«Nanolayer surface passivation schemes for silicon solar cells PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Technische Universiteit ...»
Nanolayer surface passivation schemes for silicon solar cells
ter verkrijging van de graad van doctor aan de
Technische Universiteit Eindhoven, op gezag van de
rector magnificus, prof.dr.ir. C.J. van Duijn, voor een
commissie aangewezen door het College voor
Promoties in het openbaar te verdedigen
op woensdag 21 december 2011 om 14.00 uur
geboren te Tilburg
1. Solar energy
2. Silicon solar cells, loss mechanisms and surface passivation
3. Framework and goal of this research
4. Outline of this thesis
5. Summary in ten points B. ALD Al2O3 for Si surface passivation, an overview. 21
1. Introduction to surface passivation: basics and applications
2. Al2O3 properties and synthesis methods
3. Atomic layer deposition of Al2O3
4. Surface passivation properties of Al2O3
5. Implementation in solar cells
6. High-efficiency solar cells featuring Al2O3 C. Publications I. Technological
1. Stability of Al2O3 and Al2O3/SiNx stacks for surface /SiN 77 passivation of crystalline silicon
2. Silicon surface passivation by ultrathin Al2O3 films 84 synthesized by thermal and plasma ALD
3. Influence of the deposition temperature on the c - Si surface c- 90 films passivation by ALD and PECVD Al2O3 films
4. Excellent Si surface passivation by low- temperature SiO 2 low- 98 using an ultrathin Al2O3 capping film II. Fundamental
5. Influence of the oxidant on the chemical and fieldfield- 105 effect passivation of Si by ALD Al2O3
6. Hydrogen induced passivation of Si interface by Al2O3 passivation 113 and SiO2/Al2O3 stacks
7. Nanoscale interface engineering via SiO2 interlayers to 120 control the charge density of Al2O3 films for Si(100) surface passivation
8. Effect of Al2O3 composition and annealing on the 131 the hydrogenation of the Si/SiO2 interface
Summary List of publications related to this work Acknowledgements Curriculum Vitae Part A Introduction A.1 Photovoltaic energy After the discovery of the photovoltaic effect in 1839 by Becquerel, various successive scientific and technological milestones led to development of the first silicon solar cell based on a diffused p-n junction at Bell labs in 1954. The use of solar cells in areospace applications stimulated rapid technological developments. The energy conversion efficiency would soon reach 14%. Now, after more than 50 years of research and development, the market for solar energy is no longer a niche market. Tremendous progress has been made pushing the efficiency above 20% while continuously cutting costs. For the technological advancement of solar cells, progress in the understanding, development and implementation of thin functional films—the central theme of this thesis—has been crucial.
The potential for photovoltaic (PV) electricity production is immense—the sun is a virtual unlimited source of energy. PV holds the promise of clean and decentralized energy for the developed and developing world. The people in the developing world may benefit by gaining access to cheap electricity for the first time, especially given the abundance of sunlight in most of these areas. This may contribute to eradicating poverty. Yet, in SubSaharan Africa, 70% of the population has no access to electricity. Despite its huge potential, PV and other renewable energy sources remain a fraction of the worldwide electricity production with a share of ~20% in 2010 or a mere ~3%, when excluding hydroelectricity.[3,4] PV accounts for only 0.2% at present. While the world appears to be addicted to hydrocarbon energy (oil, gas and coal), the extraction of these limited resources has reached peak levels. As a consequence, the emission of greenhouse gasses in the atmosphere, most notably CO2, is on the rise and causes global-warming.[5,6] Scientists agree that rapid climate change can have dire future consequences for all life on the planet.
Yet, the energy consumption is projected to increase significantly in the coming decades and may double by 2050. To generate the capacity needed, and at the same time avert global warming, a huge opportunity for the deployment of renewable energy lies ahead.
