Scanning Tunnelling Microscopy (STM) and X-ray Photoelectron Spectroscopy (XPS) have been used in this work at various temperatures to study the mechanism of NOx storage reaction using model catalysts based mainly on the li02 (110) and the Pt (111) surfaces. The metals Pd and 8a were deposited using metal vapour deposition (MVD». The key findings of this work are summarised below. Prior to the investigation of the 8a/Pdm02 (110) model catalyst, the NOx reactivity with the clean li02 (110) surface and Sa reactivity with li02 (110) were studied. NO and N02was adsorbed on the clean li02 (110)-(1 X2) surface in the molecule state at room temperature and dissociated 'at 373 K in the XPS. However, the amount of adsorbed NO or N02 was small, - less than 0.1 ML, and there was no evidence of or~ered structure of any reactivity between NO and li02 (110) in the STM. Upon increasing the·annealing temperature, liN was formed at 873 K by reaction of dissociated N with interstitialli3 + diffused from the bulk. On the other hand, for a small amount of 8a deposition on the li02 (110) surface, the surface was disordered, but rows of 8a were seen running in the [001] direction of the li02 substrate. Upon sintering at 1073K in UHV, a (2x2) pattern was seen in LEED that originated from the Sa, although no ordering could be seen with STM. NO and O2were adsorbed on the Pd/Sarn02 (110) model catalysts, which was prepared by the 8a deposition and then Pd deposition at 673 K. However it was less reactive due to a low NO sticking probability and it was impossible to obtain atomic resolution images of model catalysts. A new approach, using inverse catalysts, was adopted, that is, Pt (111) was used as the support and Sa was deposited at room temperature. A locally ordered (2x2) structure was obtained. In the case of the annealing temperature at 1273 K, the variety of structures were formed, which these structures might be Sa overlayer, Sa/Pt alloy and BaO. One of these structures is the banded zig-zag structure on the terrace, with the unit cell can be defined as (8 ~ J. After the introduction of O2 at· 573 K, large scale images of BaO were obtained. The average spacing is - 8.6 A, twice that expected for the (111) plane of BaO, which is due to reconstruction, resulting in the formation of (2x2) structure. There was a metastable state of a top layer of BaO/Pt (111) model catalyst at 573 K in the presence of O2, which is likely to be due to the formation of Ba02 confirmed by the atomic structure of Ba02 in the STM. Upon dosing NO and O2 at 573 K, the BaO particles grew and some growth was confirmed at step edges, especially O2 rich state. This effect is observed by an approximate increase in particle volume of 100%, which is consistent with about half of the oxide being converted to the nitrate. It is therefore postulated that a film of the nitrate effectively encapSUlates the oxide. This surface is unreactive to S02' However, when both S02 and O2 were co-dosed, the , ; atomic structure of BaO has rapidly disappeared which might be converted to the SUlphate, 8aS04, Which would result in poisoning of the real catalyst.