Patent Application
20100037937
Limited supply and increasing demand of fossil energy resources and associated global environmental damage have driven global efforts to diversify utilization energy resources and related technologies. One such resource is solar energy, which employs photovoltaic (PV) technology for conversion of light into electricity. In addition, solar energy can be exploited for heat generation (e.g., in solar furnaces, steam generators, and the like). Solar technology is typically implemented in a series of PV cells, or solar cells, or panels thereof that receive sunlight and convert the sunlight into electricity, which can be subsequently delivered into a power grid. Significant progress has been achieved in design and production of solar panels, which has effectively increased efficiency while reducing manufacturing cost thereof. As more highly efficient solar cells are developed, size of the cell is decreasing leading to an increase in the practicality of employing solar panels to provide a competitive renewable energy substitute to dwindling and highly demanded non-renewable sources. To this end, solar energy collection systems like solar concentrators can be deployed to convert solar energy into electricity which can be delivered to power grids, and to harvest heat as well. In addition to development of solar concentrator technology, development on solar cells directed to utilization is solar concentrators has been pursued.
High Intensity Solar Cell technology, referred to as a vertical multi-junction (VMJ) solar cell, is an integrally bonded series-connected array of miniature vertical junction unit cells that are edge illuminated with electrical contacts on the ends. The unique VMJ cell design can inherently provides high-voltage low-series resistance output characteristics, making it ideally suited for efficient performance in high intensity photovoltaic concentrators. Another key feature of the VMJ Cell is its design simplicity that leads to low manufacturing cost.
The efficacy of VMJ can be evidenced on performance data taken on an experimental VMJ cell with 40 series-connected junctions over the range of 100 to 2500 suns intensities where the output power density exceeded 400,000 watts/m2 at 25 volts with near 20% efficiency. It should be appreciated that the foregoing performance in VMJ solar cells is accomplished with low manufacturing cost(s) and low manufacturing complexity. Such aspects are believe to be the needed drivers for feasible technical performance and economic efficiencies needed to enable photovoltaic concentrator systems to be significantly more cost effective and viable in solving global energy problems. Furthermore any increase in cell efficiency (e.g., more watts in output) should directly decrease concentrator system size (e.g., less cost associated with bill of materials) resulting in lower $/watt photovoltaic power cost.
It is to be noted that lower $/watt cost is substantially relevant to solar cell technology adoption and market penetration since global energy demand is steadily increasing, not only in emerging but in developed countries as well, while traditional fossil fuel costs are escalating. Also there are widespread increasing concerns for all associated problems; such as environmental pollution, global warming, and national security and economic perils linked with dependency on foreign fuel supplies. These environment, economic and security factors coupled with growing public awareness are driving intense interest in finding more cost-effective and environmentally friendly renewable energy solutions. Of all available renewable energy resources, solar has the substantially greatest potential for satisfying demand in an efficient and sustainable manner. In fact, the earth receives more energy in the form of sunlight every periods of few minutes than mankind can consume from substantially all other resources over an entire year.
Even though photovoltaic power is widely recognized as an ideal renewable energy technology, its associated cost(s) can be a major impediment to adoption and market penetration. Before gaining market share and adoption, photovoltaic-based power needs to become cost-competitive with traditional power sources, including coal-fired power which is well developed, adopted among consumers and substantially cost effective. Moreover access to low cost electrical power is considered essential in all global economies; so terawatts (e.g., thousands of Giga Watts) of photovoltaic power systems can be needed. Market studies show installed photovoltaic power systems must drop to a benchmark cost of $3/watt, or less, before being cost-competitive without subsidies in large utility scale applications. Since installed photovoltaic system costs currently exceed $6/watt, substantial cost improvements are still required.
Attempting to achieve lower $/watt performance has been the principal goal of most research and development in photovoltaic technologies during the past several decades. Despite the industry spending billions of dollars pursuing a variety of technologies with the objective of rendering photovoltaic energy more cost-effective, existing photovoltaic industry still requires substantial subsidies to support sales, which can be an indicator of detrimental conditions for market development and industry development.