The scale of adjustment required to move from a hydrocarbon to a renewable energy economy is tremendous. Given the urgency of this challenge, the scale-up of renewable energy production in the coming decades is a topic that deserves great attention in science, business and politics and demands for vision and concerted action.
While the global PV capacity in 2010 was only about 40 GW, or the equivalent of ~25 coal-fired plants, the growth in production volume keeps on increasing quite significantly.[7,8] In fact, solar energy is the fastest growing renewable energy source with about 40% growth annually. Figure 1 shows the increase in the production of solar cells (expressed in GW) over the last decades. It is forecasted—undoubtedly using an optimistic scenario—that the production may exceed a terawatt by 2030, which may lead to an installed capacity that could account for ~5% of electricity consumption. In the last few years, production has largely shifted from Europe to Asia and the dominance of Europe in the production of solar cells appears to be over (inset, Fig. 1). However, Europe and most notably Germany remain the largest market. Europe was responsible for 80% of the total PV capacity (16.6 GW) installed in 2010. This points to the important fact that governmental support schemes (such as Feed-in Tariffs) as implemented by some countries in Europe remain essential for realizing significant growth. At the same time, the cost
reduction due to (technological) innovation and economies of scale has been spectacular:
the costs associated with solar power have dropped by ~20% each time world PV energy supply has doubled. At present, the module price is on average below €2.5/Wp. It is expected that grid-parity, a point in time at which the costs to finance solar electricity are equal to the market price of electricity from the grid, may be reached in a number of (European) countries already by 2015.
Figure 1: Drastic increase in the global solar cell production over the last decade. The insets show the location where the cells are produced and the technology mix in the year 2010.
Regarding the technology, the commercially available solar cells fall into two categories:
wafer-based silicon photovoltaics and thin film technologies. Wafer based Si solar cells use monocrystalline or multicrystalline Si wafers. These c-Si cells had, and still have, the largest market share with currently over 80% of the total production volume (inset, Fig. 1).
It is expected that this dominant position will be maintained for at least a decade. The most important commercial thin film alternatives are CdTe, Cu(In,Ga)Se2 (CIGS) and Si thin film. The benefit of these thin films technologies is especially related to the potential for (very) low-cost manufacturing. The amount of absorber material required is typically a factor 100 less than for wafer-based Si cells. Given the terawatt challenge that lies ahead, low-costs manufacturing is essential. This is also the reason for the interest in organic solar cells, and explains recent efforts in finding unconventional earth-abundant solar cell materials (e.g. FeS2, CuO). However, another important figure of merit is the solar cell efficiency, η. This is a factor that also directly influences costs per Wp, or electricity price per kWh. Table 1 shows an overview of the record energy conversion efficiencies for research-scale photovoltaic devices and large-scale modules. The highest efficiencies are found among crystalline Si solar cells, also after integration in modules (approaching η = 20%). The significantly lower efficiencies obtained for thin film Si in conjunction with the rapidly falling prices for c-Si have led to a (temporal?) less compelling potential for this thin film technology. For CIGS and CdTe, the promising cell efficiencies achieved so far do not easily translate into production efficiency. Then again, these technologies are less mature than crystalline Si, and we may expect considerable developments in this field.
For Si cells, the main challenge is probably the development and implementation of novel technologies that increase the efficiency of large-area cells but are, at the same time, compatible with mass production at low costs. This also appears to be an area of significant scientific challenge—which can arguably compete with research on advanced nextgeneration concepts for photovoltaics—as this thesis may illustrate. Only by fundamental understanding of the mechanisms involved can the optimal technology be developed and implemented. In the end, what will have the largest impact? A 4 cm2 cell with a recordefficiency of 25.5% or GW production of 22% cells at competitive costs?
To further provide context for the work in this thesis, the next section will discuss the fundamental processes that determine the solar cell efficiency and addresses some recent technological developments in the field of c-Si cell technology.
Table 1: Record solar cell and module efficiency.
