Your daily dose of photovoltaic technology developments and solar news

RSS | Newsletter | Advertisers' Hub | About Us | Subscription Shop

Plastic PV people: Ascent Solar makes steady progress toward market disruption (corrected version)

24 March 2009 | By Tom Cheyney | Chip Shots

Although there are 30-40 companies working to commercialize copper-indium-gallium-(di)selenide thin-film photovoltaics products, one firm has an approach unlike any other. Rather than depositing its films on glass or metal foil, Ascent Solar does its CIGS module processing on flexible plastic in a vacuum-based, roll-to-roll--and ultimately monolithically integrated--production flow.

While the Ascent team basks in the glow of the dedication of its new headquarters/factory in the Denver suburb of Thornton (with ceremonies featuring Colorado Gov. Bill Ritter, among others), they know there's still alot of heavy lifting to do--literally and figuratively--before the freshly renovated site can truly be called a manufacturing facility and the company's electronics-integrated PV and building-integrated PV products can make a disruptive play in the marketplace.

ashu_ascentsolar The company (which is publicly traded, although partner Norsk Hydro owns 35%) continues its conservative, fairly transparent, "steady as you go" strategy, with little of the hyperbolic chest-thumping--and pushbacks of production and commercial shipments--that have plagued the rollout of other firms' CIGS PV products.

Ashutosh "Ashu" Misra has been waiting a long time for this moment, so he's learned the value of patience when it comes to working with CIGS and CIGS-on-plastic in particular. Ascent's senior VP joined ITN Energy--the parent company which spawned both Ascent and its CIGS-on-foil cousin, Global Solar--in 1997 (with a break of a few years to serve as boss of Data Access America). ITN, which did yeoman's work pioneering thin-film PV process R&D with various government agencies in the 1990s and early 2000s, spun off Ascent in 2005.

We spoke in advance of today's ribbon-cutting about Ascent's new digs, its recently dedicated 1.5-MW pilot line in Littleton, details of the company's process flow and tooling requirements, current R&D efforts, and the outlook for Ascent and CIGS TFPV in general. (That's Ashu in front of a CIGS process tool in the accompanying photo.)

The 140,000-sq-ft Thornton site will house both the company headquarters and its first 30-MW volume production line. What once was a Fischer Imaging X-ray tool facility has been completely refurbished, with across-the-board upgrades (or brand-new installs) in power, HVAC, DI water, and the like, according to Misra. After laying out $5.35 million to buy the building last year, Ascent has spent $10 million on the renovations and some $110 million on manufacturing and support equipment.

"Every peripheral system we had to build," said Misra. "This was not like a semiconductor fab where you'd walk in with everything in place."

The senior exec told me that they finished specifying and then ordered all process tools by the end of the third-quarter 2008, and that the equipment has already started to arrive. He put the average lead times of the tool suppliers in the six- to nine-month range, with OEMs choosing to manufacture one system at a time, others assembling two platforms simultaneously. Misra raved about the improvements that he's seen with the toolsets over the past couple of years, calling the amount of progress "humongous," especially compared to the early days at ITN and Global Solar.

(CORRECTED PARAGRAPH) "We have received our first CIGS tool, and it is on the floor for installation," he noted. The next to arrive will be a Stangl-made cadmium-sulfide buffer-layer wet process system, followed by the deposition tools that will lay down the molybdenum and indium tin oxide films--both manufactured by General Vacuum. (NOTE: Original version incorrectly said "General Plasma" as MO and ITO tool supplier.)

As with other CIGS companies' processes, the art of production-line balancing requires mixing and matching the various toolsets to meet the specified capacity and a few years to fully optimize yields and efficiencies, according to Misra; in the case of Ascent, while three CIGS absorber and three monolithic integration systems will be required to hit that 15 MW mark, different numbers of Mo, CdS, and ITO tools will be installed.

The back-end equipment--including the proprietary laser etch/monolithic integration equipment and a fairly standard laminator tool--will be in place by year's end, keeping Ascent on track to ramp its first 15 MW of production capacity at the Thornton fab in the first half of 2010. Misra said the company will then ramp the second 15 MW of capacity in the latter half of next year.

The two most customized pieces of gear are the aforementioned CIGS tool and the monolithic integrator. An unnamed Japanese toolmaker, which works with Ascent through its trading partner Itochu, builds the shell of the CIGS platform, "and we do the rest of the integration of the sources and other stuff," he said.

ascent_solar_rolls Once the one-third-meter-wide, 150-meter-long rolls of plastic substrate complete their run through the molybdenum/back contact, CIGS/absorber layer, CdS/buffer, and ITO/transparent conductor layer processes, the total film-stack ends up between 2- and 2.5-µm thick, with an impressive across-the-web film deposition uniformity of ±2%.

