We have a closed system with reactor connected to that of a gas chromotography. From out catalyst, we expect only hydrogen peak, but there is oxygen and nitrogen peak detected also. The main issue is the area of the peaks keep increasing drastically for each run. We do not inject any gas manually, so that we can reduce the manual errors, but still we are facing issue with understanding from where these two peaks are coming from. If there is any leakage, we should be seeing change in the pressure inside the reactor, but there also it is not increasing drastically.