摘要: | 利用掃描式電子穿隧顯微鏡(in situ scanning tunneling microscopy,STM)和循環伏安法(cyclic voltammetry,CV)分別觀察UPD 銅/鉑(111)電極上聚乙二醇(PEG)的吸附結構,以及PEG 修飾UPD 銅/鉑(111)電極過後對銅沉積的影響。在1 M H2SO4 + 1 mMCuSO4 + 88 μM PEG200 溶液中,UPD 銅/鉑(111)電極上有整齊的(√3 × √7)硫酸氫根結構存在,當加入1 mM KCl 後,此硫酸氫根結構即會被氯和PEG200 所置換。當穿隧電流從0.5 nA 降低到0.2 nA 時,可觀察到面上由氯的(√3 × √3)R30˚,變成(4 × 4)的PEG200 結構。在溶液中含有氯及PEG 時的銅沉積過程,會先經過一成核的過程,電極面上會生成大小約2~3 nm 寬的CuCl 島狀物。當銅沉積的過電位小於50 mV 時,電極面上的CuCl 會生成較多,進而導致三維銅塊的沉積;當銅沉積的過電位大於150 mV時,電極面上的CuCl 生成較少,銅膜會優先從缺陷及台階處生成,則可使銅膜呈現層狀的二維成長,而形成一平整銅膜。此現象在不同分子量PEG 中皆有觀察到,分子量小的PEG 溶液中易生成較大量的CuCl 島狀物,而大分子量的PEG 對銅沉積的抑制效果較佳。提高硫酸的濃度對銅沉積的抑制效果也有幫助。不同層的銅膜有不同的結構,第一層銅膜為雙條紋結構(double-lined pattern),第二、三層為波浪狀結構(moiré pattern),其中包含一些凹陷的三角形缺陷。是因為載體鉑原子和銅的原子大小不同,當銅沉積在UPD 銅/鉑(111)電極時,使金屬原子的排列產生錯排的現象。為了解添加劑對析出銅的形貌有何影響,我利用STM 觀察在1 M H2SO4 + 0.1 mMCuSO4 + 1 mM HCl 初始溶液中,分別添加PEG 及SPS 時,銅沉積於多層銅/鉑(111)電極上的形貌及析出的過程。在含PEG4000 時,這些高分子可能形成捲曲的結構,而均勻的鋪蓋在電極表面,這些吸附的高分子造成銅緩慢的析出,唯獨凹陷的周邊無PEG分子吸附,因此銅會優先在這些區域,以較平台和缺陷底部快的速率沉積,最終造成粗糙的表面形貌,填孔效果不佳。當初始溶液中添加SPS 時,銅沉積會優先從台階邊緣開始,但會形成雪花狀的銅膜,而使新生台階邊緣不規則,凹陷底部可能未填滿時即被新生銅膜覆蓋,導致填充銅中出現孔洞。而初始溶液中同時含有PEG 及SPS 時,銅沉積於凹陷底部的速率明顯的高於只含一種添加劑的情況,所以銅膜中的凹陷最終都被銅完全填滿,此一結果顯示SPS 和PEG 共存時可獲得較佳的填孔效果。比較兩種不同分子量的PEG 後發現,PEG8000 較PEG4000 能更有效的抑制銅的沉積。In situ scanning tunneling microscopy (STM) and cyclic voltammetry (CV) were used to examine the structure of adsorbed polyethylene glycol(PEG) and its effect on theelectrodeposition of copper on Pt(111) modified by a monolayer of copper adatoms in 1 M H2SO4 + 1 mM CuSO4 + 88 μM PEG200. A highly ordered Pt(111)-(√3 × √7)-Cu + HSO4- structure due to underpotentially deposited Cu monolayer was observed in 1 M H2SO4 + 1 mM CuSO4. This adlattice restructured upon the introduction of polyethylene glycol (PEG, molecular weight 200) and chloride anions. At the onset potential for bulk Cu deposition (~0V), a Pt(111) - (√3 × √3)R30° - Cu + Cl- structure was imaged with a tunneling current of 0.5 nA. Lowering the tunneling current to 0.2 nA yielded a (4 × 4) structure, presumably due toadsorbed PEG200 molecules. The subsequent nucleation and deposition processes of Cu in solution containing PEG and Cl- were examined, revealing nucleation of 2-3 nm wide CuClclusters on an atomically smooth Pt(111) surface at overpotentials less than 50 mV. With larger overpotential (η > 150 mV), Cu deposition seemed to bypass the production of CuCl species, leading to layered Cu deposition, starting preferentially at step defects, followed by lateral growth to cover the entire Pt electrode surface. These processes were observed with both PEG200 and 4000, although the former tended to produce more CuCl nanoclusters. Raising [H2SO4] to 1 M substantiate the suppressing effect of PEG on Cu deposition. It couldobserve different structures along with the the copper growing layer by layer. On the first layer the chloride adlayer produced a double-lined pattern, followed by a moiré pattern along with minor triangular depressions seen with Cl adsorbed on the second layer of metal film.Copper atoms would have to occupy different sites on the Pt substrate, yielding unlike corrugation heights of the adsorbed chloride adatoms. To decipher the effects of organic additives on the electrodeposition of copper, Iemployed in situ STM to reveal the surface morphology of Cu deposit and the dynamics of this event. The electrode was made of multilayer Cu on Pt(111). The base electrolyte used in this study was 1 M H2SO4 + 0.1 mM CuSO4 + 1 mM HCl. The organic additives used were PEG4000, PEG8000, and SPS. Results obtained with PEG4000 showed that Cu wasdeposited uniformly on the electrode, except at the perimeters of depressed features or pits. Presumably, PEG4000 admolecules assumed coiled structures on the electrode surface, including terraces and pits. In comparison, perimeters of pits were less likely to be covered by PEG molecules and thus were more open to Cu deposit. With more Cu deposition, the resultant surface morphology became rougher than the pristine Cu surface. In the presence of SPS alone, Cu deposit grew in 2D, producing flake like features whose edges were apparentlyrugged, suggesting that SPS molecules, rather than chloride, were the most important adspecies on the Cu deposit. It is thought that pits in the Cu substrate might not be entirely filled, leading to voids in the Cu fillings. In the end, only the combination of PEG and SPSwas able to guide the deposition of Cu at depressions in the Cu substrate. Judged from the results seen with PEG4000 and 8000, the latter was superior in inhibiting Cu deposition. |