Achievements to Date in Cost and Technology

Over the past 6 years, Sunpreme has driven rapid learning cycles at its R&D Labs in Sunnyvale, California, as well as in its manufacturing plant in Jiaxing, China near Shanghai.

We have established core competencies in: a) Materials, equipment, and devices with talent trained at Applied Materials, Intel, and Bell Labs, b) Advanced PV with team members educated at Stanford and  Osaka  University’s  Professor Hamakawa’s group that invented the heterojunction cell, and c) Manufacturing quality and reliability, thanks to leadership previously developed at  Intel, SunPower and Canadian Solar.

Sunpreme’s innovative Hybrid Cell Technology (HCT) platform synergizes the efficiency advantage of amorphous Si thin-film based amorphous Si with cost advantages of ultra thin-film materials processed at lower temperatures in a TFT-FPD based PECVD and PVD reactors.

We have many innovations under our belt both on the Cost/ Technology side and on the Business Side.

  • Cost: World class LCOE and AC cost per Watt, even at limited production volume
  • Efficiency: Cells at 23.5%, modules at 22% with roadmap to 28%
  • Device Technology: TF HCT with LTPECVD deposited a-Si:H thin-film p/i//i/n junctions
  • Manufacturing Technology: Lean, Moore’s law compatible as proven on TFT-FPD
  • Quality and Reliability: Unsurpassed Aesthetics, Fire Class A, Wind resistance to 300 kmh
  • Warranty: Performance warranty backed by Munich Re re-insurance
  • Deployment: Across 5 continents, in 24 countries
  • Growth driver: Premium performance, commodity cost, CAPEX Lite
  • Differentiation: SNPM premium panels earn high ratings from independent bodies

Symmetric Bifacial Architecture


The symmetric Bifacial HCT cell architecture uses thin amorphous Silicon films to produce a p-i-(Si substrate)-i-n junction. Each of the four amorphous Si films is ultra-thin at ~10nm thick. Both interfaces between the films and to the substrate are atomically abrupt, hence chemically inert, unlike crystalline Si PV with diffused p-n junctions. Moreover, the low process temperatures ensure no c h e m i c a l interactions throughout the cell manufacturing. The role of Si substrate is to provide a light absorbing m a t e r i a l with a stable, stress-free interface for the deposited amorphous film emitters and collectors. The cell current is collected by blanket layers of transparent conducting oxides, and carried out to the outside world through patterned metal interconnects. These cells are packaged in modules made of double glass panels providing extreme ruggedness combined with unsurpassed aesthetics. Installations are clean and fast with clamps and adapters interfacing into standard trackers on the roof tops.


The production process is a lean 5-step sequence with a cycle time of only 6 hrs, yields >99% and up to 6σ SPC. It includes PECVD for a-Si:H stack, both undoped and doped, and a PVD step for TCO films. The films are deposited both on the front and back side of the substrate for bifacial action.

Both of these critical deposition systems use a similar design as has been proven out for TFT LCD displays. But instead of glass, we deposit on Si substrates laid out on trays.


Our cell efficiencies range from 23 – 23.5% for HCT cells utilizing a precision Cu patterning process technology. With fill factors at 0.80 – 0.82 and Voc’s at 740-750 mV, these cells deliver a jaw-dropping Pmax of >5.6W. The modules deliver 400W STC.


Our Bifacial panels provide 10-20% energy boost from reflected and diffuse light. With a 3x higher level of reliability than industry certification requirements. And, superior wo r l d – c l a s s aesthetics. The module efficiencies can easily reach 22% in real-world applications. Moreover, the efficiency remains high even at low irradiance seen around dawn and dusk, as well as in haze or fog.

HCT platform

Bridging the technology chasm with our HCT platform

A vast majority of PV cells being made in the world still use the low-cost p-type (multi- or mono- crystalline) Si based Solar 1.0 technology, with an annual production capacity greater than 60 GW. The price for such products has continued to decline, due to enormous scaling and overcapacity. However, PV efficiencies are capped at 18-19%, and the module output at ~300W. A higher cost embodiment called PERC is able to improve performance somewhat to 20% efficiency.

