A Seminar Report On Losses in Solar Cells by
A Seminar Report On Losses in Solar Cells by
Energy Systems Engg Indian Institute of Technology Bombay, Powai, Mumbai May 2006
Abstract Direct energy conversion from sunlight to electricity is obtained through solar cells. The present eciency limit and cost of the solar cell are not in comparison with conventional energy sources. Eciency of solar cells is about 24.7% in laboratory and 14% comercially. Low eciency is due to high percentage of losses. Reection loss, resistive loss, recombination loss and thermal loss, tend to reduce the eciency of solar cell. The present report discusses in detail the dierent mechanisms of losses occuring in the solar cells. Also techniques that are implemented to reduce, losses are addressed briey.
Contents
1 Introduction 2 Reection Losses 3 Recombination Losses 3.1 Surface recombination . . . . . . . . . 3.2 Depletion region recombination . . . . 3.3 Bulk recombination . . . . . . . . . . . 3.4 Recombination at Metal Semiconductor 1 3 6 8 9 9 10 11 12 13 16 17 18 19
. . . . . . . . . . . . . . . contacts
. . . .
4 Series Resistance Losses 4.1 Emitter resistance . . . . . . . . . . . . . . . . . . 4.2 Metal semiconductor contact . . . . . . . . . . . . 4.3 Metal ngers and Bus bars . . . . . . . . . . . . . 5 Thermal Losses 6 Eciency Dependence on Losses 7 Conclusion
List of Figures
1 2 3 4 5 6 7 8 9 10 11 12 13 Eciency vs Band-gap energy[L.L.Kazmerski, 2006] 1 Losses in solar cell . . . . . . . . . . . . . . . . . 3 Internal Reection . . . . . . . . . . . . . . . . . 5 Eect of ARC [Mller, 1993] . . . . . . . . . . . . 6 o reectivity of RTO SiN stack and other ARC [J.Y.Lee and S.W.Glunz, 2006] . . . . . . . . . . . 6 Recombination process . . . . . . . . . . . . . . . 7 Surface eld . . . . . . . . . . . . . . . . . . . . . 10 PERL and PERT solar cells [J.Zhao, 2004] . . . . 11 Resistance in Solar cell [R.J.Handy, 1967] . . . . . 12 Current ow emitter . . . . . . . . . . . . . . . . 12 Eect of junction thickness and doping level on eciency[J.C.Dur`n et al., 1991] . . . . . . . . . . 13 a Ohmic contact . . . . . . . . . . . . . . . . . . . . 14 Eect of concentration ratio and Rc on eciency and loss in solar cells [T.A.Gessert and T.J.Coutts, 1992] . . . . . . . . . . . . . . . . . . . . . . . . . 15 Contact resistance for dierent grid thickness [Y.K.Fang et al., 1979] . . . . . . . . . . . . . . . . . . . . . 15 a). Interdigitated Back contact solar cell, b). Burried contact solar cell c). H grid design solar cells . . . . . . . . . . . . . . . . . . . . . . . 17 Thermal loss . . . . . . . . . . . . . . . . . . . . . 18 Losses in solar cell [Markvart, 1994] . . . . . . . . 19
14 15
16 17
ii
List of Tables
1 2 Work function of metal semiconductor [S.M.Sze, 1969] . . . . . . . . . . . . . . . . . . . . . . . . . 14 Losses aecting the parameters . . . . . . . . . . 18
iii
Introduction
Renewable energy sources are gaining importance due to increase demand, cost and failure of conventional energy sources to meet the energy requirements of various sectors. Renewable energy sources like Solar, Wind, Bioenergy, Ocean thermal etc are used for energy generation. Important is the electrical energy generation from these sources. Direct electricity generation from sunlight is possible with the help of Solar Cells(cells). Advantage of solar cells is, they do not have any mechanical parts which makes them easy to use, maintenance free and leading to longer life. Silicon (Si) is widely used semiconductor material for solar cells. Advantage of Si over other semiconductor devices is due to well developed microelectronics industry which has considerable knowledge of working with Si. This makes Si a better candidate for solar cells as compared to other semiconductor material such as gallium arsenide(Ga As) or germanium(Ge). Fgi.1 shows the maximum theoretical eciency vs band-gap at air mass(1.5) of various semiconductor materials. Si which has band gap of 1.1ev has eciency in the range of 26% as com30 28 26 Efficiency (%) 24 22 20 18 16 14 12 0.5
GaAs Si
aSi:H
Efficiency vs Bandgap
Ge
1 1.5 Bandgap (ev) 2 2.5
Figure 1: Eciency vs Band-gap energy[L.L.Kazmerski, 2006] paerd to the GaAs which has 30% eciency, but due to good knowledge of processing and lower cost, Si is a better choice. Solar cells reaching eciency of 25% have been fabricated at UNSW Australia. Research is continuing to improve the eciency upto its theoretical limit. Which could be achived by reducing the losess in the solar cells which are quite high as compared to con1