Currently silicon solar cells, which remain substantially the same as at the time of initial discovery and development in 1960s, dominate ˜93% of photovoltaic markets. Existing photovoltaic industry in an endeavor to lower costs has relied heavily on the availability of low cost scrap-grade semiconductor silicon to manufacture conventional solar cells. It should be noted that such scrap-grade silicon, often referred to as solar-grade silicon, is primarily the heads and tails of ingots left over from wafer production and off-spec material rejected by semiconductor device manufacturers requiring higher quality prime-grade silicon wafers. Although photovoltaic sales have increased rapidly, growing ˜40% annually over the past decade with production volume estimated at 3.8 Gigawatts (GW) in 2007, sales are now hampered by shortages and higher prices in solar-grade silicon. Although prime-grade silicon is available, it is not considered an option since it would further increase manufacturing costs several fold.
For typical conventional solar cells over half the manufacturing cost is raw semiconductor poly-silicon used to produce the wafers for solar cells. As a result, a typical 14% efficiency solar cell is rated at 0.014 Wcm−2 and has more than $3/watt (or $0.042/cm2) in silicon wafer cost before any additional manufacturing. Consequently, the existing photovoltaic industry has to address and resolve the fact that starting silicon material cost(s) alone exceeds the benchmark price utilities need for large scale applications. As a contrasting aspect, semiconductor manufacturers producing microprocessor chips that sell at over $100/cm2 on an area basis can afford cost(s) associated with utilization of prime-grade silicon wafers.
The shortages in solar-grade silicon and the photovoltaic industry's inability to achieve important benchmark cost, along with the advent of new more efficient triple-junction solar cells developed for space applications, have recently generated considerable renewed interest photovoltaic concentrators. The obvious advantage of photovoltaic concentrators is the potential cost benefit resulting from using large areas of inexpensive materials (glass mirror reflectors or plastic lenses) to concentrate sunlight onto much smaller areas of expensive solar cells, hence using cheap materials to replace expensive materials. Designing photovoltaic concentrators for 1000 suns intensity would significantly reduce expensive semiconductor silicon requirements by ˜99.9%, which means 1000 MW of VMJ cells are possible using same amount of expensive semiconductor silicon currently required for 1 MW of conventional solar cells. Pragmatically, this is considered a practical approach to alleviate any silicon shortage concern.
Substantial work on solar concentrators has mostly focused on developing silicon concentrator solar cell designs for high intensities; much of work considerable developed during the era of the 1970s energy crisis, which at the time demonstrated moderate to unsatisfactory results cost benefits. Research and development initially targeting silicon cells for concentrator systems for operation at 500 suns intensity was conducted; however that target was lowered to 250 suns when unresolved development difficulties were encountered in attempting to overcome series-resistance problems in the solar cell designs being investigated. For example, high series-resistance losses in concentrator solar cells were well recognized as being a major problem, which conventional VMJ solar cell technology has addressed and resolved. It is to be noted that a substantial portion of solar cells developed for concentrator technology are quite complex and expensive to manufacture, with 6 or 7 high-temperature steps (>1000° C.) and 6 or 7 photolithography masking steps. This complexity was attributed to design attempts to minimize series-resistance losses that basically limited maximum intensity operation in the best of these designs to no more than 250 suns. Such complexity and associated costs hindered substantial development of concentrator technologies and associated solar cell technologies, and promoted development of alternative technologies like thin-film solar cell technologies.
Vertical Multi-Junction (VMJ) solar cell technology is substantially different from conventional concentrator solar cells. The VMJ solar cell technology provides at least two advantages with respect to other technologies: (1) it does not require photolithography, and (2) a single high-temperature diffusion step, at temperatures greater than 1000° C., can be employed to form both junctions. Consequently, lower manufacturing cost is a given. In addition, VMJ solar cells can be operated at high intensities; e.g., operation at 2500 suns. It is readily apparent from such operation that series-resistance is not a problem in VMJ cell design; even at intensities an order of magnitude higher conventional wisdom suggested it was not economically viable. Also the current density in VMJ unit cells at 2500 suns is typically near 70 A/cm2, a radiation level that can be substantially detrimental to most solar cells based on other technologies.
As stated above, the renewed interest in photovoltaic concentrators is largely due to the development Triple-Junction Solar Cells made with III-V materials containing gallium (Ga), phosphorus (P), arsenide (As), indium (In) and germanium (Ge). Triple-junction cell may use 20 to 30 different semiconductors layered in series upon germanium wafers: doped layers of GaInP2 and GaAs grown in a metal-organic chemical vapor deposition (MOCVD) reactor where each type of semiconductor will have a characteristic band gap energy that causes it to absorb sunlight most efficiently at a certain color. The semiconductors layers are carefully chosen to absorb nearly the entire solar spectrum, thus generating electricity from as much of the sunlight as possible. These multi-junction devices are the most efficient solar cells to date, reaching a record high of 40.7% efficiency under modest solar concentration and laboratory conditions. But since they are expensive to manufacture, they require application in photovoltaic concentrators.
