Prevention and control of intermetallic alloy inclusions that form during reflow of Pb free, Sn rich, solders in contacts in microelectronic packaging in integrated circuit contact structures where electroless Ni(P) metallization is present
In using Ni(P) and Sn-rich solders in Pb free interconnections, the prevention and control of the formation of intermetallic compound inclusions, can be achieved through a reaction preventive or control layer that is positioned on top of an electroless Ni(P) metallization, such as by application of a thin layer of Sn on the Ni(P) or through the application of a thin layer of Cu on the Ni(P
The invention is related to the fabrication of high reliability integrated circuit interconnection structures with lead free solder and in particular to the prevention of the formation of intermetallic compound inclusions that form during reflow connecting where electroless Ni(P) metallization is present.
BACKGROUND OF THE INVENTIONIn integrated circuit interconnection structures, a stack of layers are assembled in which the individual metal layers provide different functions. There are situations where efforts are required to prevent unwanted interaction between layers. There is such a situation in connection with the use of lead free solder and in particular to the prevention of the formation of intermetallic compound inclusions that form during reflow connecting where electroless Ni(P) metallization is present.
Electroless nickel-phosphorus (Ni—P) films are widely used in the microelectronic industry for several types of metallizations. They have such characteristics as excellent solderability, corrosion resistance, uniform thickness and selective deposition.
The electroless nickel phosphorous technology and its applications is well known and described in such publications as: Wiegele et al, in the Proceedings of IEEE Electronic Component and Technology Conference, 1998, p. 861; Mei et al in the Proceedings of the
IEEE Electronic Component and Technology Conference, 1998, p. 952; Lin et al in the Proceedings of IEEE Electronic Component and Technology Conference, 2001, p. 455; K. C. Hung in the Proceedings of IEEE Electronic Component and Technology Conference, 2002, p. 1650; and O.Villalobos in the Proceedings of IEEE Electronic Component and Technology Conference, 2002, p. 732.
In this technology, when an Ni—P film reacts with Sn—Pb eutectic solder, a part of the film underneath the solder crystallizes into Ni3P with a (P-rich layer); that forms at about the reflow temperature of about 200˜240° C. This low temperature reaction is referred to in the art as “solder reaction-assisted crystallization” and is described in a publication by Kim et al in the J Appl.Phys 85, 8456(1999). The “solder reaction-assisted crystallization” is different from the well known self-crystallization of Ni-P that occurs at a higher temperature.
The solder reaction-assisted crystallization is accompanied by the formation of inclusions of Ni—Sn intermetallic compounds and the formation of voids in the layer known as Kirkendall voids as described by Hung et al in the Mater. Res. publication Vol. 15, pg 2534, (2000); and by P. L. Liu et al in the Metall. Mater. Trans. publication Vol.A 31A, pg 2857, (2000).
Such interfacial reactions affect reliability and are often attributed as being the source of formation of a weak and brittle interface between Ni—P and Sn—Pb solder; as described in publications by; R Wiegele et al.in the Proceedings of IEEE Electronic Component and Technology Conference, 1998, p. 861; by Mei et al in the Proceedings of the IEEE Electronic Component and Technology Conference, 1998, p. 9520 and by Villalobos in the Proceedings of IEEE Electronic Component and Technology Conference, 2002, p. 732.
When Sn—Pb solder is replaced by Sn-rich Pb-free solders, the reliability issue of the Ni—P interface is expected to be even more important since Pb-free solders have a higher Sn content and a higher reflow temperature as described by K. C.Hung in the Proceedings of the IEEE Electronic Component and Technology Conference, 2002, p. 1650; and by. K. Zeng et al. in Materials Science and Engineering R 38, 55 (2002).
Electroless Ni(P) is a good candidate as a reaction barrier for Pb-free, Sn-rich solders, because the intermetallic compounds forming on an electroless Ni(P) surface tend to grow more slowly than on Cu metallization during soldering.
However, severe inclusions or spalling of intermetallic compounds from Ni(P) have been reported by Kang et al in the Proceedings of the 51 st ECTC May 2001 pgs 448-454; when Pb-free solders such as pure Sn, Sn-3.5Ag, Sn-3.5Ag-3Bi (in weight %) are applied in a form of solder paste onto an electroless Ni(P) layer. Typical examples of intermetallic compounds inclusions occur where for example Sn-3.5Ag solder paste is applied on an Ni(P) layer and reflowed at 250C for durations of between 2 min and 10 min. A further example of intermetallic compound spalling occurs when Sn solder is electroplated onto Ni(P) and the reflow condition is severe such as for 10 min reflow at 250C or the Ni(P) is electroless Ni(P).
The delamination or spalling of intermetallic compounds at the soldering interface is a reliability risk factor in thermo-mechanical solder joints.
