength and the PC surface were measured in regular time intervals and studied in dependence of the integrated anode charge. We observe that MCP-PMTs treated with an atomic layer deposition (ALD) technique are by far the best devices available now. A lifetime of up to 10
C/cm2 integrated anode charge was reached with these sensors. This is sufficient for both PANDA DIRCs.
The charged particle identification at the PANDA experiment will be mainly performed with DIRC detectors. Because of their advantageous properties the preferred photon sensors are MCP-PMTs. However, until recently these devices showed serious aging problems which resulted in a diminishing quantum efficiency (QE) of the photo cathode. By applying innovative countermeasures against the aging causes, the manufacturers recently succeeded in drastically improving the lifetime of MCP-PMTs. Especially the application of an ALD coating technique to seal the material of the micro-channels proves very powerful and results in a lifetime of ~6 C/cm2 integrated anode charge without a substantial QE degradation for the latest PHOTONIS XP85112. This paper will present a comparative measurement of the lifetime of several older and recent MCP-PMTs demonstrating this progress.