However the demand and cost of III-V solar cell materials are rapidly increasing. As an example, in 12 months (12/2006-12/2007) the cost of pure gallium increased from about $350 per Kg to $680 per kg and germanium prices increased substantially to $1000-$1200 per Kg. The price of indium which was $94 per Kg in 2002 increased to nearly $1000 per Kg in 2007. In addition the demand for indium is projected to continue to increase with large-scale manufacturing of thin-film CIGS (CuInGaSe) solar cells started by several new companies in 2007. Moreover, indium is a rare element that is widely used to form transparent electrical coatings in the form of indium-tin oxide for liquids crystal displays and large flat-panel monitors. Realistically, these materials appear not viable long term photovoltaic (PV) solutions needed to provide terawatts of low cost power in solving major global energy problems.
While III-V semiconductor solar cell of area 0.26685 cm2 may generate a power of 2.6 watts, or about 10 W/cm2, and it has been estimated that such technology may eventually produce electricity at 8-10 cents/kWh, substantially similar to the price of electricity from conventional sources, further analysis may be needed to support such estimate. However, VMJ solar cells showed output power exceeding 40 W/cm2 at 2500 suns intensity using the least costly semiconductor material with low cost manufacturing. (This output power is over 400,000 W/m2.) In addition to complex PV technologies based on advanced materials, Si-based solar cell technology remains substantially dominant in photovoltaic elements and applications. Moreover, should a global need occur, silicon is the only semiconductor material with an existing industrial base that would be capable of supplying terawatts of photovoltaic power within the foreseeable future for widespread global application.
The following presents a simplified summary in order to provide a basic understanding of some aspects described herein. This summary is not an extensive overview nor is intended to identify key/critical elements or to delineate the scope of the various aspects described herein. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is presented later.
The subject innovation provides semiconductor-based photovoltaic cells and processes that mitigate recombination losses of photogenerated carriers. In an aspect, to reduce recombination losses, diffuse doping layers in active photovoltaic elements are coated with patterns of dielectric material(s) that reduce contact between metal contacts and the active PV element. Various patterns can be utilized, and one or more surfaces of the PV element can be coated with one or more dielectrics. Vertical Multi-Junction (VMJ) solar cells can be produced with patterned PV elements, or unit cells. Patterned PV elements can increase series resistance of VMJ solar cells, and patterning one or more surfaces in the PV element can add complexity to a process utilized to produce VMJ solar cells; yet, reduction of carrier losses at diffuse doping layers can increase efficiency of solar cells and thus provide with PV operational advantages that outweigh increased manufacturing complexity. A system that enables fabrication of the semiconductor-based PV cells is also provided.
Aspects or features described herein, and associated advantages, such as reduction of recombination losses of photogenerated carriers, can be exploited in any class of photovoltaic cells such as solar cells, thermophotovoltaic cells, or cells excited with laser sources of photons. Additionally, aspects of the subject innovation also can be implemented in other class(es) of energy-conversion cells such as betavoltaic cells.
To the accomplishment of the foregoing and related ends, certain illustrative aspects are described herein in connection with the following description and the annexed drawings. These aspects are indicative of various ways which can be practiced, all of which are intended to be covered herein. Other advantages and novel features may become apparent from the following detailed description when considered in conjunction with the drawings.
The subject innovation is now described with reference to the drawings, wherein like reference numerals are used to refer to like elements throughout. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It may be evident, however, that the present invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form in order to facilitate describing the present invention.
In the subject description, appended claims, or drawings, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless specified otherwise, or clear from context, “X employs A or B” is intended to mean any of the natural inclusive permutations. That is, if X employs A; X employs B; or X employs both A and B, then “X employs A or B” is satisfied under any of the foregoing instances. Moreover, articles “a” and “an” as used in the subject specification and annexed drawings should generally be construed to mean “one or more” unless specified otherwise or clear from context to be directed to a singular form.
