4. Conclusion

In this work, we demonstrate the potential of incorporating our in-house developed industrial relevant electron-selective (thermal-SiOx/poly-Si(n<sup>+</sup> )) and holeselective (thermal-SiOx/poly-Si(p<sup>+</sup> )) passivated contacts into double-sided passivated contact solar cells. Using measured properties of our developed contact passivation layers (i.e., determining the recombination current density j<sup>0</sup> and the contact resistance Rc), we predict a practical efficiency potential approaching 24%, if device integrating them into a front-side textured, electron-extracting, and rearside planar, hole-extracting solar cell architecture, applying conventional screen printing for contact formation (using a n-type 6-inch Cz wafer with a resistivity of 3.4 Ω cm). Thus far, we have reached a solar cell efficiency of 21.7%, rear side only integrating an electron-extracting SiOx/poly-Si(n<sup>+</sup> ) passivated contact and using conventional screen printing.

(i.e., short-circuit current) loss of 1 mA/cm<sup>2</sup>

DOI: http://dx.doi.org/10.5772/intechopen.85039

5 mA/cm<sup>2</sup>

SiOx/poly-Si(p<sup>+</sup>

SiOx/poly-Si(p<sup>+</sup>

obtain 10-nm-thin poly-Si(n<sup>+</sup>

electron-extracting SiOx/poly-Si(n<sup>+</sup>

However, contacting hole-extracting poly-Si(p<sup>+</sup>

damage-free contacting of our (thick or ultrathin) poly-Si(p<sup>+</sup>

a textured silicon surface.

and a 250-nm poly-Si(n<sup>+</sup>

electron-extracting poly-Si(n<sup>+</sup>

extracting poly-Si(p<sup>+</sup>

poly-Si(p<sup>+</sup>

125

the underlying poly-Si(p<sup>+</sup>

are currently initiated.

whereas the photogeneration loss saturates at 0.05 mA/cm<sup>2</sup> for thicknesses lower than 25 nm, if rear-side integrated. A 25-nm-thin, front-side integrated

the relation between the parasitic absorption loss and the thickness of the front-side integrated poly-Si contact passivation layer is rather exponential than linear in this case). In order to develop ultrathin (≤ 10 nm) contact passivation layers, two different process methodologies have been developed: (1) using etch-back technology that is starting from 250-nmthick poly-Si "standard" layers (as described above) and applying a slow silicon etch (SSE) to reduce the thickness in a controlled way. Using this technology, we were able to obtain ultrathin 3–4-nm hole-extracting

passivation properties of the thick layers (our corresponding 3–4-nm-thin

voltage iVoc of 690 mV on a planar silicon surface). However, etch-back technology for electron-extracting passivated contacts was possible only down to a thickness of 70 nm. (2) Therefore, we re-optimized the diffusion conditions for ultrathin (10 nm) LPCVD of intrinsic poly-Si layers, which were subsequently subjected to phosphorous tube diffusion in order to

passivation layers reached an implied open-circuit voltage iVoc of 720 mV on

(II) Compatibility with conventional screen printing: For "thick" (250–150 nm)

SiNx passivated standard emitter at the front side of the solar cell, subsequently being metalized by conventional bifacial screen printing).

Ag/Al pastes (as used to contact p-doped silicon material), we observe several local "punch-through" contact regions, where the paste is completely consuming

passivation quality underneath the metal contact (i.e., there is no more contact passivation). This issue can be attributed to local aluminum alloying processes, which take place during fast-firing of Al containing screen-printing pastes: Al alloying is known to partially consume crystalline silicon material; thus, our thin

just outlined local "punch-through" effects. Therefore, the chemical composition of the screen-printing paste itself has to be altered, in order to enable a subsequent

Corresponding research activities, in cooperation with a paste manufacturer,

) capping layers will be consumed upon contact firing, leading to the

conventional screen printing creates no issue. Correspondingly, rear-sideonly passivated contact solar cells have been processed, reaching a solar cell efficiency of 21.7% (exhibiting a wet-chemically formed SiOx tunnel layer

rear side of the solar cell and exhibiting a conventional boron-diffused, AlOx/

for front-side device integration. Our 10-nm SiOx/poly-Si(n<sup>+</sup>

, which is no longer suited for device integration (please note that

) passivated contacts, which are basically maintaining the

) contact passivation layers reached an implied open-circuit

) electron-extracting capping layers suitable

)/SiNx contact passivation layers,

) layers or ultrathin

) capping layers.

