Strategic program: Solar+
According to all energy scenarios for the coming decades, made up by leading instances (e.g. White Paper European Commission), energy agencies (e.g. International Energy Agency) and R&D-institutions indicate that the part of renewable energy sources within the future energy generation schemes is expected to grow. Photovoltaic energy generation will become a relevant part of the total electrical energy generation after 2030 and, eventually, photovoltaic generation could be supplying 20-30% of the global electrical energy demand in 2100.
The number of solar cells being produced yearly (in terms of Watts produced by these solar cells under a standardized spectrum) is growing consistingly the last 10 years with growth rates between 15 and 60% as shown in figure 1.
Figure 1: Evolution of yearly production of photovoltaic modules and yearly increase (in red).
In 2007 this growth has continued with a total production level over 3GWp representing a total turn-over of the sector well over 15 Billion Euro.
The present cost of a photovoltaic system is between 4 and 6€/Watt, which in North-Western Europe translates in a electricity cost of 0.5-0.6€/KWh, whereas in Southern Europe and regions with an equivalent amount of sunshine, this cost can be divided by a factor of two. Reducing the cost by a factor 4 to 5 is to be done by economies of scale (larger plants) but will also require technological breakthroughs to reduce the cost of materials to manufacture cells and modules, reduction of energy input to realize these components and an increase of the energy conversion efficiency.
At IMEC this objective is translated into the four photovoltaic technologies the PV-Program is working on: thin crystalline Si solar cells, organic solar cells, high-efficiency photovoltaic stacks for terrestrial concentrators (for solar intensities between 100 and 1000 suns) and thermophotovoltaics for electricity co-generation purposes.
In parallel with the appearance of distributed electricity generation, one predicts a similar evolution for electronic systems, an evolution often described by the term “ambient intelligence”. Most of these electronic systems contain a sensing part associated with data processing capability as well as RF-features for data communication. Within this vision, ensuring the energy autonomy of freestanding and (or) portable circuits, is a crucial task. It turns out that, even at low illumination levels of typically 0.1-1% of standard sunlight, photovoltaic cells are the most obvious means to ensure this required energy autonomy with lowest area or volume requirements. For this purpose high-efficiency backside-contacted Si solar cells and flexible organic solar cells represent attractive solutions.
The whole of the solar cell activity at IMEC as well as the drivers of the different technologies can be represented as shown in figure 2.
Figure 2: Schematic representation of the technologies under development at IMEC as a function of illumination intensity as well as its drivers.
Crystalline Si solar cells
Crystalline Si solar cells have been and still are the workhorse of the photovoltaic industry with a market share of about 90% of the total world solar cell production in 2007. IMEC has developed an evolutionary roadmap which predicts a gradual transition from thin wafer-based crystalline Si solar cells towards thicknesses as low as 40µm or monolithic thin-film crystalline Si solar cell modules on a low-cost carrier. This is schematically shown in figure 3. This scheme is slightly different from last year's scheme in that recent innovative approaches under development at IMEC allow the realization of very thin Si-foils with thickness as low as 50µm. The development activities within the crystalline Si solar cell program are oriented towards the development of solar cell processes for very thin Si-substrates (<200µm down to 40µm), advanced backside-contacted solar cell technologies, the implementation of industrial cell techniques for epitaxial thin-film crystalline Si solar cells and innovative approaches for thin-film polycrystalline Si solar cells which. All this should reduce the cost of crystalline Si solar cells by at least a factor of 3 compared to nowadays and a reduction by an order of magnitude of the amount of Si/Watt - which presently is slightly below 10g/Watt.
Figure 3: Schematic representation of the IMEC-vision for crystalline Si solar cell technologies.