ventional power sources. Solar cells manufactured through industrial process have eciencies of the order 12 15% this shows about 88 85% losses occur in solar cells. If these losses are overcome, eceincy improvement could be obtained. It becomes necessary to study dierent losses that occur in the solar cells and the methods which can be implemented to reduced them. Some practices which are followed to reduce the losses are, better material selection, proper manufacturing techniques or changing design procedure of the cell. Dierent mechanism govern the losses in solar cell some of them are mention below, also Fig.2 gives the regions in the solar cell were these losses occurs. 1. Reection losses: Occurs at top surface of the cell were light is incident. Reection of light results in low absorption of photons in the solar cell. 2. Recombination losses: Occurs every were in the volume of the solar cells. Carriers generated get recombine with each other in order to maintain equilibrium condition. Areas were these losses occur in large magnitude are Bulk region(base region), Top surface, Metal to semiconductor contact areas, Junction region. 3. Series resistance losses: Resistance contributed by the Metal ngers, Metal to semiconductor contact resistance, Bus bar, Emitter region and Bulk region, is called as series resistance. Voltage drop and power loss results due to high value of series resistance and hence reduce eciency. 4. Thermal losses: A very small quantity of light absorbed by the cell is utilized in generation of power, remaining photon energy goes untiled in the form of heat which increases the temperature of the cell. Cell parameters such as open circuit voltage(Voc ), short circuit current(Isc ) are function of temperature.
Top Surface Loss Recombination at Top surface Recombination at Metal Semi contact Recombination at junction Recombination in Bulk region Recombination at Metal Semi contact
Series resistance due to emitter region Series resistance due to Bulk region
Figure 2: Losses in solar cell The present report gives an brief description of various mechanism which causes losses in solar cells. Also techniques that are used to reduced these losses are discussed. The discussion is limited to the Si solar cells. Though same losses occur in other material used for solar cells but nature of the loss and their percentage may be vary.
Reection Losses
The reection losses occurs from top surface of the solar cells which receives the light. Reection losses aect the Isc short circuit current of the solar cell. Reection reduces the absorbed carriers and hence the Isc . It becomes necessary to improve the absorption and reduce reection to improve short circuit current. For a bare Si these losses account for more than 30%. Photons striking the top suface are reected due to high reectivity of Si(0.3)[Green, 1982] in the UV and visible region, resulting in absorption of a very small portion of the incident light leading to poor eciency. The reectivity of Si is given as [Mller, 1993] o R()Si = (nSi () 1)2 + ()2 (nSi () + 1)2 + ()2 (1)
As seen in Eq.1 the reective losses depends on the refractive index (nSi ()) of the Si and on the extinction coecent(()), which are function of wavelength () of the light incident on surface. Typical value for refractive index of Si is (Si = 3.9) and extinction coecent ( = 0.05) [Green, 1987] these values are in visible region of the solar spectrum. Depending on these pa3
rameters, photon ux ((x, ) absorbed by the Si is given as [Mller, 1993] o (x, ) = [1 R()Si ]0 ()exp()x (2)
were R()Si is the loss due to reection. 0 is the number of photons incident on the surface of Si. () is the absorption coecent of Si and is dened as the number of photon absorbed by Si for a particular wavelength (). Absorption coecient is a property of material and cannot be modied. () in visible spectrum is about 6 103 /cm [Green, 1987]. The number of carriers that are absorbed by the solar cells is given by the quantum eciency dened as Internal Quantum Eciency(IQE) and External Quantum Eciency(EQE), they indicate the amount of the carriers generated due to the photon absorption. External quantum eciency is the current generated due to the photon absorption at the surface and is given as EQE = J q() (3)
were J is the current density generated in the cell, () the photons absorbed , q the charge of the electron. Internal quantum eciency is due to the photons incident on the surface and is given as EQE IQE = (4) 1 R() T () were, R() is the reectivity of the Si, T () is the transmitivity of the Si. Carrier generation rate G(x, ) depends on the IQE, absorptions coecient (()) and the incident ux () and is given accordingly G(x, ) = (IQE)()()[1 R()]exp()x (5)
The losses due to reection in a bare Si are about 30% as discuss earlier. To reduce the reectivity in solar cells a common approach is, use of an antireective coating (ARC 60nm) deposited using CV D (chemical vapour 4
deposition) and texturing (which is in the form of pyramids usually formed by etching the surface with the acid(H2 SO4 or HN O3 in H2 O2 )) on top surface of the solar cells. The relective losses occuring in the solar cells account to about 15% for properly textured solar cells a reduction in the reective losess from 30%to15%. Materials used for ARC are SiO2 or Si3 N4 earlier M gF2 (n = 1.38) or ZnS(n = 2.25) [M.A.Green et al., 1984] were used. A better surface passivation is obtained by using Si3 N4 as compared to SiO2 as it gives the better adhesion property with the N type Si material. Reection of photons also occurs from the back surface of the solar cell i.e when the photon strikes the cells back surface it gets reected from the back surface and in some cases may also go out of the cell. Such type of reection loss is neglegible in crystalline Si cells due to larger thickness of the base material. Proper thickness and refractive index are important physical parameters required for good ARC. Si3 N i4 has refractive index (nSi3 N i4 = 2) and SiO2 (nSiO2 = 1.4) in the visible region. Total internal reection is desired when the photons strikes the surface of cell (interface between the Si and the ARC) as shown in the Fig.3 The refrac-