) capping layer further passivated by SiNx at the

) layers is a challenge. Trying to contact hole-

) capping layer, causing a severe degradation of contact

) layers by screen printing, using conventional fire-through

)/SiNx contact

passivated contact will already exhibit a parasitic absorption loss of

Double-Sided Passivated Contacts for Solar Cell Applications: An Industrially Viable Approach…

, if front-side integrated,

Our methodology of developing/optimizing (ultrathin) contact passivation layers is outlined as follows: First, we were comparing different tunnel oxides for their suitability to form passivated contacts when capped with highly doped poly-Si, i.e., we analyzed ultrathin (<1.5 nm) industrial relevant SiOx tunnel layers (i.e., wet-chemically formed silicon oxide (wet-SiOx), UV/ozone photo-oxidationformed silicon oxides (ozone-SiOx), and in situ formed thermal silicon oxides, using low-pressure chemical vapor deposition (LPCVD) (thermal-SiOx)). Combining specifically designed lifetime and dark I–V test structures, we were able to extract the single-sided saturation current density j<sup>0</sup> and its associated contact resistance R<sup>c</sup> for our developed electron-selective and hole-selective passivating contacts. A subsequent optimization of the LPCVD of intrinsic poly-Si capping layers followed by conventional tube diffusion was undertaken, maximizing doping efficiency while minimizing in-diffusion of dopants from the poly-Si capping layer through the SiOx tunnel layer (which can act as a diffusion barrier) into the silicon wafer. After a subsequent standard SiNx passivation step, we reached implied open-circuit voltage iVoc values exceeding 730 mV for electronselective SiOx/poly-Si(n<sup>+</sup> ) passivated contacts and exceeding 710 mV for holeselective SiOx/poly-Si(p<sup>+</sup> ) passivated contacts, formed on a planar silicon surface, using an in situ LPCVD-grown thermal-SiOx tunnel layer prior the LPCVD of the (intrinsic) poly-Si capping layer, and a subsequent tube diffusion. Applying these layers on a textured silicon surface, the electron-extracting SiOx/poly-Si(n<sup>+</sup> ) passivated contact still performed well (iVoc > 700 mV), whereas the hole-extracting SiOx/poly-Si(p<sup>+</sup> ) passivated contact showed unsatisfying performance on a textured surface (iVoc 630 mV).

Subsequently, an asymmetric, front-side textured electron-extracting, rear-side planar hole-extracting passivated lifetime structure was processed, reaching an iVoc of 713 mV. Two key challenges have been identified when aiming at a double-sided passivated contact device integration of these layers:

(I) Parasitic absorption: There is always a significant amount of parasitic absorption within the poly-Si capping layer, which reduces the absorbed photogeneration current within the silicon wafer, and therefore the maximum possible short-circuit current of the solar cell. Thus, the poly-Si capping layers have to be designed to be as thin as technologically possible. This issue is even by far more important, if aiming at an additional front-side (i.e., double-sided) device integration of passivated contact. Numerical simulations (calibrated toward our developed contact passivation layers) indicate that a 10-nm-thin poly-Si layer still leads to a photogeneration