One of the IMEC-highlights of the last years is the development of a new industrial process perfectly tailored to thin crystalline Si substrates. This process is named the i-PERC process and relies on local rearside contacts and a dielectric passivation stack in between the rear contacts. All the steps are in principal compatible with the high throughputs needed for industrial PV-production and screenprinted metallization. Using the i-PERC process an efficiency of 17.6% was obtained on large-area Si solar cells (thickness: 130µm), which is the best ever for such thin cells with screenprinted contacts C15031. Efficiencies near 19% with currents of nearly 40mA/cm2 obtained on small-area cells confirm the efficiency potential of this approach.
IMEC is also developing a proprietary process for thin-film crystalline Si solar cells, epitaxially grown on very low-cost Si substrates based on metallurgical grade Si. he inclusion of special types of buried Bragg reflectors based on porous Si allows substantial current increases allowing large-area epitaxial cells with efficiencies near 14% C14754. In addition the use of epitaxial emitters allows to improve substantially the blue response of thin-film epitaxial cells resulting in an efficiency of 15% C15060.
Also very encouraging results were obtained in polycrystalline Si layers on ceramic and high-temperature glass. By an innovative emitter approach the effectiveness of the hydrogen passivation of the polycrystalline Si layer is increased which resulted in open-circuit voltages ≈540mV, the highest ever reported for this type of material, and efficiencies near 8% P15573, P14865, P14254 whereas on high-temperature glass substrates an efficiency of 6.7% was obtained.
Organic solar cells
Active layers of organic solar cells are typically in the order of 100nm to several µm's. The low material consumption and the fact that the technologies to deposit these layers (printing) are compatible with extremely high production throughputs (1 to 2 orders of magnitude in comparison with the present solar cell technologies) could result in costs a factor 5 to 10 lower than the present solar cell technologies. One of the most promising concepts in the field of organic solar cells is that of the bulk donor/acceptor heterojunction. Here, the active layer consists of an intimate mixture of two different conjugated organic materials sandwiched between metallic electrodes.
Investigations on this device concept at IMEC are focused along two main routes. The first one considers polymer-based organic solar cells in which the active layer can be processed from solution. Typically spin coating is used, though successful steps have already been made towards the introduction of printing technology like screen printing. Secondly, active organic layers can be deposited by vacuum evaporation if small conjugated molecules are considered. The upscaling of the evaporation of small conjugated molecules is pursued by means of the organic vapor-phase deposition (OVPD) technique. The roadmap in terms of activities, efficiency objectives and stability is shown in figure 4. Details about the results in 2007 are to be found in the part of the scientific report 2007 (see section Organic photovoltaics).
Figure 4: Roadmap of the Organic Solar Cell activity at IMEC for the three major issues: efficiency, stability and the development of a cell & module manufacturing technology.
On the route of polymer-based solar cells, a lot of effort has initially gone into the improvement and understanding of bulk donor-acceptor heterojunction cells based on blends of PPV-polymers (poly-para-phenylene vinylene) with a soluble derivative of C60. This resulted in efficiencies near 3.5%. Gradually, the focus of the organic solar cell research has moved to P3HT (poly-3-hexylthiophene) as donor material because it has a higher absorption coefficient close to the maximum photon flux in the solar spectrum. In 2007, careful optimization of the solvent, the thermal treatment, the evaporation conditions of the solvent and the use of a novel rearside contact allowed to realize P3HT:PCBM solar cells with efficiencies near 5%. These results were obtained for spincoated layers, but the final technology will rely on linear casting approaches. In the past attempts were already made by screenprinting the active layer, but the viscosity of the P3HT:PCBM is too low to obtain well defined patterns. In 2007, we extended this study to inkjetprinted and sprayed active layers. Efficiencies between 2 and 2.5% were readily obtained using these methods C15855.
