- THE MAGAZINE
- NEW PRODUCTS
The current creates extremely high heat at the contact points between particles. The ON-OFF DC pulse disperses the heat phenomena throughout the specimen, resulting in a rapid and thorough heat distribution, high homogeneity, and, in the end, consistent densities. (For additional details regarding the SPS process, read “Spark Plasma Sintering,” Ceramic Industry, May 2008, pp. 24-26.)
The simplest approach to volume processing with SPS technology is batch processing, where two or more parts or “samples” are made during one cycle. In the case of flat shapes, samples can commonly be stacked within the die using die spacers. Parts can also be oriented in a circular fashion around a circular die, much like the hour markings on a clock. Combining the two-part orientations is also possible. The key is that both force and electrical density need to be applied evenly across all of the samples within the die.
Barriers to CommercializationSPS has been around since the early ’90s, and SPS-like technology has been around for substantially longer. Though there have been attempts to apply SPS to a few commercial applications, the majority of interest in the method has come from R&D and academia. Several basic reasons are behind SPS technology’s slow migration to commercial use.
First and foremost, the few attempts at commercial applications have been collaborations between SPS manufacturers and end users, and non-disclosure agreements have limited detailed discussions. This lack of discourse and the general “black box” view of SPS have resulted in a very slow migration from academic and R&D to commercial use.
Second, SPS requires a significant up-front investment in time and money. The operational variables include, but are not limited to, material properties, die design and material, heat ramp rates, hold temperatures and times, force strategies, vacuum and atmospheric conditions, power settings, and cooling conditions. Commonly, test samples require sophisticated analysis. This development is necessary to draw the most benefits from SPS technology, and these variables need to be examined thoroughly before a high-throughput strategy can be built around an application.
Third, an important SPS market is only now emerging. Though there are many benefits to SPS processing, one of the most important is its ability to process very fine particulate materials with minimal grain growth and minimal negative effect on microstructure. Due in part to SPS technology, many advanced super-fine particle materials are only now becoming available on a commercial level.
High-Throughput ApproachThe key to advancing commercial SPS technology lies in the development of an efficient and reliable high-throughput system. The ideal system should have a minimal footprint and only one SPS processing section or press chamber. It must produce high-volume output and include a strategy for protecting the die during disassembly. Lastly, the price needs to be reasonable in relation to the potential part volume.
A new patented high-throughput SPS design has been developed that addresses all the above requirements (see Figure 2).* The process operates as follows:
Section one: two-stage glove box
- First stage-load material into the die (atmosphere-controlled fill hopper)
- Second stage-die assembly
- First stage-SPS heat/force cycle
- Second stage-the die is “hot-stripped”
Section four: load lock chamber for die-set removal
It’s important to keep several factors in mind. First, the die-set is automatically moved throughout the system on three separate reciprocating carriages. The press chamber isolation valves are water cooled. It’s important that the die be hot-stripped immediately after processing. If the die-set is allowed to cool while assembled, there can be substantial scoring damage on the die during disassembly because the material and die cool at different rates. Hot-stripping the die prevents this damage. The cooling chamber stages can be increased or decreased, depending on cooling needs.
The entire system is controlled via a single control unit, with each individual step programmed using macro variables. In other words, the process movements are triggered by other processes and are not just timed. In addition, the system is also protected by interconnect and thermal switches.
Tooling ConsiderationsTooling requirements and processing issues must be weighed carefully before settling on a final design. Tooling strategies are commonly pre-tested and may include several layers of materials, including semiconductor, non-conductive, and strength materials (see Figure 3).
Generally, when processing high-temperature materials (above 1200°C), graphite is the norm. SPS graphite tooling is usually made of fine-grain, high-strength, high-density, very high-purity material. It has a very high melting temperature, is a good semiconductor, is relatively inexpensive, and is easy to machine. Graphite also produces a reducing atmosphere within the tooling during processing, which can be very valuable when processing materials that are easily oxidized.
The negative aspects of graphite include possible carbon contamination, high wear, and low compressive and flexural strength. These drawbacks mean that net shaping is not likely unless the final parts do not require tight tolerances. It also means that tools need to be routinely replaced due to high wear and breakage.
Other tooling materials include tool steels, tungsten carbide and ceramics doped with conductive materials. All of these choices work well with high-volume and net-shaping applications but have limited temperature capacities. Generally, maximum processing temperatures for a given tooling material fall in the range of 60% of the melting temperature of the tooling, which is the lowest temperature material in the mix if it’s a composite.
Negative aspects of high-strength materials include high cost and chemical incompatibility with required atmospheres and processing materials. As mentioned previously, it’s not uncommon to insert a graphite sleeve in a high-strength die material. This combines all the benefits of graphite with the strength of steel or ceramic, and the insert is relatively inexpensive to replace.
Processing BenefitsBinder-less processing and net and near-net processing are major benefits of the SPS process. Because SPS technology does not require binders to bond powders, the pre-press operation (or “green” stage) is unnecessary. With high-strength tooling, net and near-net shaping is possible with limitations, one being size. The largest parts that have been produced using SPS are in the range of 450 mm diameter x 7 mm thick.
