With the dual-layer DVD-9 becoming standard issue for DVD-Video, there is continuing discussion on the material options for the semi-reflective L0 layer. Target puts the case for silver alloy as a cost-effective and high yield option, comparing its cost and process characteristics with gold and silicon.

Han Nee Target Technology Company LLC

 

Among all materials, pure silver and silver alloys have the highest optical reflectivity. Because of this inherent high reflectivity, silver alloy L0 only needs a thickness of about 10nm to balance the signal between L0 and L1, as compared to 15nm of gold and 17nm of silicon. So for sputtering targets of the same size and shape, a silver alloy one produces about 50% more DVD-9 discs than a gold target, and about 70% more than a silicon one. L0 cost using silver alloy is around 0.5 to 1.0 US cent depending on target shape and sputtering machine efficiency. Silver alloy is comparable to silicon’s L0 cost per disc, but substantially lower than gold’s cost, which is about 4 to 6 cents per disc.

In addition to certain cost benefits, the balance reflectivity of silver alloy L0 with aluminum L1 is about 23%. Comparing with silicon’s L0 reflectivity balance of about 20% and gold at 22%, discs made with silver alloy L0 are far less likely to suffer rejects due to poor reflectivity. Because silicon’s reflectivity is the lowest among the three materials, it is more susceptible to reflectivity dropping below the DVD specification of 18%.

Although thin silver alloy L0 is desirable in lowering cost, it presents challenges in covering the DVD pit walls. Only properly designed and optimised silver alloy compositions work well in ensuring pit side wall coverage. A typical 10nm thickness silver alloy L0 thin film’s transmission electron microscope photomicrograph is presented in Fig 1 with a single DVD pit. The left micrograph showed silver alloy of inferior design with only partial coverage of the pit side wall resulting in signal problems. The micrograph on the right shows silver alloy of optimised design with good pit coverage and low jitter value.

 

It is seen in Fig 2 that the UV transmission rate of silver alloy L0 is about six times higher than that of silicon at 350nm wavelength. On average, six times more UV power is required to cure the UV resin through silicon than through silver alloy. This extra high UV curing power for silicon could potentially generate more heat in the disc causing more tilt rejects and lowering process yield.

 

Typical sputtering parameters in Fig 3 for different L0 materials using Unaxis Cube-Lite metallizer show that the sputtering yield or rate of either silver alloy or gold is six to seven times higher than that of the silicon under the same process condition. Sputtering silver alloy or gold requires less power, which results in lower heat generation. On the other hand, silicon requires increased sputtering power and argon flow, and longer sputtering time. These conditions demand very efficient and fast cooling to remove the generated heat. Silicon’s low sputtering rate could also present a bottleneck in the overall replication process as DVD makers continue to push for short cycle time.

Silicon’s low thermal conductivity, depicted in Fig 4, also contributes to other manufacturing issues. Both silver alloy and gold have similar and high thermal conductivity and coefficient of thermal expansion as compared to typical backing plate material such as copper which presents no problem. But silicon’s thermal conductivity and coefficient of thermal expansion is about five times smaller than copper’s. The low thermal conductivity, together with the high power sputtering of silicon, (which results in high target surface temperature) will lead to a high temperature gradient across the silicon

target thickness during production. This, coupled with the brittle nature of the silicon target, could at times crack the silicon target and lead to failure in the sputtering process. Metallic materials, such as silver alloy and gold, are not susceptible to this type of target cracking.

 

Environmental test results (80°C, 95%RH, 4 days) for a DVD-9 made with silver alloy TTP40-A for both the L0 and L1 are depicted in Fig 5. Following aging tests, no significant change in jitter value, PI error, or reflectivity for both the L0 and the L1 layers were observed. A DVD-9 made with the same stamper with gold L0 and aluminum alloy L1 also passed the same

aging test with no significant change in signal as shown in Fig 6. For the disc made with two silver alloy layers (L0 and L1), the jitter values for both layers are nearly the same.

Because L1 jitter values for Ag in Fig 5 are lower and better than the Al L1 values in Fig 6, higher L1 stamper yields are possible with a dual silver alloy layer disc. It is also possible to use a single and fast metallizer to deposit both the L0 and L1 of the same silver alloy, thus saving the cost of one metallizer as compared to the typical 2 metallizers arrangement for different L0 and L1 materials. With an efficient metallizer designed to sputter both L0/L1 silver alloys, the sputtering target cost per disc of Ag/Ag could be in the range of 1.2 to 1.5 US cents. This makes the so called ‘single component sputtering’ of silver alloy with a single metallizer per DVD line an attractive alternative line configuration.

As seen in Fig 7, an Ag/Al DVD-9 produced with a suitable UV resin along with clean polycarbonate and an optimised manufacturing process results in a disc that passes a very harsh environmental test at 80°C, 85 %RH for 20 days. High quality, low-cost DVD-9 with silver alloy L0 can be routinely made in mass production.

For best aging test results, UV resins specifically designed for silver alloy DVD-9 should be used. Most major UV resin makers carry products designed for silver alloy DVD-9 and are competitively priced. A limited number of UV resins designed for Au/Al DVD-9 can also be used for Ag/Al DVD-9 with acceptable aging test results. In general, UV resins for silver alloy DVD-9 should have low impurity levels and low water absorption to avoid corrosion of the aluminum L1 and the silver L0.

 

Because both Au and Ag are good electrical conductors, the sputtering process does not usually result in significant arcing. However for semiconductors such as silicon, which is an insulator, it can cause arcing on the target surface, interfering with the sputtering process. Silicon also reacts with residual oxygen or water in the sputtering chamber during production to form silicon oxide or silicate on the surface of the target and the mask. Silicon oxide, silicate, or sand particles on the mask, can result in pinholes on the L0 film.This necessitates a more frequent machine shutdown to clean silicon from the sputtering chamber and mask, and additional sand blasting to remove the silicon from the mask; shortening the mask life. Silver alloy or gold can be easily peeled off if the mask is coated with a suitable parting agent. Sputtered silver alloy and gold layers also present easily wetted surfaces for the bonding resin, resulting in very few air bubbles entrapped between L0 and L1. But a silicon L0 surface, which is more difficult to wet, tends to result in more bubble rejects.

 

A silver alloy DVD-9 shows easy process characteristics, similar to a gold DVD-9 but with

substantially lower L0 material cost per disc. Silver alloy results in about the same L0 material cost per disc when compared to silicon, but does not have the considerable process complications and yield problems. Production experiences at major DVD manufacturers around the world have consistently established the technical advantage and financial benefits of silver alloy when used for the reflective layers of DVD-9.