Your description is a bit coarse, was your experiment like this:

-sample placed in UHV

-treatment with 98%He, 2%NH3 for a fixed time

-pumpdown

-subsequent TPD measurement

and what you want to know is the concentration, aka coverage, of NH3 on the surface after this treatment? That's not that straightforward, unfortunately. You would probably have to do a coverage series and check when a saturation is reached. That's the coverage (theta)=1. The problem here is that you may have to start at cryogenic temperatures because the saturation coverage may not be held by the substrate.

If you're looking for a concentration in the sense of molecules per area, you can prepare a material for which that is known and you can saturate it. Then you can use that as a reference concentration.

Comparing absolute intensities from tool to tool is basically impossible because geometries may vary strongly and detection efficiencies have the same problem.

The easiest way is to calibrate the TCD signal(area) by injecting small gas sample of 2% NH3 in He in your system. You can easily to calculate the amount of the concentration from the curve.

Expanding on what Sanggi Lee wrote: Using the same equipment and settings (flow rate, TCD current etc) inject a known volume of the 2% mix, or even pure NH3, into the pure helium carrier stream and record that area. Integrate areas under sample TPD peak and injection and by direct comparison calculate the desorbed amount. If using the 2%, remember to divide the injection volume by 50 to get the true volume of ammonia, and ALSO convert to STP (760 torr and 273.15K) from ambient/instrument conditions of syringe/loop to get the molar quantity.

Best way is to do the calibration injection through the sample cell while it is isothermal at the end of the TPD run. In this way the peak will be broadened in a similar way to the TPD itself, and at the exact same flow rate.

Some folk do the TPD into the 2% NH3 without changing to pure He but that displaces the TPD peaks in temperature and can make interpretation difficult, and also reduces signal height. I strongly urge the TPD be done into pure NH3. Also, always remember to "pre dry" or "degas" the sample to remove surface water to prevent the ammonia from interacting with it. Do that at as high a temperature as the solid can withstand without undergoing any other chemical change or physical/structural alteration.

Don't be surprised if the TCD baseline does not come back to a perfect zero after the TPD peak(s) when running protonated zeolites like HZSM5... the structure will start to break down and release H2O from that proton and structural oxygen.