On the other hand, for the higher performance, n-Si based Solar 2.0 market there is a severe under capacity, and demand is growing faster than production capacity. This technology has evolved over time, starting with B-diffused junction into n-Si with interdigitated back contact. This was followed by a thin-film junction based HIT™ (Panasonic) technology which holds high PV efficiency records in the industry.

While it will be logical to expect the production capacity to transition from Solar 1.0 to 2.0, this is not happening due to a deep technology chasm. It is not possible to transition the p-Si factories into n-Si production, because the latter requires a more complex process technology and a different equipment set.

With one exception – Sunpreme’s unique HCT thin-film PV platform has been able to bridge the chasm by changing the role of Si substrate from a diffused p-n junction element into an inactive substrate for a thin-film based floating junction.


The HCT platform drives a multi-generational Efficiency and Cost reduction roadmap

Back in 2012, we commercialized our thin film based PV junctions consisting of amorphous n-i//i-p films sandwiching a p-type MG-Si substrate with cell efficiencies of 14-16%. This platform included double glass, frameless modules. Next, in 2013 we were able to utilize the same equipment set, but a different process sequence to make 19% mono-facial cells with thin-film amorphous p-i//i-n junction on n-type substrates, but with a reflective back contact consisting of AZO/Al/NiV. This was the first demonstration of any factory being able to produce both Solar 1.0 and Solar 2.0 product back to back without changing the equipment set.

In 2014, Sunpreme introduced a Bifacial Solar 2.0 product based on the ultra-thin film amorphous Si junctions still using the common HCT platform.  We were able to implement this ultra-thin film junction technology

across three product generations. This was a third generation high performance Bifacial cell where the back side reflective metallization was replaced with a TCO and metal collectors. The front side efficiency increased to 22% range. The backside cell efficiency was measured at ~21%, resulting in a Bifacial coefficient of 0.95.

Going forward, we expect our module level effective efficiencies to continue their upward trajectory to 24% and 26% at module level. And, thereafter, we forecast a paradigm shift with new 6th generation superstructures over the HCT platform that will enhance efficiencies into 28% range.  All these improvements are planned with manufacturing-worthiness as a primary boundary condition. And, they will lead to step-wise reductions in production costs, which we are able share with the customers as a reduced price per Watt, without a disruptive loss of profitability.


Cost: Think all premium panels are the same? Think again!

Recent (Q416) spot market price for p-PERC modules have been reported to be around 40 ¢/W. We believe a fair market price for Sunpreme modules can be justified at twice this amount, as illustrated in the accompanying water fall diagram. This example is for a large commercial roof top in Southern California.

The comparison concerns Sunpreme’s GxB 420W equivalent panel vs a top-of-the-line p-mono PERC module rated at 350W. Staring from a nominal price of 84 ¢/W for Sunpreme product, we have applied various credits relating to a 15% Bifacial gain for C&I (Commercial & Industrial roof top), a lower thermal coefficient of efficiency (0.28%/C vs 0.45%/oC), an LID benefit of 3%, a Voc benefit of 735mV/cell vs 560 mV/cell) bringing the effective price down to 60 ¢/W for an “apple to apple” module level comparison. Moreover, at systems level an additional benefit accrues for Balance of Systems (BOS) simply because we need 22% fewer modules in case of higher power SNPM product. This is reflected in a significant savings for materials, i.e. racking and wiring portion of the BOS (R&W), and labor portions of the BOS (L). The end result of all these savings is an effective price of 35.6 ¢/W, which is in the same ball park as 40 ¢/W. We have not monetized here SNPM’s Class A fire and 300 kmh typhoon wind rating, nor a 0.1% less power degradation over 25 yrs, backed by a Munich Re performance warranty, uniquely for SNPM products.

It needs to be emphasized that this chart is a snap shot of current (Q42016) situation. With time, we expect aggressive cost reductions driven by Sunpreme’s innovations & scaling, thereby keeping a sustainable competitive edge in the market.