Moreover, with respect to nomenclature of impurity doped materials that are part of the photovoltaic cells described herein, for doping with donor impurities, the terms “n-type” and “N-type” are employed interchangeably, so are the terms “n+-type” and “N+-type.” For doping with acceptor impurities, the terms “p-type” and “P-type” are also utilized interchangeably, and so are the terms “p+-type” and “P+-type.” For clarity, doping type also appears abbreviated, e.g., n-type is labeled as N, p+-type is indicated as P+, etc. Multi-layer photovoltaic elements or unit cells are labeled as a set of letters, each of which indicates doping type of the layer; for instance, a p-type/n-type junction is labeled PN, whereas a p+-type/n-type/n+-type junctions is indicated with P+NN+; labeling of other junction combinations also adhere to this notation.
The subject innovation relates to improving performance of photovoltaic cells, e.g., solar cells, particularly high-intensity solar cells such as edge-illuminated or vertical junction structures that can produce a substantially high power output under high intensity radiation levels. Various designs of PV elements that form unit cells employed to fabricate VMJ photovoltaic cells are set forth herein unit to reduce recombination losses of photogenerated carriers via patterned contacts.
The VMJ cell has an inherent theoretical limit efficiency exceeding 30% at 1000 suns intensity so further performance improvements are possible using experimental understanding and insight from computer simulations and modeling analysis. Although conventional one-sun solar cells are easily modeled with good results using analytical equations, such is not the case for edge-illuminated VMJ cells at operating at high intensities, because at high intensities, even second order effects can have substantial effect(s) on the cell operating efficiency. While aspects or features of the subject innovation are illustrated with solar cells, such aspects or features and associated advantages, such as reduction of recombination losses of photogenerated carriers, can be exploited in other photovoltaic cells, e.g., thermophotovoltaic cells, or cells excited with laser source(s) of photons. Moreover, aspects of the subject innovation also can be implemented in other classes of energy-conversion cells such as betavoltaic cells.
The physics of electron-hole carrier pairs produced in solar cells at high intensities is rather complex as many physical parameters come into play, including, but not limited to: surface recombination velocities, carriers mobility and concentrations, emitters (e.g., diffusions) reverse saturation currents, minority carrier lifetimes, band gap narrowing, built-in electrostatic fields, and various recombination mechanisms. Mobility decreases rapidly with increasing carrier density and Auger recombination increases rapidly with intensity as the cube of the carrier density. To incorporate such aspects into modeling of VMJ solar cell performance, computer simulations (e.g., two-dimensional numerical computational analysis of photogenerated carrier transport in a semiconductor) can provide insight into physical parameters in vertical junction unit cells, or PV elements, operating or for operation at high intensities. Such simulations provide an analysis and design instrument to understand possible sources of performance efficiencies and to increase performance of VMJ cells at high intensities. It should be appreciated that while even though conventional one-sun solar cells are easily modeled with good results using simple analytical equations, such is not the case for edge-illuminated VMJ photovoltaic cells operating at high illumination intensities, because at high intensities, even second order effects can have a dramatic effect of the cell operating efficiency
Computational simulations based upon models of contact-to-contact VMJ unit cells that incorporate an array of semiconductor physics reveal specific regions in VMJ unit cells where recombination losses of photogenerated carriers occur at high intensities. At least some of such regions present complex loss mechanisms that are intensity dependent. Computer simulation(s) reveal regions in PV elements, or VMJ unit cells, that can be improved upon in order to reduce recombination losses and improve performance of VMJ cells. Aspects of the subject innovation provide such improvements.
Series resistance has been considered a significant source of design issues for conventional concentrator solar cells. The VMJ photovoltaic cell design proved more than adequate in this regard, showing series resistance is not a problem even at 2500 suns intensity. However, in some situations, it can be advantageous to tradeoff an increase in series resistance for less design simplicity, in order to improve efficiency of VMJ photovoltaic cells for photovoltaic concentrators operating near 1000 suns.
It should be appreciated that design for operation at substantially higher intensities, such as 2500 suns where VMJ cells are still capable of operating efficiently, can require substantially more demanding and expensive concentrator system engineering in optics, structures, sun tracking, and thermal control, while not likely contributing any better overall performance or economic benefits. Therefore, aspects or features of solar cells, and associated process(es) for production thereof, set forth in the subject innovation can increase efficiency performance of high-intensity VMJ cells operating in the range of 1000 suns or higher. Efficiency increase can make VMJ solar cells or other solar cells that exploit aspects of the subject innovation more cost effective and viable, even though it can involve additional manufacturing and a potential increase in series resistance for intensities greater than 1000 suns. Aspects or features described herein can provide adequate engineering tradeoffs to make photovoltaic concentrator systems using solar cells, VMJ cells or otherwise, that exploit aspects of the subject innovation more viable and cost effective in providing lower $/watt performance.