(i.e., short-circuit current) loss of 1 mA/cm<sup>2</sup> , if front-side integrated, whereas the photogeneration loss saturates at 0.05 mA/cm<sup>2</sup> for thicknesses lower than 25 nm, if rear-side integrated. A 25-nm-thin, front-side integrated passivated contact will already exhibit a parasitic absorption loss of 5 mA/cm<sup>2</sup> , which is no longer suited for device integration (please note that the relation between the parasitic absorption loss and the thickness of the front-side integrated poly-Si contact passivation layer is rather exponential than linear in this case). In order to develop ultrathin (≤ 10 nm) contact passivation layers, two different process methodologies have been developed: (1) using etch-back technology that is starting from 250-nmthick poly-Si "standard" layers (as described above) and applying a slow silicon etch (SSE) to reduce the thickness in a controlled way. Using this technology, we were able to obtain ultrathin 3–4-nm hole-extracting SiOx/poly-Si(p<sup>+</sup> ) passivated contacts, which are basically maintaining the passivation properties of the thick layers (our corresponding 3–4-nm-thin SiOx/poly-Si(p<sup>+</sup> ) contact passivation layers reached an implied open-circuit voltage iVoc of 690 mV on a planar silicon surface). However, etch-back technology for electron-extracting passivated contacts was possible only down to a thickness of 70 nm. (2) Therefore, we re-optimized the diffusion conditions for ultrathin (10 nm) LPCVD of intrinsic poly-Si layers, which were subsequently subjected to phosphorous tube diffusion in order to obtain 10-nm-thin poly-Si(n<sup>+</sup> ) electron-extracting capping layers suitable for front-side device integration. Our 10-nm SiOx/poly-Si(n<sup>+</sup> )/SiNx contact passivation layers reached an implied open-circuit voltage iVoc of 720 mV on a textured silicon surface.

(II) Compatibility with conventional screen printing: For "thick" (250–150 nm) electron-extracting SiOx/poly-Si(n<sup>+</sup> )/SiNx contact passivation layers, conventional screen printing creates no issue. Correspondingly, rear-sideonly passivated contact solar cells have been processed, reaching a solar cell efficiency of 21.7% (exhibiting a wet-chemically formed SiOx tunnel layer and a 250-nm poly-Si(n<sup>+</sup> ) capping layer further passivated by SiNx at the rear side of the solar cell and exhibiting a conventional boron-diffused, AlOx/ SiNx passivated standard emitter at the front side of the solar cell, subsequently being metalized by conventional bifacial screen printing).

However, contacting hole-extracting poly-Si(p<sup>+</sup> ) layers or ultrathin electron-extracting poly-Si(n<sup>+</sup> ) layers is a challenge. Trying to contact holeextracting poly-Si(p<sup>+</sup> ) layers by screen printing, using conventional fire-through Ag/Al pastes (as used to contact p-doped silicon material), we observe several local "punch-through" contact regions, where the paste is completely consuming the underlying poly-Si(p<sup>+</sup> ) capping layer, causing a severe degradation of contact passivation quality underneath the metal contact (i.e., there is no more contact passivation). This issue can be attributed to local aluminum alloying processes, which take place during fast-firing of Al containing screen-printing pastes: Al alloying is known to partially consume crystalline silicon material; thus, our thin poly-Si(p<sup>+</sup> ) capping layers will be consumed upon contact firing, leading to the just outlined local "punch-through" effects. Therefore, the chemical composition of the screen-printing paste itself has to be altered, in order to enable a subsequent damage-free contacting of our (thick or ultrathin) poly-Si(p<sup>+</sup> ) capping layers. Corresponding research activities, in cooperation with a paste manufacturer, are currently initiated.

4. Conclusion

Silicon Materials

selective (thermal-SiOx/poly-Si(p<sup>+</sup>

conventional screen printing.

selective SiOx/poly-Si(n<sup>+</sup>

selective SiOx/poly-Si(p<sup>+</sup>

whereas the hole-extracting SiOx/poly-Si(p<sup>+</sup>

passivated contact device integration of these layers:

SiOx/poly-Si(n<sup>+</sup>

124

integrating an electron-extracting SiOx/poly-Si(n<sup>+</sup>

In this work, we demonstrate the potential of incorporating our in-house devel-

ated contact solar cells. Using measured properties of our developed contact passivation layers (i.e., determining the recombination current density j<sup>0</sup> and the contact resistance Rc), we predict a practical efficiency potential approaching 24%, if device integrating them into a front-side textured, electron-extracting, and rearside planar, hole-extracting solar cell architecture, applying conventional screen printing for contact formation (using a n-type 6-inch Cz wafer with a resistivity of 3.4 Ω cm). Thus far, we have reached a solar cell efficiency of 21.7%, rear side only

