Insitu Observation of Growth Dynamics on Cu100

In a general sense, STM may be used for dynamic studies whereby the long unresolved issue of the influence of adsorbates on the microstructural evolution of electrodeposited films may be addressed. In this vein a variety of notable reports of copper deposition on gold have been described [16]. Our initial studies focus on homoepitaxial growth on Cu(100) under conditions relevant to our multilayer synthesis program. During multilayer growth copper reduction occurs under diffusion control and this raises the possibility of developing morphological instabilities. However, the fact that laminar multilayers have been successfully grown [1,2] suggests that there is a significant degree of surface or step mobility associated with the electrocrystallization reactions such that compact surface structures are formed [2].

This argument is supported by our STM observations. Immersion of Cu(100) in 10 mmol/1 HC1 followed by polarization at -0.250 V vs. Cu/Cu+ results in the formation of a stable chlorine adlattice as described previously. In the presence of a dilute cuprous

Fig. 5. A series of 125 nm x 125 nm STM images of step flow associated with Cu+ deposition on Cu(100) at -0.25 V vs. SCE, with Etlp = 0.010 V , 7t = 15 nA and a frame time of-19 s for images collected over a period of 3 min; scanning (a) down 0-18.7s, (b) up 18.9s-37.6s, (c) up 55.6-74.3s, (d) down 74.4-93.2s, (e) up 93.25-112s, (f) down 112.1-130.8s, (g) up 130.7-149.7s, (h) down 149.7-168.5s, (i) up 168.5-187.3s.

Fig. 5. A series of 125 nm x 125 nm STM images of step flow associated with Cu+ deposition on Cu(100) at -0.25 V vs. SCE, with Etlp = 0.010 V , 7t = 15 nA and a frame time of-19 s for images collected over a period of 3 min; scanning (a) down 0-18.7s, (b) up 18.9s-37.6s, (c) up 55.6-74.3s, (d) down 74.4-93.2s, (e) up 93.25-112s, (f) down 112.1-130.8s, (g) up 130.7-149.7s, (h) down 149.7-168.5s, (i) up 168.5-187.3s.

chloride solution, nominally 1 mmol/1 Cu+ in 10 mmol/1 HC1, metal reduction proceeds under diffusion control at -0.250 V. Growth is generally observed to occur by step propagation in the <100> direction with the energetics associated with the chlorine adlattice dominating the evolution of step morphology as shown in Fig. 5. The electrocrystallization reaction develops according to a combination of two competing mechanistic paths. In the first instance, reduction of the cuprous chloride complex may be at catalyzed the kink-saturated <100> steps. This is consistent with the marked difference reported for the exchange current for silver deposition at step edges versus the adatom state [17]. The distinct <100> step orientation results from subsequent energy minimization which is facilitated by the rapid transport of reduced atoms along step edges.

An alternative scenario for step-mediated growth would entail the capture of adatoms from a two-dimensional (2D) gas on neighboring terraces. In this model the terrace adatoms which form the 2D gas are presumed to be sufficiently mobile so as not to be imaged by the STM [18]. Similarly, in this instance the adatom density relative to the mean diffusion length must be such that adatoms are trapped by the steps as opposed to nucleating as islands on the large terraces. This second model follows from a direct analogy with physical vapor deposition [19] and thus ignores the possible influence of the higher-coordination step edges on the inner-sphere reduction reaction. Nonetheless, at least two possible homogeneous island nucleation events are visible in Fig. 5(a) and (i).

Further experimentation under controlled conditions of saturation and crystal miscut will enable us to address the growth mechanism more accurately. As noted earlier, it is encouraging that several theoretical works already exist for describing homoepitaxy of Cu on Cu(100) from the vapor phase [11].

0 0

Post a comment