Computer modeling analysis of conventional VMJ unit cell design, e.g., P+NN+ slab with deep junctions, using realistic parameters for good silicon processing (minority-carrier lifetimes, surface recombination velocity, etc.) at intensities greater than 500 suns, showed the following percentage recombination losses for some specific regions:
Therefore, this analysis suggests the heavily doped P+ and N+ diffused emitter regions with their metal contacts account for over half of all recombination losses in unit cells that form the VMJ solar cell, and that an optimized diffused N+ emitter may be different in design from an optimum diffused P+ emitter, due in part to differences in mobility. Relative magnitude of recombination losses originated in N+ and P+ regions can be switched for N+PP+ unit cell(s), or P+NN+ unit cell(s) with shallow P+N junction(s). In an aspect, the subject innovation is directed to reducing recombination losses in the foregoing diffusion regions in order to improve the performance of VMJ cells.
High minority-carrier lifetimes and low surface recombination velocities were successfully achieved in conventional VMJ cell development with open-circuit voltage Voc=0.8 volts per unit cell junction at high intensities. Voc is determined by sunlight-generated currents and diffused emitter reverse saturation currents (Jo), with both the P+N and NN+ junctions present in the unit cell(s) of a VMJ solar cell contributing to the open-circuit voltage. The optimum junctions from an electrical point of view are the lowest Jo; using Jo=1×1−13 Acm−2, which is representative of high-quality low reverse saturation currents in diffused junctions, the analysis showed diffusion depths of approximately 3 to 10 μm are sufficient depths for both the P+ and N+ diffusions, even when considering infinite recombination velocities at the ohmic metal contacts.
It is to be noted that even though deep and gradual NN+ diffusion profiles will provide a built-in electrostatic drift field that will enhance the minority carrier movement towards the junction barrier for ultimate collection and reduce recombination in this region, computer simulations reveal NN+ junction enhancement becomes less effective at high intensities, which can result in higher recombination in N+ region as shown above.
Experiments and computational modeling and simulation have identified that prime areas for improving performance are in reducing recombination losses in the heavily doped P+ and N+ diffused and metal contacts regions for VMJ unit cells operating at high intensities. Since a high-quality oxide passivated surface can have a recombination velocity as low as a few cm/second, which is significantly less than that at the metal contacts, and considering that the drift fields created by diffusion profiles become less effective at high intensities, aspects of the subject innovation provide reduced metal contact area and diffusion area via patterned dielectric coating of PV elements, or VMJ unit cells, to improve performance of VMJ solar cells.
With respect to the drawings,
While aspects or features of the subject innovation are illustrated for an initially N-type slab of semiconductor material as precursor of PV element 110, such aspects or features can also be implemented or accomplished in an initially intrinsic, e.g., nominally undoped, precursor of PV element 110. Moreover, in alternative or additional scenarios, P-type precursor(s) can be employed: PV element 110 can be a slab of P-type doped semiconductor material that can include P+ diffuse layer 116 on a first surface, and its vicinity, of the slab and N+-doping diffuse layer 114 a second surface, and its vicinity, substantially parallel to the first surface, as described supra.
In an aspect of the subject innovation, patterned dielectric coating 120 reduces formation of metal-diffuse doping layer interface (e.g., metal and N+ layer 114 interface) upon metallization of active PV element 110—openings in a patterned dielectric coating are the regions where the metal and diffuse doping layer form an interface. Since such interfaces have high recombination losses, the reduction of the metal-diffuse doping layer contact thus mitigates nonradiative losses of photogenerated carriers (e.g., electrons and holes), with ensuing increase in photovoltaic efficiency of PV element 110. In addition, coating a PV element, e.g., 110, with dielectric material produces passivation of surface states and thus reduces surface recombination losses. Patterning of dielectric coating can be accomplished through photolithographic techniques, or substantially any other technique that allows controlled patterning of a dielectric surface; for instance, wet etching. Such photolithographic techniques generally afford pattern formation through multiple processing steps of masking and removal of the dielectric material in the dielectric coating. Alternatively or additionally, patterning of dielectric coating can be accomplished through deposition techniques, e.g., vapor coating like chemical vapor deposition (CVD) and its variations, plasma etched CVD (PECVD); molecular beam epitaxy (MBE), etc., in the presence of a mask that shadows deposited material in order to dictate a specific pattern.
