添加銅鎳至銲料中探討抑制微孔洞以及Cu3Sn的機制

Abstract

銲料在微電子工業上的挑戰，如封裝相容性，潛變，微孔洞等等，而其中在經過高溫熱處理的試片，沿著Cu3Sn與Cu的界面，我們可以觀察到一層連續micro voids在Cu3Sn中生成，Micro voids將提高銲點發生脆化的潛在風險，影響銲點強度，降低產品使用壽命，我們若能對於銲料與金屬墊層間的反應情況有更深的瞭解與控制勢必可增加銲點強度。銲料與Cu墊層反應後，在Cu3Sn中有micro voids生成已經被許多文獻報導，由於Cu3Sn生長與micro voids息息相關，愈薄的Cu3Sn生長對銲點強度的提升也許會有幫助。目前銲料的發展為微量元素的添加去抑制Cu3Sn的生長，然而微量元素如何去抑制Cu3Sn的生長機制還尚未瞭解。本實驗主要目的為探討銲料中Cu濃度對micro voids生成以及添加Ni至銲料中探討Cu3Sn生長機制，實驗分為兩大部分，第一部分我們討論銲料中Cu濃度與micro voids的關係，第二部份討論添加Ni與Cu3Sn生長機制的關係，實驗參數與結果如下： 第一部分： 所使用的銲料組成分別為Sn-xCu(x=0, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 wt.%)，與Cu板進行頂溫235oC、90 sec迴銲，再將銲接後的試片經過160oC，500、1000、2000 h熱處理，研究重點在於探討銲料中Cu濃度對micro voids生長的影響。由實驗結果得知在經過高溫熱處理後，Cu6Sn5與Cu3Sn同存在於界面處，當反應時間500 h尚未有micro voids生成，直到反應時間達1000 h，當銲料中Cu濃度≦0.46 wt.%時，在Cu3Sn中有micro voids生成，當反應時間達2000 h銲料中Cu濃度≦0.50 wt.%時，在Cu3Sn中有micro voids生成。由實驗結果可知，在熱處理溫度160 oC反應時間達2000 h，可藉由維持在銲料中的高Cu濃度(e.g.≧0.58 wt.%)來有效抑制micro voids生成。 第二部分： Micro voids只生長在Cu3Sn中，目前抑制Cu3Sn生長的有效方法為微量Fe、Co、Ni添加，然而其抑制機制尚未明確，為了研究Ni對Cu3Sn生長的影響，此部分實驗所使用的銲料為10Sn90Pb與5Sn95Pb添加0、0.03、0.06、0.1和0.2 wt.% Ni，反應條件為在350 oC下進行2 min迴銲，以及160 oC下500，1000，2000 h熱處理。在迴銲後，Cu3Sn是唯一生成的介金屬，而在固態熱處理後，10Sn90Pb-xNi界面上除了生成Cu3Sn外，也有Cu6Sn5生成，而5Sn95Pb-xNi界面上只有Cu3Sn生成，介金屬生長順序可藉由Cu-Sn-Pb三元相圖得知。實驗結果顯示微量Ni添加到高鉛銲料中無法抑制Cu3Sn的生長，然而由過去文獻可知Ni添加到無鉛銲料中可以抑制Cu3Sn，這是由於在無鉛銲料中添加Ni，當試片迴銲後會生成很厚且鬆散的(Cu,Ni)6Sn5，形貌似樹枝狀結構，讓Cu原子優先與Cu6Sn5縫隙中的Sn反應生成Cu6Sn5，因此抑制Cu3Sn生長，因此Ni抑制Cu3Sn必須透過Cu6Sn5。

