Results of Surface Analysis of RPC Glass

I built two 63.5cm x 63.5cm RPCs. RPC #1 was used as a control. RPC #2 was operated with high concentrations of water vapor in an attempt to duplicate our experience with the KLM modules at KEK. The results of this test are given in water test I succeeded in reducing the efficiency of RPC #2 to essentially zero and it showed no signs of recovery after nearly four months of flowing dry RPC gas. I measured a water vapor concentration of about 50 ppm in its exhaust at the end of the test.

I autopsied both RPCs and observed a few brown spots on the anode of RPC 1 (the control) and brown deposits on much of the surface of the cathode of RPC 2. I took samples of the anodes and cathodes to the surface analysis lab in the Virginia Tech Chemistry Department.

The Electron Surface Chemical Analysis (ESCA) involves irradiating the surface with a 1253.6 eV x-ray beam and measuring the energy of the electrons emitted from the surface. The depth of measurement is 50 angstroms. Spectrographs are shown below for six samples. I'm told that the height of the peak above background is important but that the background level is "noise" and not necessarily reproduceable. Multiple sweeps are done to reduce the noise. The vertical scale is N(E)/E, the number of electrons detected at a given energy divided by the energy. The area sampled is 1mm x 3mm.

Sample A (Unused glass surface with orange glow under UV light):

Sample B (Unused glass surface with white glow under UV light):

Sample C (Undamaged RPC #1 anode):

Sample D (Undamaged RPC #1 cathode):

Sample E (Damaged RPC #2 anode):

Sample F (Damaged RPC #2 cathode):

Measurement of the area of the peaks gives the following percentages for elements in each sample...

Note the carbon measured on all samples.

Tin surface

I was surprised to learn that the surface which glows orange under a UV lamp is NOT the surface with tin. In fact, the surface which is milky white under UV light shows clear tin peaks (Samples B and F). We built the barrel RPCs with the orange glowing surface inside believing that to be the tin surface. I intended to build the test RPCs the same way but apparently I managed to flip the sheet of glass used on the cathode for RPC #2 because it shows tin peaks and I checked it with a UV lamp and it is milky white.

Observations:

Fluorine is seen on the surface of both the control and test RPCs and is concentrated more on the anode than the cathode.

There is a high percentage of carbon measured on all the samples even the unused glass and relatively less carbon on the anodes of both RPCs.

The cathode of the damaged RPC has a significantly higher rate of emission of electrons than any of the other samples.