It should be appreciated that dielectric coating layer 120 can adopt various planar geometries and configurations that provide electrical contact among N+-doping diffuse layer 114 and metal contact 125. As indicated supra, in example diagram 120, dielectric coating 120 adopts a square-lattice arrangement of elliptical disconnected areas. Other lattices of dielectric regions also can be formed. Such lattices can include triangular lattice, monoclinic lattice, face-centered square lattice, or the like. Alternative or additional arrangements of portion(s) of dielectric material within a patterned dielectric coating can include disconnected or connected stripes of dielectric material. It is to be noted that a patterned dielectric coating, such as coating 120, can be placed among metal contact 135 and P+ diffuse doping layer 116 (see, e.g.,
It should be appreciated that substantially any pattern of dielectric material (e.g., a disconnected array of openings, such as the space between dielectric elliptic areas in dielectric coating 120) can reduce recombination losses at a single diffuse layer (e.g., N+ layer 114) because metallization applied in a later step can assure all or substantially all open, contact areas are mutually connected when fully bonded to the next planar unit cell within the VMJ cell structure. Unit cell(s) employed to produce a VMJ photovoltaic cell as described herein consist of PV element 110 coated with a dielectric pattern and metalized as described supra. Thus, such unit cell(s) are different from conventional unit cell(s) employed for fabrication of conventional VMJ solar cells. It is noted that smaller contact area(s) amongst metal and doped layer may contribute to an increase in series resistance in a stack of PV elements such as 110 that form a solar cell; thus, an advantageous pattern for reducing the contact area ratio is a high density of closely spaced smaller openings for optimizing performance for a given intensity. Recombination losses can include radiative or nonradiative recombination of photogenerated carriers, wherein nonradiative recombination can comprise Auger scattering, carrier-phonon relaxation, or the like. Auger recombination rate increases as the cube of carrier density, e.g., density of photogenerated carriers; doubling the volume of a photovoltaic device can lead to a sixteen-fold increase in recombination losses when Auger bulk scattering in accounted for. Thus, thinner slabs 110 or substantially any design modification that renders PV element 110 thinner, such as the use of light trapping with textured surfaces, such as V-grooved surfaces, U-grooved surfaces . . . , or back side reflectors, can be utilized to mitigate bulk Auger recombination at high intensities through reduction of the thickness of unit cells that form a VMJ photovoltaic cell. Collection efficiency in PV cells can increase significantly when VMJ unit cells as designed in accordance with aspects described herein afford a 50% reduction in recombination losses.
It should be appreciated that substantially any dielectric material can be employed for dielectric coating 120. In an aspect, dielectric coating can be a thermal oxide layer, which has a low surface recombination velocity. It should further be appreciated that making electrical contacts to end of unit cells, or PV elements, of semiconductor-based (e.g., Si-based) VMJ photovoltaic cells with patterned openings in the dielectric can require a full electrical contact that can be provided by low resistivity silicon that thermally matches or substantially matches the thermal expansion coefficient of the unit cells, or a metal such as molybdenum or tungsten which have thermal coefficient(s) that nearly matches the thermal coefficient(s) of silicon. Likewise, for a VMJ solar cell based on a semiconductor material or compound other than silicon, metallization of patterned dielectric coating, e.g., 120 or 160, can be effected with conductive material(s), e.g., metals or low-resistivity doped semiconductors, that have thermal coefficient(s) that nearly matches thermal coefficient(s) of semiconductor material of the unit cells that form the VMJ solar cells.
With respect to metal layers, metal contact layer 125 and metal contact layer 135 can be disparate. For example, a first metal contact layer (e.g., layer 125) can include dopants, and a second contact layer (e.g., layer 135) can incorporate a diffusion barrier in order to mitigates autodoping.
It is to be noted that mitigation of recombination losses of photogenerated carriers and ensuing increased PV element performance provided by the introduction of the second patterned dielectric coating outweighs the added complexity and possible extra expense(s) of additional processing act(s) associated with preparation of a second patterned dielectric coating.
To ensure efficient operation of PV element 110 in a photovoltaic device, the first pattern in dielectric coating 120 is to be correlated with the second pattern in coating 160 so as to have a set of one or more opening(s), and section(s) of metal layers 125, in opposition. When patterned dielectric coating 120 is “out-of-phase” with respect to patterned dielectric coating 160, and the dielectric coatings mutually occlude section(s) of respective metal layers 125, resistance among unit cells in a stack of PV elements 110 increases and efficiency of a VMJ solar cell decreases.