Air
ARC
Si
>
Si
ARC
>
air
Figure 3: Internal Reection tive index and thickness of the ARC material has to be adjusted to achieve internal reection. The idea is to obtain the thickness t of ARC layer in such a manner that the light which is reected from the surface of ARC and 5
from the interface of ARC and Si is out of phase, resulting in destructive interference. So maximum possible light is absorbed by the Si. The total internal reection is given as, t= cos 4 ARC (6)
Also for internal reection to occur the refractive index of the ARC has to be n2 (7) ARC = nair nSi and the Reection coecient is given as Rmin = {
2 ARC air Si 2 } 2 ARC + air Si
(8)
Use of ARC the reducing of reectivity of Si is given in the Fig.4. Due to decrease in reectivity there is an increase in absorption of photons and more generation of carriers. The reectivity of a solar cell with ARC is about 13%[Green, 1982] or less as comapred to the more than 30% for bare Si. Fig.4 also shows the reectivity of Si under glass which lies between the bare Si and the Si using ARC of Si3 N4 or SiO2 . A new technique using
rapid thermal oxidation (RTO) and Si3 N4 systems is introduce by [J.Y.Lee and S.W.Glunz, 2006]. The RTO/Si3 N4 system(consists of SiO2 and SiN ) gives better front surface passivation and low optical reectivity (5.4%) Fig.5. A solar cell fabricated by this ARC/passivating layer gives an high Voc but lower Isc and F.F. SiO2 /SiN combine stack ARC [J.Y.Lee and S.W.Glunz, 2006] is investigated for it passivating property. It is found that such stack layer gives low reectivity on textured surface as compared to other materials used for ARC.
Figure 5: reectivity of RTO SiN stack and other ARC [J.Y.Lee and S.W.Glunz, 2006]
Recombination Losses
Photon incident on the solar cell generates electron hole pairs, these generated pairs are called as carriers. Generated carriers needs to be separated before they recombine, with emission of energy. Recombination causes loss of carrier and aects the performance of the cell. Open circuit voltage Voc of the cell is aected by recombination of carriers. As recombination increases the Voc reduces. Various techniques are used to reduce the recombination in the solar cells and improve Voc . Generation of carriers is in the entire volume of the solar cell material. The carriers generated near depletion region are 7
separated out very quickly as they get swept away by the electric eld present in the depletion region. Were as the carriers which are generated away from the depletion region that is in the bulk region, on the surface, or at the back surface have less probability of getting separated. These carriers will be lost and would not contribute to the current ow if they recombine. Recombination of carriers generated in the Solar cells due to photo excitation is one of the most dominating loss occurring in the solar cell. Various mechanism which contribute to the recombination phenomenon include 1. Band to band recombination: It is a radiative form of recombination in which an electron from conduction band combines with the hole in the valence band with emission of energy (light) Fig.6. Band to band recombination occurs in direct band gap semiconductors. 2. Trap assisted recombination: Dominant form of recombination mechanism in most of the solar cells. Due to the impurities present in the semiconductor an additional energy level Et is introduce within the forbidden energy gap Fig.6. This energy level acts as a trap and captures electrons and holes, leading to recombination. Trap assisted recombination is a two step process in which electrons and hole recombine in traps and then fall back into the valence band, completing the recombination process. 3. Surface recombination: Surface of the solar cells have large number of dangling bonds due to abrupt termination of crystal structure. These dangling bond acts as recombination centers. Carriers generated at the surface fall in to the dangling bond and recombine with the hole. 4. Auger recombination: Solar cells when exposed to high intensity of photons, exhibit the phenomenon of Auger recombination. It involves three particle electron-electron-hole or hole-hole-electron. When hole from the valence band gets recombine with the electron in the conduction band the excess energy released during recombination is absorbed by the neighboring electron in the conduction band which then goes to some higher energy level and then again falls back to the conduction band with release of energy Fig.6.
