GTS meeting 7 05 16

• We have tried to reach 1 mA (300kV) stable running, but the gun HVPS trips off on current a few minutes later, max time is about 30 minutes, see link here: [1]
  • We think is mainly due to halo since the entire photocathode is activated.
  • We have tried to see if vacuum, or biasing the anode at +430V improves stability, but did not seem to help much. What it seemed to help is to increase the beam current SLOWLY.
  • Next is to activate a photocathode using a mask to limit emission area and therefore reducing halo.
    • HOWEVER, it has been very very difficult to activate photocathode in this chamber. Main problems: Not being able to bias the photocathode and non-adjustable distance between photocathode and chemical sources.
  • Observed trips are very similar to those observed in the FEL while running >8 mA CW, and also observed by the Cornell team
  • Trips are likely related to ions in the beam line and beam dump. Adriana has some input from papers.

• Below is a summary of what we have learned so far:

• 1. Using an inverted geometry ceramic insulator gun, and home-made multi-alkali photocathode, we have demonstrated 300 keV beam a, at 1-mA DC sustained albeit for about 30 minutes. This shows that inverted insulator guns are capable of reaching 300 kV level and produce mA level currents, therefore making a good case to upgrade the FEL gun. In summary, I believe we have demonstrated proof of principle attacking many fronts: the high voltage engineering of connecting the cable to the insulator, the electrostatic design of the shed, and the in-house multi-alkali R&D.

• 2. We have move forward the photocathode R&D led by Mamun from the test deep chamber to the gun producing actual electron beam. With this, the QE is actually higher than that estimated in the prep chamber, we believe due to space charge effects with very low bias voltage and relatively high laser fluence.

• 3. The robustness of these photocathode is truly amazing! to give you a perspective: Every time the FEL Cs:GaAs NEA cathode suffered a "damage spot event" just like the ones we observed in the GTS, it would take us nearly 24 hours to recover. In the GTS, every "damaging event" decreases the QE only by about 20%, so we keep on running beam! See the following link for QE curves after each damaging event: [2], and also here for pictures of damaged photocathode

• 4. Mamun and Yan were able to re-activate the photocathode from 0.1% QE to 8% QE just by adding more alkali while keeping the substrate at about 150C. But since the photocathode cannot be biased (we bias the anode ring instead to collect photocurrent), it takes several hours until the photocurrent finally levels off. It is very difficult to activate the cathode w/o being able to bias it.

• 5. Operating with a higher >5% QE cathode is essential to reduce laser light that in turns leads to electron beam halo.

• 6. We do not know exactly the reason for the gun trips caused by current bursts, but the events look of the same nature as those observed in the FEL gun while running >8 mA CW beam, and those observed by the Cornell team. In summary: the trips are caused by ions generated way downstream of the gun, i.e. the dump in case of the GTS, and in the 1/4 cryo-module in the FEL. The Cornell teams has several papers on this: they resolved the issue by addressing ion trapping between the gun and the dump!

• 7. Although biasing the anode seemed to help reaching 1 mA, I'm not totally convinced is the final solution. In fact, biasing the anode seemed to make a damage spot deeper than those with the anode un-biased. I think the best way to solve the problem is with a reduced active area photocathode to reduce electron beam halo, therefore reducing ions in desorbed gas in the beam line.

• 8. Fay demonstrated with beam measurements that the beam comes off the gun with a VERTICAL kick of 2 degrees. Yes, vertical, not tilted horizontal as we thought based on the fact that the electrode is rotated w.r.t. gun/beam axis. This is consistent with our observation that the damage spot is on top of the photocathode (at 12 o'clock position). Fay says this is related to the asymmetry of the electrode geometry, i.e. the insulator! When we biased the anode, the damage spot moved to the center of the photocathode. This is why I think that biasing the anode is not the ultimate solution, or at least we may have to bias at higher voltages.

• 9. The Beam dump vacuum shows classical signature of the ceramic break discharging. Is there a shield? is it placed in the correct orientation? It is evident at high currents. This is not helping with running high current in a stable fashion. In summary, I think the gun is tripping off because there is a straight shot from the photocathode to the dump, and because of the halo beam causing havoc in the beam line and the dump.

• 10. We expected to see photocurrent decaying with time, but we have not seen this so far! except with the very first photocathode that had 0.3% QE. So, no lifetime measurements yet, but Adriana has posted really cool QE graphs showing the robustness of the photocathode as a function of running conditions and damaging events.

• 11. Yan was able to take the first shot at emittance measurements. Adriana collected beam size vs solenoid strength.

• 12. Below are a few slides wooing the potential around the anode when biased at +430V and +1000V.
  • Media:Potential_Near__Anode checking biasing.pptx