The second route that is strongly investigated is that of the small molecule based bulk heterojunction solar cells. Vacuum evaporation offers the possibility to control the growth and deposition of the photo-active layer in a more appropriate way by e.g. the evaporation rate of the different compounds and the substrate temperature. Subphthalocyanine was tested as donor material in a planar donor-acceptor heterojunction structure: ITO (100nm) / SubPc (14nm) / C60(33nm) / BCP(10nm) / Al(80nm). Electrical characterization demonstrated a maximum power conversion efficiency of 3% P14945. Organic vapor phase deposition (OVPD) is considered as a more cost-effective alternative to the vacuum evaporation technique. A carrier gas transports the organic material in its vapor phase through a showerhead, requiring only a rather low vacuum environment. Thereby, it allows much faster, yet controlled, thin-filmgrowth of small molecular weight organic semiconductors. In 2007, it was proven that small-molecule organic solar cells made with vacuum evaporation and OVPD showed similar performance C15146.
Photovoltaic stacks for terrestrial concentrators
The primary goal of this activity is the development of an innovative technology to produce 4-terminal high-efficiency mechanical stacks capable of efficiencies up to 35%, as shown in figure 5. This activity comprises the manufacturing of thin-film InGaP/GaAs topcells and Ge-bottomcells.
Figure 5: Concept of the high-efficiency photovoltaic stack at IMEC. The objective is to have top- and bottomcel contacts available at the same side.
In 2007, we have been successful in obtaining GaAs solar cell results at least identical to the ones obtained on germanium substrates made using standard manufacturing technologies. This was evidenced by a new record conversion efficiency of 24.7% obtained for a single-junction GaAs solar cell realized on a substrate that received an alternative surface finish/dry.
Next to the further development of the individual top and bottom cells, attention has been given in the last year to the actual process of stacking top and bottom cells together, and the handling of the thinned-down topcell (temporary bonded to a carrier) before attaching it to the bottom cell. A procedure has been developed incorporating temporary bonding of the III/V top cell to a carrier.
Present state-of-the-art multijunction solar cells use an In0.49Ga0.51P and GaAs top and middle cell, which can be grown lattice-matched on a Ge substrate. III-V materials, matched to a Ge-substrate, however do not offer an optimum bandgap combination for photovoltaic conversion; an In0.65Ga0.35P/In0.17Ga0.83As/Ge cell offers a theoretical efficiency increase of 4%. However, as in this configuration top and middle cell are no longer lattice-matched to the substrate, misfit dislocations are formed in these cells during the epitaxial growth, which reduces the minority-carrier lifetime and hence the device performance. In 2007, attention was focused on the effects of these dislocations on the performance of the tunneljunction. More details can be found in the part about III-V epitaxial layer growth for multijunction solar cells (see section III-V epitaxial layer growth for multijunction solar cells).
Low-bandgap cells for thermophotovoltaic application
Low-bandgap, stand-alone Ge solar cells have been under investigation at IMEC in the past couple of years. These cells offer numerous application possibilities, e.g. as bottom cell in a mechanically stacked multijunction solar cell or as converter in a thermophotovoltaic or co-generation system, as shown in figure 6.
Figure 6: Schematic representation of the different parts of a thermophotovoltaic system.
Whereas solar cells convert solar radiation to electricity, thermophotovoltaic cells are optimized to convert the radiation from heat sources which are at lower temperature as compared to the sun. This requires the use of materials with a lower bandgap as compared to silicon. For this purpose, several low-bandgap III-V compounds are being investigated in a number of institutes worldwide. Principally, germanium is also suited because of its low bandgap but problems related to proper surface passivation were hindering the further development. The significantly lower cost of germanium as compared to the low-bandgap III-V alternatives incited IMEC to tackle the surface passivation issue. The combination of improved surface passivation and novel contacting technologies led recently to Ge-cells with an open-circuit voltage over 270 mV, an AM1.5 efficiency of 8.4% and a broad spectral response from 400 to 1700nm, values which exceed significantly the values reported by other groups P14229. The cells behaved well for concentration levels between 10 and 20 whereas initial stability tests under thermal cycling conditions showed promising results.
In order to improve the rear surface reflection (very important for efficient optical confinement) and rear surface passivation, a proprietary process based on a-Si:H passivation and local contact formation by laser treatment was developed at IMEC. This new approach led to efficiencies of 6.3% with a clear improvement of the spectral response beyond 1500nm C14643.