Phase boundaries of nanometer scalec(2×2)-O domains on the Cu(100) surface

Abstract

A scanning tunneling microscope ͑STM͒ showed the formation of nanometer size c(2ϫ2) domains of adsorbed oxygen on the Cu͑100͒ surface when oxygen coverage is low, though large well-ordered c(2ϫ2) domains were not observed. The STM image of the phase boundaries of these nanometer size c(2ϫ2)-O domains showed complex zigzag bright lines surrounding the c(2ϫ2) domains, and the low-energy electron diffraction pattern of this surface gave a ''four-spot'' pattern, which can be explained by a local c(2ϫ2) domain model.

Oxygen adsorption-induced nanostructures and island formation on Cu{100}: Bridging the gap between the formation of surface confined oxygen chemisorption layer and oxide formation

“…8,58 A recent investigation also revealed that the minority sites such as vacancies provide a nonactivated adsorption channel. 59 In agreement with Fujita et al, 9 we find that upon oxygen adsorption the surface phase of c͑2 ϫ 2͒-O is converted to the ͑2 ͱ 2 ϫ ͱ 2͒R45°-O reconstruction as indicated by the depressions in Fig. 1͑d͒.…”

“…This is also in agreement with the earlier experiments by Fujita et al who observed a mixed phase already after 8 L exposure ͑1.3ϫ 10 −7 mbar for 80 s͒ but at a significantly higher surface temperature ͑450 K͒. 9 As the exposure is further increased from 551 L, the depressions grow isotropically ͓Fig. 1͑e͔͒ indicating the growth of a missing-row reconstruction at the expense of the c͑2 ϫ 2͒-O phase.…”

“…The depressions do not appear to be centered at any obvious nucleation sites and their density is ϳ0.02 nm −2 . Following the earlier studies by Fujita et al, 9 we attribute the initial roughening of the surface to the c͑2 ϫ 2͒-O phase and the depressions in Fig. 1͑d͒ to the formation of ͑2 ͱ 2 ϫ ͱ 2͒R45°-O missing-row type of a reconstruction.…”

“…8,58,59 Below 0.3 ML coverage ͑1 ML=1 monolayer= 1.53ϫ 10 15 cm −2 ͒, oxygen atoms occupy fourfold hollow sites resulting in nanosized Cu͑100͒-c͑2 ϫ 2͒-O domains. 9 At higher coverage oxygen induces a missingrow-type Cu͑100͒-͑2 ͱ 2 ϫ ͱ 2͒R45°-O reconstruction which is highly stable and, under ultrahigh vacuum conditions, leads to a saturation coverage of oxygen at 0.5 ML. 10,11 However, as our recent molecular beam experiments demonstrate, the 0.5 ML phase remains reactive towards O 2 dissociation.…”

“…The onset and growth of oxide on Cu͑100͒ and vicinal surfaces consisting of Cu͑100͒ terraces have been studied extensively by in situ TEM ͑above 423 K͒, [36][37][38][39][40][41][42][43][44][45] and by utilizing combinations of HREELS/XPS, [24][25][26] hyperthermal O 2 molecular beam ͑HOMB͒/XPS, 27,28 NEXAFS, 29 and MBSS. 17 All previous STM investigations [9][10][11][12][13][14][15][16] on Cu͑100͒ have been limited to oxygen coverages below 0.5 ML except our recent XPS/STM study that focused on the formation of subsurface Cu 2 O and CuO in the temperature range of 300-423 K. 60 In this paper we investigate physicochemical phenomena leading to oxide island nucleation on Cu͑100͒ by means of variable temperature scanning tunneling microscopy ͑VT-STM͒ and quantitative XPS as a function of O 2 pressure ͑8.0ϫ 10 −7 and 3.7ϫ 10 −2 mbar͒ at 373 K. We revisit the extensively studied surface confined oxygen adlayer and demonstrate that the ͑2 ͱ 2 ϫ ͱ 2͒R45°-O reconstruction is inert in the low pressure regime. The nucleation and growth of well-ordered two-dimensional Cu-O islands between two ͑2 ͱ 2 ϫ ͱ 2͒R45°-O domains is revealed by time-resolved STM experiments up to 0.5 ML of oxygen.…”