Our methodology of developing/optimizing (ultrathin) contact passivation layers is outlined as follows: First, we were comparing different tunnel oxides for their suitability to form passivated contacts when capped with highly doped poly-Si, i.e., we analyzed ultrathin (<1.5 nm) industrial relevant SiOx tunnel layers (i.e., wet-chemically formed silicon oxide (wet-SiOx), UV/ozone photo-oxidationformed silicon oxides (ozone-SiOx), and in situ formed thermal silicon oxides, using low-pressure chemical vapor deposition (LPCVD) (thermal-SiOx)). Combining specifically designed lifetime and dark I–V test structures, we were able to extract the single-sided saturation current density j<sup>0</sup> and its associated contact resistance R<sup>c</sup> for our developed electron-selective and hole-selective passivating contacts. A subsequent optimization of the LPCVD of intrinsic poly-Si capping layers followed by conventional tube diffusion was undertaken, maximizing doping efficiency while minimizing in-diffusion of dopants from the poly-Si capping layer through the SiOx tunnel layer (which can act as a diffusion barrier) into the silicon wafer. After a subsequent standard SiNx passivation step, we reached implied open-circuit voltage iVoc values exceeding 730 mV for electron-

) passivated contacts and exceeding 710 mV for hole-

) passivated contact showed

) passivated contacts, formed on a planar silicon

) passivated contact still performed well (iVoc > 700 mV),

Subsequently, an asymmetric, front-side textured electron-extracting, rear-side planar hole-extracting passivated lifetime structure was processed, reaching an iVoc of 713 mV. Two key challenges have been identified when aiming at a double-sided

absorption within the poly-Si capping layer, which reduces the absorbed photogeneration current within the silicon wafer, and therefore the maximum possible short-circuit current of the solar cell. Thus, the poly-Si capping layers have to be designed to be as thin as technologically possible. This issue is even by far more important, if aiming at an additional front-side (i.e., double-sided) device integration of passivated contact. Numerical simulations (calibrated toward our developed contact passivation layers) indicate that a 10-nm-thin poly-Si layer still leads to a photogeneration

(I) Parasitic absorption: There is always a significant amount of parasitic

surface, using an in situ LPCVD-grown thermal-SiOx tunnel layer prior the LPCVD of the (intrinsic) poly-Si capping layer, and a subsequent tube diffusion. Applying these layers on a textured silicon surface, the electron-extracting

unsatisfying performance on a textured surface (iVoc 630 mV).

)) and hole-

)) passivated contacts into double-sided passiv-

) passivated contact and using

oped industrial relevant electron-selective (thermal-SiOx/poly-Si(n<sup>+</sup>

Furthermore, it seems that our current ultrathin, 10-nm electron-extracting poly-Si(n<sup>+</sup> ) layers are not firing stable, especially if deploying high peak firing temperatures (they still do outperform conventionally diffused front-side contacts, though). Interestingly, this is not the case for our "standard" 250-nm-thick layers. ECV measurements confirm that after contact firing (fast-firing in order to form low resistivity contacts), the dopants within the poly-Si(n<sup>+</sup> ) capping layer have outdiffused into the silicon wafer bulk, thereby effectively reducing field-effect passivation and thus the observed implied open-circuit voltage of the samples after contact firing. Thus, more efforts to render our ultrathin contact passivation layers firing stable, i.e., by deploying lower peak firing temperatures and/or changing the chemical composition of the ultrathin LPCVD of poly-Si capping layers itself, are necessary.

References

net/

[1] Fischer M. International Technology Roadmap for Photovoltaic (ITRPV). 2018. Available from: http://www.itrpv.

DOI: http://dx.doi.org/10.5772/intechopen.85039

for ohmic contacts. Nature. 2016;539:

[9] Yoshikawa K, Kawasaki H, Yoshida W, Irie T, Konishi K, Nakano K, et al. Silicon heterojunction solar cell with interdigitated back contacts for a photoconversion efficiency over 26%.