Figure 6: Recombination process These losses account for major portion total input power. Dierent recombination losses that occur at dierent regions of the solar cells are given Fig.2, 1. Surface recombination. 2. Bulk recombination. 3. Depletion region recombination. 4. Recombination at metal Semiconductor contact.
3.1
Surface recombination
Surface recombination is high in Si due to the presence of incomplete bonds also called as dangling bonds. These bonds appear due to sudden disruption of crystal structure. The incomplete bonds traps the generated carriers and get recombined. Another important factor that results in the surface recombination is surface recombination velocity S(SRV), it is a function of Nst surface trapping state [S.M.Sze, 1969] Sp = p vth Nst were Nst is the number of surface trapping states. 9 (9)
vth is thermal velocity of carriers. p is capture crossection for holes. SRV is dened as the velocity with which the generated carriers recombine. As seen from the Eq.9 more are the number of Nst more is the SRV and more recombination. It is desired that the surface recombination velocity to be low. It depends on the material property and on the nature of surface as is clear from the Eq.9. Typical values of SRV for bare Si are in the range of (6 8 104 cm/s)[S.M.Sze, 1969]. To reduce the eect of surface recombination and SRV, passivation is needed at the surface. This is accomplished by depositing a layer of Si3 N4 or SiO2 at the top surface. These layers acts as ARC as well as passivating layers performing both the functions. Si3 N4 has better passivating property than SiO2 because when Si3 N4 is deposited the hydrogen from the SiH4 and from N H3 precursor lls the dangling bonds and helps in passivating the surface of Si. It is also found that the temperature at which the passivating layer is deposited aects the SRV [S.A.Ponce et al., 2005]. Higher temperature causes SRV to reduce, an SRV value of about 165cm/s at temperature is obtained by deposition of SiN at temperature of 4500 C[S.A.Ponce et al., 2005]. The passivating layer is deposited by using chemical vapor deposition(CVD). The SRV value of 165cm/s obtained by using plasma enhanced chemical vapor deposition (PECVD). As explained in section.1 the rapid thermal oxidation with SiN deposition is one of the new technique in reducing the SRV.
3.2
Recombination occurring in the depletion region is less signicant as compared to the surface recombination due to the presence of electric eld. Charge carriers generated in depletion region are separated by electric eld very quickly avoiding any chance of recombination. Any recombination occurring in the depletion region is mostly driven by the trap assisted recombination or band to band recombination.
10
3.3
Bulk recombination
Trap assisted recombination is dominating in the bulk region of the solar cell. As explained earlier the impurities if present in the semiconductor create a energy state which acts as the trap Fig.6. A model given by Shockley and Reed hall is use to represent the trap assisted bulk recombination phenomenon [P.Ashburn et al., 1975]. USHR = pn n2 i po (n + n1 ) + no (p + p1 ) (10)
were USHR is Shockley Reed-hall recombination rate the po and no are the minority carrier lifetime in N and P region, p1 ,n1 are the equilibrium carrier concentration in P and N type semiconductor. p and n are the intrinsic carrier concentration at the equilibrium state. To reduce the trap assisted recombination a high purity semiconductor material is required. Wafer grown by utilizing oat zone (F z) method have low impurity content than those from the czochkralski (Cz) grown ingots. Float zone (F z) grown wafers have higher minority carrier lifetime n and the diusion constantDn . These parameter determines the diusion length Ln for the minority carrier(n are considered as minority carriers in the P type region). (11) L n = n Dn It is desired that the minority carrier diusion length should be as large as possible, it is dene as the distance traveled by the minority carriers before they gets recombine. So higher Ln menace greater distance traveled by the carrier and more is the probability that the carriers are available for contributing to the current ow. This requires larger minority carrier lifetime which indicates bulk material having low impurities and less crystal defects. losses can be reduced with the help of highly pure material as a substrate(base region).