The drive for solders in the microelectronics industry presents some reliability challenges. Examples include package compatibility, creep and micro voids. Along the Cu3Sn/Cu interface, we can find a series of micro voids. These micro voids were the true culprit responsible for the weakening of the interface. It is of importance for the reliability to have better understanding and control of the solder/metallization interactions during soldering. In the reactions between solders and Cu substrate, the formation of micro voids within the Cu3Sn layer had been report by many research groups. Because the Cu3Sn growth had been linked to the formation of micro voids, which in turn increased the potential for a brittle interfacial fracture, thinner Cu3Sn layers might translate into better solder joint strength. It is widely accepted that the formation of these micro voids is related to the growth of Cu3Sn. Main thrust of solder development has shifted to the minor alloy additions to retard Cu3Sn growth. However, the mechanism explaining how adding minor alloys can reduce the Cu3Sn thickness is still lacking. The main objectives were to investigate the effect of Cu concentration and the effect of Ni addition on the formation of micro voids and the mechanism of Cu3Sn growth. This study divides two parts. In first part, we discuss the relation between apparent Cu concentration in solder and micro voids. In second part, we discuss the relation between Ni addition and Cu3Sn growth mechanism. The experiment parameters and results are as follows: First part experiment: The interfacial reaction between Cu and Sn-xCu (x= 0, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7 and 0.8 wt.%) solders were examined. Emphasis was placed on the effect of Cu composition on the formation of micro voids in the Cu3Sn phase. Reaction conditions included one reflows and subsequent aging at 160 oC for up to 2000 h. After aging, both Cu6Sn5 and Cu3Sn formed at interface, and Cu3Sn was void-free, and remained void-free after 500 h of aging for all of the compositions. The micro voids appeared in Cu3Sn after 1000 h of aging, but only in those solder joints with their apparent Cu concentrations that were 0.46 wt.% and lower. After 2000 h of aging, micro voids formed only in solder joints with their apparent Cu concentrations that were 0.50 wt.% and lower. This observation suggests that the micro voids in Cu3Sn can be effectively inhibited during aging at 160 oC for 2000 h by maintaining a high Cu concentration in solder (e.g ≧ 0.58 wt.%). Rationalization for this Cu concentration effect is presented. Second part experiment: In the reactions between solders and Cu, minor alloy additions, such as Fe, Co, or Ni, to solders often reduce the Cu3Sn growth rate. Nevertheless, the mechanism for this effect remains unresolved. In order to investigate the effects of Ni on Cu3Sn, the solders used for this study are 10Sn90Pb and 5Sn95Pb doped with 0, 0.03, 0.06, 0.1, and 0.2 wt.% Ni. Reaction conditions included one reflow at 350oC for 2 min and solid-state aging at 160 oC for 500, 1000 and 2000 h. In reflow study, Cu3Sn was the only reaction product observed for all the different solders used. In solid state aging study, both Cu3Sn and Cu6Sn5 formed in 10Sn90Pb-xNi solders, but only Cu3Sn formed in 5Sn95Pb-xNi solders. The growth order of intermediate component can clearly know by using the ternary Cu-Sn-Pb phase diagrams. The key observation to be made in this study objective is to see whether the growth rate of Cu3Sn can be reduced by Ni additions under these situations. The experimental results show that minor Ni addition to high-lead solder can’t retard Cu3Sn thickness. However, the Ni addition to lead-free solder can retard Cu3Sn. This is because adding Ni to lead-free solder transformed the microstructure into a much loose (Cu,Ni)6Sn5. The Cu atoms prefer to react with Sn atoms of Cu6Sn5 interstice. They tend to form Cu6Sn5. In addition, the morphology of (Cu,Ni)6Sn5 gradually become layer-type (Cu,Ni)6Sn5 after aging. Thicker Cu6Sn5 layer-type morphology becomes a good diffusion barrier stopping the necessary atomic flux necessary for the growth of Cu3Sn. Therefore, Ni retards the growth of Cu3Sn through Cu6Sn5. The objective of this study is to investigate individually the influences of Cu concentration on micro voids and Ni addition on Cu3Sn growth. Emphasis is placed on a systematic comparison study on the effects of Cu and Ni addition.