Figure 4 illustrates examples of shapes that are easily processed with SPS technology, while Figure 5 represents shapes that could not be processed with SPS technology, at least not in a single procedure. The primary issues are force and electrical density. If force and electrical density can be maintained evenly throughout 100% of the part, it qualifies as a candidate for SPS processing.
Calculating CostMany potential commercial applications exist for SPS technology. For cost calculation purposes, we’ll discuss a 200 mm diameter x 10 mm thick sputtering target with an approximate processing temperature of 1800°C. (This study is an examination of SPS operation and consumables; it does not include peripheral expenses such as the cost of material, material preparation, labor, or post-processing expenses.)
It is assumed that full density is required (above 98% relative density). Processing cycle times (ramp up and hold) are in the range of 14 minutes for one part and 17 minutes when two parts are stacked in the die. This example details the latter (two plates per run). Note: There is minimal out-gassing during processing, meaning there is no need to suspend processing to enable the vacuum to catch up.
In detailing the total processing cycle times, the process steps previously listed need to be reexamined. Assuming this is an ongoing process, it’s safe to say that the operator will load the empty die-set into the glove box while the SPS is processing another die-set. While in the glove box, the die is filled, weighed, vibrated, assembled and lightly pre-pressed. At this time, the operator activates the “ready” switch, notifying the control that the die-set is ready to be moved into the press chamber.
In the press chamber, as soon as the previous sample’s hold time is complete and the die is hot-stripped, the die-set is then moved into the seven-stage cooling chamber. Because the cooling chamber and the press chamber use the same atmosphere, very little vacuum is lost. Then the next die-set is loaded into the press chamber. If the following die-set is not in location and the operator has not hit the “ready” switch, the system simply waits until the switch is activated.
After seven cooling cycles, the die-sets are moved into the exit load-lock chamber and the atmosphere is neutralized. The operator is then notified that the load-lock chamber is ready for unloading.
Table 1 shows the time cycle breakdown. Rounded to a 24-minute cycle time, 60 SPS cycles or 120 sputtering targets could be achieved in 24 hours, equating to 720 plates per week or 2880 per month (assuming six-day operation each week).
Any high-throughput system requires regular cleaning and lubrication, plus ongoing minor adjustments. It is recommended that users of such a system plan for one day of downtime for each six full days of operation.
Required utilities include power, water and compressed air. The ongoing expense of operating a cooling tower and utilizing compressed air is approximately $100 each per week.
Maximum power consumption is estimated at 780 kVA. Maximum power is active during roughly 71% of the cycle (17 of 24 minutes); otherwise, the system runs on approximately 20 kVA. In a given week, which would include 20 kVA for power usage during maintenance, total power consumption would be in the range of 81,016 kWh. At today’s energy costs (the 2008 U.S. national average was $.11 per kWh), that would average approximately $8900/week.
A budgetary quote for the system detailed in this report, including initial tooling, is $2.8 million. For this example, the total initial expense will be depreciated over eight years. This puts a total monthly purchase burden of $29,167 on the system, which brings total monthly expenses to $108,779. When that total is divided by the monthly production throughput of 2880 plates, the total cost per plate equates to $37.77.
Hot isostatic pressing (HIP) and hot pressing (HP) are the standard conventional technologies for consolidating powder sputtering targets. With targets of this size, manufacturing expenses (including consolidation, secondary machining, labor, and general business) commonly run in the range of $1000 to $1200 per target. Generally, the consolidation process and equipment (including equipment purchase and depreciation) would cover approximately 40% of these values. Not including labor, consolidation expenses would therefore run $400.00 to $480.00 per target.
In the case of 350 mm targets, the general approach is the same. The processing time per press increases to 25 minutes with a 32 minute cycle, which would enable 90 targets per day. With similar general expenses and an increase in power consumption, the cost per target would be $52. If we look at 50 mm targets, the throughput per press would be 36 targets. With similar processing expenses and cycle times, the throughput would be 2160 targets per day at a cost per target of $2.09. As a reminder, high-strength tooling can be used when processing lower-temperature materials. This would open the door to net shaping, thus increasing profitability even further.
As mentioned above, the materials used in these examples produce minimal out-gassing during processing, and there was no need to suspend processing for vacuum. Many materials do out-gas heavily during processing, especially in volume processing. Let’s consider the 200 mm plate example with a 57 minute cycle (approximately 20 minutes at full power, 780 kVA; 30 minutes of 20% power, 158kVA; and seven minutes at 20 kVA). In this scenario, the throughput drops to 50 plates per day, but the power consumption and the tooling expenses also drop. When calculated, the cost per plate would rise to just $59.11, still a significant savings compared to conventional processing.
Moving Toward CommercializationPrevious impediments to the commercial growth of SPS have been removed with the development of a comprehensive high-throughput system. This new technology, which is an economically sound and viable approach for manufacturers in a range of industries, is poised for tremendous growth.
For additional information regarding high-throughput spark plasma sintering, contact Thermal Technology LLC at 1911 Airport Blvd., Santa Rosa, CA 95403; (707) 571-1911; fax (707) 571-8098; e-mail firstname.lastname@example.org; or visit the website at www.thermaltechnology.com.