Additionally or alternatively, openings formed through pattern dielectric coating 120 can be different in size, e.g., different area, that openings generated via dielectric coating 160. For instance, it can be more desirable to have the openings area for the N+ contacts wider than those for the P+ contacts in PV element 110, or P+NN+ unit cells, to more effectively reduce overall losses, particularly if there are higher losses at the N+ diffused region and metal contacts. As described above, such disparate among opening sizes can be implemented or exploited irrespective of the particular pattern of the dielectric coating.
In addition, the monolithic stack of PV elements 2701-270M that produces the VMJ solar cell can be processed, e.g., sawn, cut, etched, peeled, or the like, in order to expose or nearly expose a specific crystalline plane (qrs), with q, r, s Miller indices, which are integer numbers, to sunlight when the VMJ solar cell is part of a PV module or device. In an aspect, to achieve substantive passivation of surface states, specific crystalline plane(s) can (100) planes.
At fabrication of vertical multi-junction solar cell(s), which includes stacking and alloying surface-patterned PV elements described herein, the differently oriented, dielectric areas when bonded together with metallization can form low-resistance contact points in a defined pattern. In an aspect, the contact points, facilitated through the openings in dielectric coatings 330 and 340, are directly aligned and mutually adjacent in a controlled pattern, with P+ contacts of one wafer interfacing at points to N+ contacts of the next wafer in order to keep series resistance low in finished VMJ cells. As described supra, in an aspect, fabricated VMJ cells can be sawn to have a preferred <100> crystal orientation at the illuminated surface in order to establish the lowest surface states for passivation. Thus, as illustrated in the
Upon metallization, e.g., surface of P+ diffuse layer 416 and patterned surface of confined, disconnected N+ diffuse doping region (e.g., set of regions 414) are coated with a metal contact, a set of metalized PV elements can be stacked, and processed, e.g., soldered or alloyed through a high temperature manufacture step, to form a VMJ photovoltaic cell with reduced recombination losses in accordance with aspects described herein.
In view of the example systems and elements described above, example methods that can be implemented in accordance with the disclosed subject matter can be better appreciated with reference to flowcharts in
At act 830, an ohmic contact is deposited onto one or more of the patterned surfaces of the PV element. The material for the ohmic contact, can be embodied in substantially any or any conductive material, e.g., a low-resistivity doped semiconductor or a metal. In an aspect, the conductive material nearly matches the thermal coefficient(s) of the semiconductor material e.g., Si; Ge; GaAs, InAs, or other III-V semiconducting compounds; II-VI semiconducting compounds; CuGaSe; CuInSe; CuInGaSe . . . , of the PV element and is suitable for alloying. As indicated supra, pattern(s) of dielectric material coating(s) ensures that deposition of an ohmic contact onto opposing patterned surfaces results in regions of low electrical resistance by aligning metalized regions on disparate surfaces (e.g., 90 degree-misoriented striped openings in patterns 530 and 540 result in metal contact regions aligned along a stacking direction (e.g., z direction 280).
At act 840, a set of patterned, metalized photovoltaic elements is stacked to form a solar cell. The set of photovoltaic elements that form the solar cell spans M elements, with M a natural number determined at least in part by a target operation voltage of the solar cell. In an aspect, the set of PV elements can be homogeneous or heterogeneous. In a homogeneous set each element, or unit cell, in the set is based on the same or substantially the same precursor, whereas in a heterogeneous set each element is based on disparate precursors. Disparate precursors can be based on the same semiconducting compounds, e.g., Si; Ge; GaAs, InAs, or other III-V semiconducting compounds; II-VI semiconducting compounds; CuGaSe; CuInSe; CuInGaSe, but differ in doping type or, for alloyed compounds, in alloying concentrations. In addition, such patterned, metalized PV elements include confined regions of diffuse doping as discussed above. At 850, the solar cell is processed to facilitate deployment in a PV device, optimize photovoltaic performance, or a combination thereof. Processing can include various manufacturing steps or procedures such as cutting procedures, polishing procedures, cleaning procedures, integrating procedures, or the like. Such steps can be intended, at least in part, to expose a specific crystalline plane to sunlight when the formed solar cell is deployed in a PV device. In one example, processing comprises cutting the formed solar cell(s) so as to expose or substantially expose (100) crystal planes to sunlight in order to establish the lowest surface states for passivation. It should be appreciated that the solar cell can be processed to expose or substantially expose other crystal planes, e.g., (qrs) planes such as (311).