A.2 Solar cell efficiency, loss mechanisms and surface passivation For a silicon solar cell the maximum energy conversion efficiency that can be achieved theoretically is approximately ~29% (operating under 1 sun illumination).[14-19] This efficiency limit is due to fundamental loss mechanisms that are intrinsically related to the fact that semiconductors have band gaps (1.1 eV in case of Si). The fundamental losses include the thermalization losses due to the excess energy of above-band gap photons and the transmission losses due to the transparency of Si to photons below the band gap. Thermalization and transmission losses reduce the efficiency by about ~55% absolute.
Radiative- and especially Auger recombination of charge carriers represent the additional fundamental losses which reduce the efficiency further to the aforementioned ~29% limit.
The highest conversion efficiency demonstrated to date for an actual Si solar cell is η = 25% (PERL-cell, UNSW Sydney),[20,21] which is fairly close to the theoretical maximum.
For large area cells, the record has recently been set to 24.2% by Sunpower.22 The gap in efficiency between actual devices and the theoretical limit can be attributed to additional technological loss processes in the solar cell. Compared to the record-efficiencies, conventional mass-produced Si solar cells typically exhibit lower efficiencies of approximately 16-18%. By transferring, adapting and implementing the technology developed for high efficiency cells, the performance of mass-produced solar cells can be further improved. The quest for higher cell efficiencies is directly related to the ongoing reduction of the costs per Wp. Other recent technological trends to cut production costs are related to a decrease in the Si wafer thickness and the reduction of the amount of expensive silver used for the metallization (for example, using electroplated copper in the future).
Over the years, a multitude of high-efficiency solar cell device architectures has been designed and many different strategies have been developed to mitigate technological losses. A large share of the work described in this thesis is concerned with one such “loss minimization” strategy: surface passivation. Surface passivation stands for the reduction in the carrier recombination through the defect states that are abundantly present at pristine surfaces. Effective surface passivation is a prerequisite for obtaining high energy conversion efficiencies. Due to various recent developments, the implementation of (rear) surface passivation schemes in (conventional) solar cells has a prime position on the technological roadmap of many solar cell manufactures. To provide a context for surface passivation, some of the other key strategies to mitigate solar cell losses will be briefly discussed next.
Mitigating fundamental and technological losses
The fundamental and technological loss mechanisms are schematically depicted in Figs.
2a/b. Whereas technological losses can be reduced to some extent by engineering and optimizing a given solar cell concept, to overcome the fundamental losses, disruptive novel “next generation” technologies are required. For instance, to reduce thermalisation losses, hot carrier solar cells have been proposed. These cells should feature energy selective contacts in order to utilize the excess energy of the photon before it thermalises to the band gap. An approach to reduce the transmission losses is photon upconversion. By the application of an upconvertor material, which typically exhibits Er3+ ions, sub band gap photons are combined to form a high energy photon that can be absorbed by the solar cell.
Other approaches include use of photon downshifting materials or tandem cells. Although these next generation concepts hold the promise of high solar cell efficiencies and are an exciting field of research, the technological feasibility, let alone the cost-efficiency, has not been demonstrated in most cases at present.
Figure 2: (a) Band gap diagram illustrating fundamental losses: (1) Transmission; (2) thermalisation and (3) Auger and radiative recombination. Fig. (b) Technological losses illustrated for a standard ptype Si cell: Optical losses including (a1) Reflection; (a2) Shading; (a3) Parasitic absorption; (b) Electronic recombination in emitter, base and at front and rear surfaces; (c) Resistive losses. Fig. (c) High-efficiency PERL cell (Passivated emitter rear locally diffused) cell design to reduce surface recombination by implementation of surface passivation schemes, local (diffused) p+ rear BSF, selective emitter and high-aspect ratio front contacts. Fig. (d). High-efficiency back-junction back contacted cell design with n-type Si base, eliminating shading losses completely.