3.4
Metal semiconductor contacts regions provide very large recombination sites. Semiconductor and metal contact junctions are formed at both front and back 11
side of the solar cell. Back side contact contributes more to the recombination as it has more contact area with the semiconductor. Surface recombination velocity is also present at the rear contact and needs to be reduced. Rear surface passivation can reduce the SRV. Most of the high eciency solar cells like PERL(passivated emitter rear locally diused), PERT(passivated emitter rear totally diused) have used rear passivation techniques utilizing a oxide layer between the metal and semiconductor layer. Carriers generated due to light and which are separated by the electric eld are required to get collected by the metallic contact to provide power to the external load. If these carriers get recombine in the defects present at the metal semiconductor junction, there is reduction in ISC leading to power loss and reduced eciency. To reduce recombination in such areas a very strong electric eld
is created between metal semiconductor contacts. Electric eld sweeps the carriers very fast with out allowing them to recombine. This enhances the collection probability for current in the solar cells. Electric eld is created by heavily doping the semiconductor region which is in contact with the metal Fig.7. The heavily doped regions are called as the surface eld and are present at both the ends front and back were the metal semiconductor meets. Another technique that is used to reduce the recombination is the use of passivating layer at the back side contact of the cell. Passivating layer used is an oxide preferably SiO2 . The function of passivating layer is similar to that performed at the front side for reducing the SRV, which is high at the back side contact. Solar cells giving world record eciency are manufactured using front and back passivating layers Fig.8[J.Zhao, 2004]. 12
Front contact
13
Series resistance losses contribute to around less than20% of the total input power[Markvart, 1994]. But these losses inscreases trememdously when solar cell is operated at high intensities. The high intensity is obtained when a sunlight is focoused on the solar cells with the help of lenses or mirrors. The intensity of the sunlight is measured in terms of geometrical concentration ratio (X). It is dene as the ratio of aperture area to the receiver area, X= Ac Ar (12)
were Ac is the aperture area of lenses or mirrors, Ar receiver area(solar cell area) Isc is proportional to the intensity incident photons striking the solar cells, and hence on the concentration ratio (X), Isc (X) = XIsc (1) (13)
were Isc (X) is short-circuit current at X concentrations and Isc (1) is short circuit current at 1 Sun. Power losses in the solar cell, due to series resistance is in direct proportion with the X 2 [J.Coello et al., 2004]. So higher X ratio increases power loss, reduces, ll factor and eciency of the cells. Series resistance of the cell is combination of, 1. Emitter layer resistance 2. Metal-semiconductor contact(front and back) 3. Metal bus-bars and ngers 4. Bulk semiconductor resistance. The metal bus-bars and ngers, emitter layer and metal-semiconductor contact resistance contribute in large magnitude to series resistance. Bulk resistance is low due to its high conductivity. A description of various resistances in semiconductor is shown in the Fig.9.
14
R1 H A R2
R1
R1
R1 E
R3 R4 R4 R5 R5 R4 R2
R4 D
R1 :Resistance of front metal contact bus bar R2:Resistance between Emitter later and Front metal contact: R3 :Resistance of grid R4 Sheet resistance between the emitter layer and bus bar. R5 Sheet resistance between the emitter layer and grid. R6 Bulk resistance. R7 Back contact resistance ,semiconductor metal. R8 Back metal contact
4.1
Emitter resistance
Sheet resistance is the measure of resistance of thin lms such as emitter layer which has thickness 1m and is given as , R = l/A, A = t w, R = w/(t l), resistance in Rs = /t , and w/l is a unit less quantity which indicates number of squares ,l = n w, n is the number of squares. This gives the sheet resistance of the emitter layer as /square. Emitter resistance of solar cell is one of the most dominating component of series resistance of solar cell. Sheet resistance is measure of emitter resistance and is measured experimentally using four point probe method[L.K.Mak et al., 1989] and it is desired to have sheet resistance value in the range of 80 100/square. Factors aecting the emitter resistance are, thickness of the emitter layer, 15
current direction in the emitter region, current collection by the metal ngers and busboys. Increase in emitter resistance is due to thin thickness of emitter layer which is around 1m as compared with the thickness of bulk region 280m. The direction of current ow in the emitter region is lateral as shown in the Fig.10 this increases path length of current due to lateral ow in emitter. Distributed nature of resistance due to variation of contact
te
tb
area of metal ngers and bus bar [R.J.Handy, 1967]. Surface texturing has an eect on the emitter resistance [M.A.Green et al., 1984]. Resulting in increase of resistance due to increase in surface lateral area and exposure to (111) crystal planes. One approach for reducing the emitter resistance is through optimization of junction thickness. This could be achieved by suitable emitter doping concentration[J.C.Dur`n et al., 1991]. Also the dopa ing concentration and the junction depth needs to be optimize depending on which technology is used for front contact grid placement i.e screen printing or the photolithography[J.C.Dur`n et al., 1991]. According to studies a by[J.C.Dur`n et al., 1991] it is seen that the junction depth and doping dena sity does not reduce the eciency in large magnitude Fig.11. This eect can be used to increase the junction depth and reduce the sheet resistance eect in series resistance of the solar cell.