There have been earlier efforts involving such metals as Au,Ag and Pd. In those efforts a thin layer of Au on top of Ni(P) metallization did not protect the Ni(P) and therefore intermetallic compound formation or spalling was observed. In the case of an Au layer, the dissolution rate of Au into the molten Sn-rich solder, such as Sn-3.5% Ag, is expected to be so rapid that it can not protect the Ni(P) metallization. A similar situation would be expected with a thin layer of Ag or Pd on top of Ni(P) metallization.
SUMMARY OF THE INVENTIONIn accordance with the invention where it is desired to control and suppress a reaction between Ni(P) and Sn-rich solders where an intermetallic compound inclusion may form adjacent to an electroless Ni(P) layer due to the P atoms that exist in Ni(P) and which would result in poor adhesion, control and suppression can be achieved through a control or protective layer on top of the electroless Ni(P) metallization such as by application of a thin layer of Sn on the Ni(P) or through the application of a thin layer of Cu on the Ni(P).
BRIEF DESCRIPTION OF THE DRAWINGS
Referring to
As reflow time is extended, intermetallic compound growth and separation from the interface, becomes more enhanced, as shown in
Referring to
It is considered that
The causes of the compound formations and resulting separations have been attributed to effects of poor adhesion between the P-rich layer on the Ni(P), intermetallic compound inclusions or spalling and insufficient protection of the Ni(P) layer during the reflow soldering operation.
Protection for the Ni(P) metallization can be provided through applying a protective layer directly on and over the Ni(P).
Referring to
As shown in
The protective layer 50 on top of Ni(P) metallization wets well the Sn-rich solder and forms Ni—Sn or Ni—Cu—Sn intermetallic compounds that operate to protect the Ni(P).
There is however a limitation in prolonged, extended reflow. As the time is extended, undesired separation progressvely appears. This is illustrated in
The effectiveness of a control or protective layer of the positioned at the interface 5 between the electroless Ni(P) layer and the metallization 1 is illustrated in connection with side by side photomicrograph views in
Considering
Considering
Considering
What has been described is the suppression of a reaction that forms intermetallic compound inclusions between Ni(P) and Sn-rich solders in Pb free interconnections that is achieved through a reaction preventive or control layer that is positioned on top of an electroless Ni(P) metallization, such as by application of a thin layer of Sn on the Ni(P) or through the application of a thin layer of Cu on the Ni(P).
Claims
1. In the fabrication of integrated circuit interconnection structures
- wherein stacked layers of metal for different purposes are caused to fuse together by subjecting said layers to a fusion temperature excursion,
- a process for inhibiting a potential undesirable interaction of first and second adjacent layers during said fusion temperature excursion comprising the steps of: exposing a surface of said first of said potential interacting layers, applying a protective layer of a metal having, during said fusion temperature excursion, at least one inhibiting property for said interaction, over said exposed surface of said first of said potential interacting layers, applying said second layer of said first and second potential interacting layers over said protective layer, and, subjecting the assembly stack of said first, said protective and said second layers to said fusion temperature excursion.
2. The process of claim 1 wherein said first layer is of electroless Ni(P) metallization, said protective layer is a metal taken from the group of Sn, and Cu, and said second layer is a further layer of the interconnection.
3. The process of claim 1 wherein said first layer is of electroless Ni(P) metallization, said protective layer is a metal taken from the group of Sn, and Cu, in a thickness range of between 0.1 and 5 micron and said second layer is a further layer of the interconnection.
4. The process of claim 1 wherein said first layer is of electroless Ni(P) metallization, said protective layer is a metal taken from the group of Sn, and Cu, in a thickness range of between 0.1 and 5 micron, said second layer is a further layer of the interconnection, and said temperature exursion is up through 250 degrees C.
5. The process of claim 1 wherein said first layer is of electroless Ni(P) metallization, said protective layer is a metal taken from the group of Sn, and Cu, in a thickness range of between 0.1 and 5 micron, said second layer is a further layer of the interconnection, and said temperature exursion is up through 250 degrees C.
6. The process of claim 1 wherein said first layer is of electroless Ni(P) metallization, said protective layer is a metal taken from the group of Sn, and Cu, in a thickness range of between 0.1 and 5 micron, and said second layer is a layer of Sn-rich Pb-free solders of a material taken from the group of pure Sn, Sn-3.5% Ag, Sn-0.7% Cu, Sn-3% Bi, Sn-3.5%-0.7% Cu, Sn-3.5% Ag-3% Bi, Sn-8%-Zn-3% Bi, Sn-3.5% Ag-0.7% Cu-0.5% Sb, and Sn-2.5% Ag-0.5% Cu.
Type: Application
Filed: Jul 30, 2004
Publication Date: Feb 2, 2006
Inventors: Sung Kang (Chappaqua, NY), Da-Yuan Shih (Poughkeepsie, NY), Yoon-Chul Son (Daejon)
Application Number: 10/903,365
International Classification: H01L 21/44 (20060101);