PMT lives matter!

Most laser scanning microscopes use PhotoMultiplier Tubes ( PMTs ) to detect the emitted photons of biomarkers. PMTs are required as the markers are weakly-emitting signals and living tissue has high scattering parameters.

The detection path needs to be optimized by the microscope manufacturer, meaning  the detection path should be minimal in length and “large” in the collection. In most cases there are multiple channels for detection. Different dichroic and filter sets need to be further optimized to increase the quality of the signal.  There are no perfect filters, but they definitely play a role in PMT protection.

Biomarkers' emission spectra usually are in the visible range. This is the primary reason for PMT photocathode selection is important. The photocathode radiant sensitivity should be higher on those wavelengths where the signals are expected. Below are two figures showing why GaAsP PMTs are the preferred option on the market today.

Other Multialkali PMTs also cover this range and even wider wavelength ranges but the radient sensitivity is significantly lower, between  - 450 – 700 nm. GaAsP photocathode provides the most in sensitivity for neuroscience but there are many other components of a good, clear image.

In two-photon imaging we sometimes hear the term that PMTs are consumables. From one perspective it is true. PMTs are like pens; after a long time of use they get blunt, meaning their sensitivity won’t always stay as high and the images will get dimmer.

Following are some examples of what we can do to get the most out of our PMTs:

First, as the GaAsP PMT sensitivity is in the visible range all two-photon experiments should be done in minimal (ideally 0 ) surrounding light. Sometimes it is challenging with existing conditions but if we have a chance to design the room for that specific area, minimize the room light reaching the system.  In this case the screens of the monitors where we work also add to the room light.

Currently PMTs have their own safety system built in that shuts down if strong room light is detected. Always use PMTs which have this functionality. Strong light can also come from an extremely bright spot in the tissue but it would be so extreme we need to check what is happening. To sum it up, it is a really useful feature.

Second, try to use some sort of shading box. There is a good example on our site with multiple useful features beyond the fact that it would further protect our PMTs, such as noise cancellation, which can be important for behavioral studies. It can also serve as a Faraday cage for electrophysiology.  It is highly recommended to ensure the operating workstation won’t affect the lifetime of PMTs.

Third, if it is possible we should cover the space under the objective. It proves important when we place screens inside the shading box so we can use visual cues for the animal. There are some good hints from labs on keeping the front aperture covered from the screen light.

In addition we also need to think about photostimulation experiments. If it is a 2p photostimulation it is less of a problem as 2p wavelength range is out of the photocathode wavelength range, above 700nm, than if it is a continuous 1p laser source or a full field LED. If experiments include similar photoactivation you must have a gating feature in your PMT circuit. During the stimulation we need to protect our PMT, which can be done with a shutter. But as mechanical shutter movements are slow the number of experiments that can use this method is limited. Timing and triggering are very important factors when we do such an experiment; most of our measurements in neuroscience require higher speed.  If our shutter would be liquid crystal it would solve the timing problem, but it is not yet practical. The most commonly used solution is gating. The PMT is electronically switched off so there is no output during the stimulating period. We need to know the photons still reach the photocathode in this case. Furthermore it is not well known, but PMT lifetime depends more on the dynode section than the cathode part.  Performing high-speed gate operation requires high-speed triggers. This switching can cause noise, but this can be extracted from the measurement easily.

Last, let’s examine the ideal gain setting on PMTs in order to have the best SNR during measurements. Gain is usually defined in percentages. The gain can be adjusted depending on the signals. It is not recommended to use the PMT gain close to its minimum or maximum. Our experience shows the best solution is to keep the gain between 50 –  90% and rather play more with the laser intensity. By knowing our signal and the effect a laser can cause, we suggest using less laser power and set the gain a bit higher.  

It is also good to know about the stability of the PMTs. Stability mainly depends on supply voltage, current and ambient temperature. Unstable measurement can be caused by the secondary dynode section after a longer period of measurement time.

Warm-up time before actual measurements is important to maintain stable signals and longer lifetime. At the beginning it should be between 30-60 minutes but this can be shortened after many operations. Still always take at least 10 -20 minutes even if you select lower anode current.

These are our suggestions for maintaining the health and life of the PMTs in your system. They are not that cheap, but always be sure your supplier holds available stock of PMTs. New GaAsP PMT lead time can be months if they are not in stock. It is important to minimize the downtime if your PMT becomes damaged or wears out.

Resource:

Photomultiplier tubes - Basics and Application ( Third edition (Edition 3a ) by Hamamatsu.