4.2
The other main component of the series resistance is the semiconductor to metal contact resistance. Metal contacts are required in the solar cells for collection of the carriers and transport them to the load to deliver power. 16
Front contact
Figure 11: Eect of junction thickness and doping level on eciency[J.C.Dur`n et al., 1991] a Metal semiconductor contacts are of two types, a]Rectifying contacts(Schotty contacts), b]Ohmic contacts. Rectifying contacts for solar cells will not be discussed, though MIS(Metal Insulator Semiconductor) cells are fabricated but their eciency is low[Markvart, 1994]. Major focus will be on the Ohmic contacts. Ohmic contact is the one which is more important for conventional solar cells. These contact is of two types, 1]Simple Ohmic contact, 2]Tunneling contact. The nature of metal semiconductor contact i.e ohmic or rectifying depends on the work function of the metal and the electron anity A of the semiconductor. A barrier is formed in both the cases and external electric eld aects the barrier height this is the case for rectifying contacts. In the case of ohmic contacts the barrier present does-not have any hindrance to the ow of carriers in both the directions. For a ohmic contact Fig.12 between metal and n+ type semiconductor the of the metal has to be less than the electron anity of the n+ material. Were as for p type material it is other way round. of metal has to be more than electron anity of p+ material. If these convention is not followed then it results in a rectifying contact. Another way of achieving the ohmic contact is by heavily doping the semiconductor region which is below the metal. The heavy doping also helps in generating surface elds which reduces the recombination at the metal semiconductor contact. Heavy doping is called as tunneling in 17
which the barrier height is reduced and the carriers can just tunnel through the junction with out any apposition in either direction from the electric eld forming an ohmic contact. Table.1 gives the Work function for various metals and semiconductors, which are used for metallic contacts in Si solar cells.
Work function volts 4.3 4.45 4.35 4.85 3.9 4.5 4.52 4
A Ec Ef Metal N
Figure 12: Ohmic contact The resistance of the metal semiconductor contact is measured with respect to specic contact resistance Rc ( cm2 ). Typically it is of the order of 105 cm2 .
18
Rc is a function of barrier height qBn in low doped semiconductor, Rc expqBn /kT for heavily doped semiconductor(tunneling) Rc is the function of doping density, Rc expaBn / Nd (14) Nd the (15)
In low doped metal semiconductor contact the specic resistance is a strong function of barrier heightBn were as in heavy doped metal semiconductor contact the Rc is a function of doping density and decreases with increase in doping density. To decrease the contact resistance(not specic contact resistance) the area of the metal to semiconductor contact needs to be ins creased so that current collection increases [Y.K.Fang et al., 1979]. When solar cells operate at high concentrations the Rc is suppose to be very low as high value would cause higher power loss and decrease eciency as shown in Fig.13. Solar cells have metallic contact at both the places top as well as
Figure 13: Eect of concentration ratio and Rc on eciency and loss in solar cells [T.A.Gessert and T.J.Coutts, 1992] bottom. The top contact is in the form of grids and bus bars were as the bottom contact is in the form of the metallic plate which covers entire back area of the solar cells. The metal semiconductor contact resistance depends 19
0.004cm
0.002cm
.0.001cm
Area=1.6x10
cm 2
Area=1.6x10 R =1.4k
cm 2
R =3.1k
Figure 14: Contact resistance for dierent grid thickness [Y.K.Fang et al., 1979] on the area of the contact. Higher the area lower is the contact resistance and vice versa. Fingers in the front contact are about 50m the number of grid lines on the front surface could be increased if 30m or less size of grid lines are fabricated. Thiner size of grid line will accommodate more grid lines in the front surface, this would improve current collection and reduce the metal semiconductor contact resistance. Also the contact resistance Rct [Y.K.Fang et al., 1979] depends on the number of grid lines n and the relation is given as Rct /n so as n inscreases the resistance decreases. This is due to reduction in current crowding at the periphery of the contact[Y.K.Fang et al., 1979] Fig.14.
