Can a discrete dynode mass spec detector get "tired?"

Question

In an electron multiplier (discrete dynode detector), one electron triggers the release of more electrons in a cascade.

Is it possible that a "large" number of electrons hitting the detector can temporarily (course of hours) reduce the number of electrons available to amplify future incident electrons? (E.g. is there a cumulative recharge time?)

Or is it possible for the anode to have temporarily reduced sensitivity if the electrons are not flushed quickly enough?

(Assume real-world, non-ideal physics. This is an actual mass spec we're talking about. Something is causing reduced sensitivity over the course of hours of usage, aside from reduced transmission and ionization due to dirt.)

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user ZachB

Comments

Short answer is "yes", all cascading detectors can have a reduced efficiency or gain immediately after a large response. Measuring and compensating for these effects, however, is difficult. My copy of Leo has gone missing (i.e. been borrowed without bound) or I'd find you a reference.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user dmckee

---

@dmckee Thanks. Estimate of order of magnitude of time? What is "large?" My best estimate of our "large number" is a few million.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user ZachB

---

Timing depends on the current draw of the detector and the pure number of electrons moved (you need to supply enough "new" electrons to get things back to "normal". It's usually pretty short.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user dmckee

---

@dmckee I think I found your reference: §5.5 (General Characteristics of Detectors: Response Time) of W.R. Leo. That discusses extensible and non-extensible dead time. Are you suggesting here that we're observing a gradual (~hours) reduction of the observed count rate due to a mixed ext and non-ext dead time model? (The section makes it sound like something is either one or the other of those models.)

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user ZachB

---

If you are seeing a change over the course of hours than it is not the effect I had in mind: that would appear on short time-scales (typically considerably less than 1 second). On the other hand, their are hours- or days-long effect in PMTs. If they have been exposed to daylight, they can have reduced efficiency for the first few days after they are put back in the dark. It is possible that a sustained high draw could cause something similar, but I don't understand the origin of this effect so I can't guide you.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user dmckee

Answer 1

An "hours long" gradual reduction in gain smells like accumulated contamination from something in the chamber. As an extreme example, silicone-based pump fluids will kill an electron multiplier; other materials can be removed by cleaning and baking. O'Hanlon's A User's Guide to Vacuum Technology has some discussion in the RGA chapter.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user Art Brown

Comments

Ah...looking at the rest of the apparatus. Good call, that.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user dmckee

---

Very good point here, a good reminder to constantly check the rest of the equipment.

This post imported from StackExchange Physics at 2014-03-24 04:19 (UCT), posted by SE-user user29350