[10] Green MA, Emery K, Hishikawa Y, Warta W, Dunlop ED. Solar cell efficiency tables (version 47). Progress

[11] F. Haase and R. Peibst. 26.1% record efficiency for p-type crystalline Si solar cells. 2018. Available from: https://isfh. de/en/26-1-record-efficiency-forp-type-crystalline-si-solar-cells/

[12] Richter A, Benick J, Feldmann F, Fell A, Hermle M, Glunz SW. n-Type Si solar cells with passivating electron contact: Identifying sources for

efficiency limitations by wafer thickness and resistivity variation. Solar Energy Materials and Solar Cells. 2017;173:

[13] Zielke D, Petermann JH, Werner F, Veith B, Brendel R, Schmidt J. Contact passivation in silicon solar cells using atomic-layer-deposited aluminum oxide layers. Physica Status Solidi—R. 2011;5:

[14] Tao Y, Upadhyaya V, Chen C-W, Payne A, Chang EL, Upadhyaya A, et al. Large area tunnel oxide passivated rear contact n-type Si solar cells with 21.2% efficiency. Progress in Photovoltaics: Research and Applications. 2016;24:

[15] Loozen X, Larsen JB, Dross F, Aleman M, Bearda T, O'Sullivan BJ, et al. Passivation of a Metal Contact with a Tunneling Layer. Energy Procedia.

Nature Energy. 2017;2:17032

in Photovoltaics. 2016;24:3-11

536-540

Double-Sided Passivated Contacts for Solar Cell Applications: An Industrially Viable Approach…

96-105

298-300

830-835

2012;21:75-83

[2] Inns D. Understanding metal induced recombination losses in silicon solar cells with screen printed silver contacts.

Energy Procedia. 2016;98:23-29

Koduvelikulathu LJ, Comparotto C, Kopecek R, Harney R. Metallization– induced recombination losses of bifacial

[4] Melskens J, van de Loo BWH, Macco B, Black LE, Smit S, Kessels WMM. Passivating contacts for crystalline silicon solar cells: From concepts and materials to prospects. IEEE Journal of

[5] Ling ZP, Ge J, Stangl R, Aberle AG, Mueller T. Detailed micro Raman spectroscopy analysis of doped silicon thin film layers and its feasibility for heterojunction silicon wafer solar cells. Journal of Materials Science and Chemical Engineering. 2013;1:1-14

[3] Edler A, Mihailetchi VD,

silicon solar cells. Progress in Photovoltaics. 2015;23:620-627

Photovoltaics. 2018;8:373-388

[6] Ling ZP, Duttagupta S, Ma F, Mueller T, Aberle AG, Stangl R. Threedimensional numerical analysis of hybrid heterojunction silicon wafer solar cells with heterojunction rear point contacts. AIP Advances. 2015;5:077124

[7] Ling ZP, Mueller T, Aberle AG, Stangl R. Development of a conductive

[8] Tang CG, Ang MCY, Choo K-K, Keerthi V, Tan J-K, Syafiqah MN, et al. Doped polymer semiconductors with ultrahigh and ultralow work functions

127

heterojunction solar cells using n-doped microcrystalline silicon and aluminumdoped zinc oxide films. IEEE Journal of Photovoltaics. 2014;4:1320-1325

distributed bragg reflector for

An alternative work plan is to investigate low-temperature metallization approaches, like inline plating.

Nevertheless, despite still having to solve a suited industrial metallization scheme for our ultrathin (≤ 10-nm) in-house developed industrial electron- and hole-selective SiOx/poly-Si/SiNx passivated contact layers, due to their excellent passivation and contact resistance properties, these layers have a huge potential to get device integrated into a double-sided passivated contact solar cell architecture, which exhibits a practical efficiency potential of 23.2%, using our measured layer properties for a corresponding numerical prediction. Double-sided passivated contact solar cells deploying bifacial contacts are definitely able to outperform rearside-only passivated contact solar cells in the near future.