4.3
Metal ngers and bus bar resistance causes considerable loss of power in the solar cells. Metal contacts are placed on the front surface and back surface of the solar cells to collect carriers and pass them to the load. Front contact metal is in the form of ne grid lines were as the back contact is a metal plate covering entire back surface. Metals used for the contact resistance are Al, T i, P d, Ag , Table.1 etc. Back contact metal is Al were as for front contact the ner grid lines are of high conductivity metal usually Ag or paste of Ag, P d or T i is used. These contacts are deposited on semiconductor by 20
using various techniques such as Photolithograpy, Evaporation, Sputtering, CVD, Screen printing, Electroplating. Conventional solar cells use screen printed contacts with Ag, Al paste. For ner grid structures of about 30m photolithography is used but this technique is very expensive and is not suitable for the solar cells. Solar cells operating at high concentration have considerable power loss in the metal grid as the losses are proportional to X 2 were X is the geometrical concentration ratio as discussed earlier. Reducing resistance is one of the major challenge in such cells. Solar cells having screen printed contacts cannot be used for concentrator solar cell due to lower metal density which results in increase resistance. To reduce the resistance of contacts, an electroplating of high conductivity metal(Ag) on screen printed contacts is a good option. Electroplating has good metal density which causes decreases in the resistance and can improve the performance of solar cells at high concentration ratio [J.Coello et al., 2004]. Shading is caused due to the ngers and bus bars on the top surface. This reduces the carrier collection. To avoid shading various other techniques are employed such as interdigitated back contact in which the both the contacts are placed at the back surface and entire front surface is exposed to light[R.A.Sinton et al., 1986]. Another approach is the burried contact solar cell which has front metallic contacts burried in the emitter layer. This improves the aspect ratio(height/width) and reduces the metal resistance[S.R.Wenham et al., 1994]. Conventional screen printed contacts cells having H grid structure are modied by using dierent ngers design to avoid shadow loss and also to reduce the metal resistance[A.R.Burgers, 1999]. It is clear from the discussion that the series resistance of the solar cell causes power loss and decrease the eciency of the cell. The possible measures that can help to reduce the series resistance are as follows, 1. High conductivity base (substrate) material oat zone(F z) , czochralski(Cz). 2. Optimizing the the junction depth for reducing the emitter resistance. Increasing the thickness to reduce the sheet resistance. 3. Increasing the number of ngers by reducing the width increases the current collection and reduces the metal semiconductor contact resistance.
21
a)
b)
c)
Figure 15: a). Interdigitated Back contact solar cell, b). Burried contact solar cell c). H grid design solar cells 4. Electroplated metal contacts to reduce the resistance of the metal contacts by increasing the metal density. 5. Dierent metalization techniques Burried contact, H design, Back contact. 6. Use of dierent techniques for depositing metal contacts at the front and back surface. This is called as the Hybrid contacts. This can reduce the cost of deposition of metals.
Thermal Losses
A major portion of loss in solar cell is a due to heat. Light absorbed by the solar cells has excess energy than that required for generation of electron hole pair(band-gap energy Eg ). This excess energy is released in the form heat Fig.16. This thermal energy causes rise of temperature of cell. The parameters that are aected by the temperature of the cell are band gap energy, diusion length, minority carrier lifetime, intrinsic carrier density. The increases in diusion length and minority carrier concentration and intrinsic carrier concentration and decrease in band gap energy causes the increases in the reverse saturation current Io Eq.16. The increase in Io reduces the
22
open circuit voltage Eq.17 which degrades the eciency of the cell. Io = qDn n2 qDp n2 i i + Ln N A Lp N D (16)
kT Isc ln (17) q Io Temperature eect is more pronounced in concentrator cells. Depending on the concentration ratio, temperature of the solar cell can rise above 1000 C. At such a high temperature the solar cell eciency decreases due to reduced Voc . If temperature rise is kept within limits with the help of proper cooling arrangements, with use of heat sinks or heat pipes, thermal losses could be maintained within limits. V oc =
excess energy as heat Eg 1.1ev Photon with energy more than Eg hole electron
Ec E ev
Ev
The eciency of the solar cells is given in terms of the Isc (short circuit current), Voc (opencircuit voltage) and the F.F (llfactor) of the solar cells. = V oc Isc F.F P in (18)
Eq.18 shows that eciency depends on Isc , Voc and F.F . These factors are solar cell parameter and every solar cell performance is measured with the help of these parameters. These parameters also indicate the losses prevailing in the solar cells as discussed. A table of losses and the parameters which 23
Table 2: Losses aecting the parameters it aects can be tabulated and is shown in Table.2 From Table.2 we can observe that any decrease in the loss will result in improving the eciency of the solar cells. Losses and the solar cell parameters are closely linked with the performance of the cell. Fig.17 shows the losses occurring in the solar cell due to various mechanisms discussed in the report.
