Adsorption/desorption experiments:
Tests of the effects of adsorption and desorption were carried out on different surfaces of materials widely used for the preparation of reference gas mixtures (e.g. PFA, PTFE, stain-less steel). In a second step, the influence of humidity on the adsorption of NH3 in the preparation of reference gas mixtures was assessed. It was concluded that adsorption on stainless steel surfaces in contact with NH3 can be minimised by up to 90 % when a silica-based coating (SilcoNert2000®, SilcoTek Inc.) is applied on gas-wetted surfaces. The results were implemented in the construction of the mobile reference gas generator ReGaS1 and in the dynamic generation of reference gas mixtures. Thus, SilcoNert2000® has been applied to all stainless steel surfaces in contact with NH3 as adsorption on coated surfaces are reduced by 90 % compared to uncoated stainless steel surfaces. The comparison of polymer (PFA, PTFE) to uncoated stainless steel surfaces yields similar results.
Adsorption of NH3 is reduced with increasing levels of humidity yet, at ambient levels of humidity, NH3 is dissolved in water H2O. Therefore, it is advised to use dried air or nitrogen N2 for the generation of reference gas mixtures, despite the fact that for spectroscopic applications, humid matrix gas is preferred.
Development and construction of a mobile reference gas generator ReGaS1:
A mobile instrument for the dynamic generation of SI-traceable NH3 reference gas mixtures has been planned and realised (ReGaS1). ReGaS1 is a development beyond the state of the art as it allows for the on-site permeation and dynamic dilution over two dilution steps and thus for the dynamic generation of NH3 at ambient amount fractions (0.5 – 500 nmol mol-1) with UNH3 ≤ 3 % (k = 2). Validation test have been conducted and the characterisation results are very promising.
Stability of temperature and pressure yield a constant mass loss of the permeation unit (UqmC ≤ 1 %, (k = 2). Mass flow rates are modifiable within seconds and thus allow for readily changing the generated amount of substance fractions over two dilution steps covering the entire range of atmospheric amount fractions (0.5 – 500 nmol/mol) with UNH3 ≤ 3 % (k = 2). The stabilisation time required to measure stable values is primarily depending on the adsorption on the surfaces of the comparator to be calibrated.
The prerequisites for ReGaS1 in terms of amount of substance fractions as well as uncertainty (0.5 – 500 nmol/mol at UNH3 ≤ 3 %, (k = 2)) could be met. The user feedbacks were positive throughout with particular emphasis on the user friendliness and the vast flexibility in the generated amount fractions.
Good Practice Guide:
The document is going to be compiled and made available to end-users with a delay of 4-5 months in the course of summer 2017.