Copper Resistivity: A Problem at Smaller Dimensions?
Peter Singer, Editor-in-Chief -- Semiconductor International, 5/1/2001
 |
There are about 22 semiconductor manufacturers worldwide engaged in the use of copper. Early adopters of copper technology implemented copper in production at the 0.25 µm device node. Most copper in production or pilot production today is at the 0.18 and 0.13 µm technology nodes. For some manufacturers, 0.13 µm is a second generation of Cu; for others, 0.13 µm is the entry point for putting Cu into production. The wafer size transition from 200 to 300 mm for copper devices will occur this year, largely at the 0.13 µm technology node. By the 0.10 µm technology node, the majority of devices are predicted to be manufactured on 300 mm wafers. The applications served by this technology are advanced logic products such as microprocessors and DSPs. The use of copper is no longer dominated by semiconductor manufacturers in the United States; approximately half of the world's users of copper technology are now in Asia.
A potential problem, however, is that the resistivity of copper may actually increase in 0.10 µm and smaller geometries, partially negating the main reasons why manufacturers switched from aluminum to copper in the first place. Michael Thomas, chief technical officer, wafer fabrication materials, at Honeywell (Sunnyvale, Calif.), says it's all about the mean free path of an electron in copper (about 450 Å). In a 0.10 µm line, the mean free path is on the order of 500 Å, he explained. This means that, in the near future, the resistivity of copper will increase with smaller feature sizes. "People have not had to deal with this yet because we haven't gotten to feature size-limited resistivities. We will see, somewhere between 1500 and 1000 Å, a resistivity increase in the copper metallurgy. This has nothing to do with alloys or barriers, it's just the physical constraint of the copper itself." The problem is compounded by the barrier films, which take up an increasingly larger percentage of the space in the via, resulting in a smaller cross section of copper, Thomas said. "What's going to happen is that the resistivity that people have planned on retaining all the way to 0.05 µm will probably rise."
Fortunately, the problem is not as bad as it might first appear, at least on the lower levels of metal, Thomas said. "Typically, when you use the finest feature size, what will happen is the wiring will be short. And therefore, in the lower levels of metal, the capacitance is really the major issue instead of the resistance of the wire; the contacts will dominate the resistive part of the RC delay." There potentially could be a problem with the intermediate and upper levels of metal, however. "If you wish to use very tight copper wires that are long, what's going to happen is the resistivity of that wire and total resistance can hurt you. You're not going to get what you think," Thomas said. Fortunately, most designers increase the size of the wires when they become long, which would delay potential resistivity problems.
The cause of reduced mean free path in copper has to do with the way it's deposited. "What happens when you plate copper is that you get seams, depending on if you go from the bottom up or if you do it from all three sides. Those seams and the sidewalls of the copper line act as interfaces that can be scattering sites," Thomas explained. He added that it will be very difficult to get around this because billions of wires are being fabricated on each die. "There's something called specular reflection and it's the way you control the reflection of electrons off of an interface or a wall. That's going to determine how resistive the copper is. There may be some orientational dependencies on the amount of loss and the increase in the bulk resistivity that you see caused by those sidewalls and interfaces. That's something that you could try to control, but to completely control it through billions of wires will be quite a challenge."
Surprisingly, aluminum has much shorter mean free path than copper, measuring about 250 Å. This means that, at some point, aluminum wires and copper wires may have equivalent resistivity at spacing between 450 and 250 Å. "I would anticipate this effect would not be as pronounced in aluminum as it will in copper as we shrink, even though aluminum has a higher starting bulk resistivity," Thomas said.
For additional information on wafer processing, go to www.semiconductor.net/wafer