100mW/sqcm
52mW/sqcm thermal loss 46mW/sqcm availabel Voltage 1.1V Current availabel 44mA/sqcm
Recombination
41mA/sqcm
Poutput 23mW/sqcm
24
Conclusion
Various mechanism of losses occurring in the Si solar cells are briey discussed. Recombination, Reective, Resistive, Thermal are the major losses occurring in the solar cells at 1 sun solar concentration. At higher concentration levels the resistive losses become dominant. Various schemes for reduction of losses such as Surface passivation using Si3 N4 , SiO2 for reduction in SRV at the front and back contacts, ARC for reducing the front surface reection and increase the absorption of photons, Dierent metalization techniques like Electroplated contacts, Burried contacts, H grid design for reducing the metal resistance, Optimum thickness of emitter layer to reduce the sheet resistance of the emitter layer and thus reducing the series resistance, Heat sinks for removing excess heat from the solar cell. Are some of the methods commonly applied over the years by researchers to reduce losses and improve eciency. Solar cells parameters i.e Isc , Voc , F.F are dependent on losses also these parameters decide the eciency of the solar cell. A table consisting of Type of loss and the parameter which it aects is tabulated.
25
References
A.R.Burgers. How to design optimal metallization patterns for solar cells. Progress in Photovoltaics:Research and Application, 7:457461, 1999. Martin Green. Solar Cells. Prentice Hall., Englewood Clis, N.J, 1982. Martin Green. High Eciency Silicon Solar Cells. Trans Tech Publications, 1987. J.C.Dur`n, G.Venier, and R.Weht. Optimization of junction depth and dopa ing of solar cells emitters. Solar cells, 31:497503, 1991. J.Coello, M.Castro, I.Antn, and M.A.Vzquez. Conversion of commercial o a si solar cells to keep their ecient performance at 15 suns. Progress in Photovoltaics Research and Applications, 12:323331, 2004. J.Y.Lee and S.W.Glunz. Investigation of various surface passivation schemes for silicon solar cells. Solar Energy Materials and Solar cells, 90:8292, 2006. J.Zhao. Recent advances of high eciency single crystalline silicon solar cells in processing technologies and substrate materials. Solar Energy Materials and Solar cells, 82:5364, 2004. L.K.Mak, C.M.Rogers, and D.C.Northrop. Specic conact resistnce measurement on semiconductors. Journal of Physics, Instrumentation, 22:317321, 1989. L.L.Kazmerski. Solar photovoltaic r& d at the tipping point:a 2005 technology overview. Journal of Electron Spectroscopy and Related Phenomena, 150:105135, 2006. M.A.Green, A.W.Blakers, J.Shi, E.Keller, and S.R.Wenham. High eciency silicon solar cells. IEEE Transcations on Electron Devices, 5:679683, 1984. Tomas Markvart. Solar Electricity. John Wiley and Sons, 1994. Hans Joachim Mller. Semiconductor for Solar cells. Artech House, Boston, o 1993. 26
P.Ashburn, D.V.Morgan, and M.J.Howes. A theoretical and experimental study of recombination in silicon p-n junctions. Solid State Electronics, 18:569577, 1975. R.A.Sinton, Y.Kwark, J.Y.Gan, and R.M.Swanson. 27.5-percent silicon concentrator solar cells. IEEE Electron Devices Letters, 7-10:567569, 1986. R.J.Handy. Theoreticle analysis of the series esistance of a solar cell. Solid State Electronics, 10:765775, 1967. S.A.Ponce, C.D.Canizo, A.Luque, and D.N.Wright. Surface passivation layer for multicrystalline silicon solar cells. IEEE,Spanish Conference on Electron Devices, pages 295298, 2005. S.M.Sze. Physics of Semiconductor Devices. John Wiley and Sons, 1969. S.R.Wenham, C.B.Honsberg, and M.A.Green. Burried contact silicon solar cell. Solar Energy Materials and Solar cells, 34:101110, 1994. T.A.Gessert and T.J.Coutts. Grid metallization and antireective coating optimization. Journal of Vacuum Science and Technology( Vacuum surface and lms), 10:20132024, 1992. Y.K.Fang, C.Y.Chang, and Y.K.Su. Contact resistance in metal semiconductor systems. Solid State Electronics, 22:933938, 1979.
27