As described above, coating a PV element with a dielectric material can occur prior or subsequent to doping. Doping subsequent to patterned dielectric coating exploits such coating as a mask for generation of confined or nearly-confined doping regions (see, e.g.,
Metallization module(s) 916 enables deposition of metallic layer(s) to a PV element that includes doping regions, extended or confined, and patterned dielectric coating(s). Metallization can be accomplished through deposition of semiconductor material with subsequent doping, or a metal material. In an aspect, such materials have thermal coefficient(s) that matches or nearly matches thermal coefficient(s) of PV element with doping regions.
Deposition reactor(s) 910 can include sputtering chamber(s), epitaxy chamber(s), vapor deposition chamber(s); electron beam gun(s); source material holder(s); wafer storage; sample substrate; oven(s), vacuum pump(s); e.g., turbomolecular pump, diffusion pump; or the like. In addition, deposition reactor(s) 910 can include computer(s), including processor(s) and memories therein, with memories being volatile or non-volatile; programmable logic controller(s); dedicated processor(s) such as purpose-specific chipset(s); or the like. Deposition reactor(s) 910 also can include software application(s) such as operating system(s), or code instructions to effect one or more processing acts, including at least those described supra. Described hardware, software, or combination thereof, facilitate or enable at least a portion of the functionality of deposition reactor(s) 910 and module(s) therein. A bus 918 allows communication of information, e.g., data or code instructions; transfer of materials; exchange of processed elements; and so forth, amongst the various hardware, software, or combination(s) thereof, in deposition reactor(s) 910.
Photovoltaic element(s) can be supplied to a package platform 930 for further processing. An exchange link, e.g., a conveyer link, or an exchange chamber and electromechanical components therein, can supply the PV element(s); at least one of the exchange link or exchange chamber illustrated with arrow 920. Assembly module(s) 932 can collect a set of PV element(s) and allow stacking of each of the PV elements through a high-temperature process or step in order to produce a solar cell, e.g., a VMJ solar cell. The stack is transferred to a specification module(s) 934 that completes the solar cell to a determined specification, e.g., the stack is sawed to allow exposure of a particular crystalline plane of the PV elements in the stack that form the solar cell. Such processing can be facilitated or allowed, at least in part, by test module(s) 960, which can determine crystallographic orientation of the PV elements, or unit cells, in the solar cell; such determination can be established via X-ray spectroscopy, e.g., diffraction spectrum and rocking curve spectra.
For quality assurance or to meet specifications, test module(s) 960 can probe precursor materials or processed materials various stages of solar cell manufacturing. As an example, test module(s) 960 can probe density of openings in a patterned dielectric coating of PV element(s) to determine whether such density is adequate for an expected sunlight intensity, or photon flux, in a solar concentrator. As another example, test module(s) can determine defect density that can arise from thermal cycling in a PV element with metallic layers, to establish if the material or process utilized for metallization is adequate. To at least such ends, test module(s) 960 can implement or enable minority-carrier lifetime measurements, X-ray spectroscopy, scanning electron microscopy, tunneling electron microscopy, scanning tunneling microscopy, electron energy loss spectroscopy, or the like. Probe(s) implemented by test module(s) 960 can be in situ or ex situ. Samples of precursor of processed materials or devices, e.g., solar cells, can be supplied to test module(s) via exchange links 940 and 950.
Processing unit(s) (not shown) can effect logic to control at least part of the various processes described herein in connection with operation of system 900. Such processing unit(s) (not shown) can include processor(s) that execute code instructions that effect the control logic; the code instructions, e.g., program module(s) or software applications, can be retained in memory(ies) (not shown) functionally coupled to the processor(s).
What has been described above includes examples of systems and methods that provide advantages of the subject innovation. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the subject innovation, but one of ordinary skill in the art may recognize that many further combinations and permutations of the claimed subject matter are possible. Furthermore, to the extent that the terms “includes,” “has,” “possesses,” and the like are used in the detailed description, claims, appendices and drawings such terms are intended to be inclusive in a manner similar to the term “comprising” as “comprising” is interpreted when employed as a transitional word in a claim.
This application claims the benefit of U.S. Provisional Application Ser. No. 61/089,389, filed Aug. 15, 2008 and entitled “SOLAR CELL WITH PATTERNED CONTACTS,” the entirety of which is incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| 61089389 | Aug 2008 | US |