The monolithic integration process, which takes place after all the thin films are deposited, "gives us the flexibility to finish the roll into multiple products," Misra explained. It employs a "combination of various types of printing techniques to lay down the insulating and conducting inks in vias or trenches."

While the opening of the Thornton plant gets all the attention, nitty-gritty process work still goes on at the newly dedicated 1.5-MW pilot line in the company's Littleton facility. Misra said they're running one shift there now, with plans to add a second shift by next quarter and a third shift by the following quarter, which will bring the line up to a 24/7-type operation and push the capacity close to nameplate status.

After spending at least six months dialing in or "freezing" the process recipes, the line is now fully stabilized and ready to ramp. Module-level conversion efficiencies are in the 7-8% range, according to Misra, and he's comfortable with the company's roadmap for 1% per year improvements to reach the 9-10% efficiency range by 2011. The National Renewable Energy Lab announced late last year that it had verified efficiencies as high as 9.64% from modules made on the Ascent pilot line.

"I think 10% is very realistic, and we're also looking at improving the substrate material, so we can operate on a high-temperature substrate." The company is working closely with Ube Industries, the Japanese company that supplies Ascent's flexible polyimide materials, to push the tolerance of the plastics so they can handle process temperatures of another 50°C, to squeeze even more efficiency out of the CIGS.

Ascent has invested heavily in "intelligent process control," with an array of primary inline sensors "extracting information in real time and hooking it back to our closed-loop system to play with the variables of pressure, process or web temperatures, and substrate temperatures."

Misra admits that although such APC tools are "really maturing and much more robust" than before, crediting "help from the semiconductor guys," he'd like to see more advancements in those technologies, which would lead to "significant improvements in our process, with better efficiencies and better yields."

It's a trickier proposition to control a large, roll-to-roll vacuum process than one where individual glass modules can be tested and the line adjusted accordingly, before yields plummet. "When you're processing a thousand-foot roll, you're stuck inside that machine. You need to make sure you have a pretty good handle on your process."

No PV module is worth anything without proper testing and certification, so Ascent has a comprehensive set of internal testing protocols and equipment to help accelerate its UL and TUV chances when the first plastic CIGS panels are sent to the labs. Since testing usually takes about six months, the company plans to submit its first samples next quarter so it can receive certification by year's end, but in the meantime it has accumulated a small mountain of internal data.

"We are developing something new and unique," he explained. "We want to make sure that before the product goes out of our door, even to customers for initial sampling, that we know exactly what is going to happen with it."

Although Ascent is laser-focused on ramping production and improving manufacturing efficiencies, it has an active R&D effort. One notable area involves efforts to come up with a cadmium-free, dry-process alternative to the current CdS buffer-layer step. In addition to the messiness of the wet CdS process now in place, Misra noted that "in Japan, it's very tough to sell a product with cadmium in it there."

Another research project examines one of the holy-grail materials of flexible CIGS BIPV--a moisture-resistant barrier encapsulation layer that is reliable, doesn't degrade quickly, and isn't cost prohibitive to make or integrate into the process.

"We're making pretty good progress there," Misra said. "We're working on an in-house solution plus another solution from three or four suppliers that are big chemical and plastic companies. We've seen good results with thermal cycling and have internal samples under damp-heat tests."

Ascent has been shipping "limited amounts" of prototypes to its customers, although Misra wouldn't cop to just how many modules are going out the door. He did say that those partners are eager to increase the number of Ascent PV products in their possession, so they can run them through their paces.

Customers of record include Icopal SAS, Texsa, TurtleEnergy, and part-owner Norsk Hydro seeking BIPV solutions, along with ICP Solar on the EIPV and portables side, with some U.S. military contracts still in the mix.

Misra remains a true believer in CIGS thin-film PV, but is realistic about the difficulties of perfecting the quarternary compound. "The greatest potential is with CIGS, but you have to give it time. The market has to be patient with CIGS." After all, he points out, thin-film king First Solar was no overnight success story, taking years to perfect its CdTe process--and burning through tens of millions of dollars of Walton family cash.

Considering Ascent's vastly differentiated value proposition for the EIPV and BIPV target markets--a competitively efficient, low-cost flexible plastic CIGS module that can be rolled onto or built into a rooftop, embedded into a building facade, or tailor-made for small electronic devices--perhaps it behooves the "market," within reason anyway, to hold the Colorado company to an alternative standard of "success."

ascent_solarhq