Yaƙin Ƙarshe na BC-TOPCon: Tsare SiOx na Ƙasa a Daskare, Tsakaita Tsakiyar Layer da AlOx/SiOx
Teburin Abubuwan Ciki
Gabatarwar Samfur
"Tsohon Zhang, waccan takardar 26.34% na gaba-da-finger da kuma cikakken tuntuɓar baya-p biyu-Layer — mataki na gaba shi ne a sanya baya-p ya zama finger-shaped ma, kuma wannan shine ainihin BC. Amma a ƙarƙashin p-finger, waccan tsakiyar SiOx Layer: shin za mu ci gaba da amfani da SiOx mai tsafta daga 26.66, ko mu canza zuwa AlOx/SiOx don amfani da tasirin filin?" Wannan tambayar ta taso a safiyar yau, bayan da na jefa takardar 26.34% a cikin ƙungiyar layin samarwa, sai mutanen aikin BC suka fara bin ta.
A cikin gaskiyar BC-TOPCon, SiOx na ƙasa — wanda ke girma pinholes masu karewa — ba za a iya taɓa shi ba. Amma layin "SiOx na tsakiya", wanda ainihin aikinsa shi ne toshe Ag da kuma rage layin tsayawa, za a iya buga shi da yanki da zarar p-fingers sun zama gida. Yankunan da ba su taɓa ba suna zuwa AlOx/SiOx don aro tasirin filin. Yankunan taɓawa, a ƙarƙashin manna, suna kiyaye SiOx mai tsabta don hana yaduwar B daga fashewa. Wannan musanya ɗaya ita ce mafi girman ɓangaren da za ku iya ci a bayan cikin tsalle 1.3% daga 26.34% zuwa 27.6%+.
Ma'aunin Fasaha
Na farko, ma'aunin BC 27%+ guda biyu a kan tebur
Zheng, Z. et al. Maximizing carrier extraction in hybrid back-contact silicon solar cells. Nature (2026). https://doi.org/10.1038/s41586-026- (Jami'ar Fasaha ta Beijing + Golden Stone Energy, certified 27.62%)
LONGi HIBC, 28.13%, ISFH CalTeC Apr 2026 (LONGi HIBC 2.0, hanyar HPBC, sabon rikodin da aka kafa). An gina shi akan HIBC 1.0 (Wang G. et al., Nature 647, 369–374, 2025, DOI:10.1038/s41586-025-09681-w, 27.81%) wanda aka maimaita tare da iPET + LIC; takarda yana jiran, takaddun shaida akan shafin hukuma na LONGi.
| Ma'auni | JinKo 26.66% (bifacial n-TOPCon) | BJUT + Golden Stone 27.62% (HBC) | LONGi HIBC 28.13% |
|---|---|---|---|
| Tsari | Gaban boron + baya n-TOPCon cikakken lamba biyu-Layi | Gaban a-SiOx + baya p/n interdigitated (HBC) | Gaban HJT-P + baya TOPCon-N yatsa (HIBC) |
| Voc (mV) | 744.6 | 740 | — |
| Jsc (mA/cm²) | ~41.2 est. | 42.54 | — |
| FF (%) | 85.57 | 87.73 | — |
| Babban motsi | Layi biyu SiOx/poly + toshe Ag | Gradient B-doped a-Si + O-doped a-SiOx gaban passivation | Bipolar hybrid passivation (P-gefe a-Si:H + N-gefe poly) |
Kalli wannan FF na 87.73%. 26.66% kawai ya kai 85.57%. Wannan 2% abs FF tazarar ba abin da "baya-Layi biyu" kadai zai iya cika ba. Ya fito daga baya yatsa geometry ingantawa, gaban-surface grid-free shading, da tura lamba recombination zuwa matakin 400-500 fA/cm² (ƙari a ƙasa).
Fa'idodin Fasaha
Yadda Layi biyu SiOx/poly ke canzawa da zarar an ƙaddara p-yatsa
Dukansu 26.66% da 26.34% "double-layers" sune cikakken tuntuɓar baya — poly yana rufe duk saman baya, tsakiyar SiOx yana gudana a matsayin takarda ɗaya mai ci gaba. Da zarar ainihin BC ya sa duka baya-n da baya-p su zama masu siffar yatsa, tarin yana tafiya daga "bargo" zuwa "interdigitated," kuma waɗannan dabaru biyu dole ne a sake raba su.
Ƙasan SiOx (~1.x nm, wurin rami) — baya motsawa.
Wannan shine gidan waccan takardar Das Solar pinhole-passivation. Yana girma 10¹² cm⁻² ajin ramukan da ke ɗauke da oxygen, kuma tura J₀ ƙasa zuwa 2-4 fA/cm² ya dogara da shi.
A cikin BC, ƙasan SiOx a ƙarƙashin p-finger yana fuskantar B-doped poly, ba P ba. B yana zaune tare da mafi girma Dit a Si/SiO₂ interface, kuma rabuwar B yana lalata SiOx stoichiometry. Don haka ƙasan SiOx na gefen p-finger a zahiri yana son ya zama ɗan kauri fiye da gefen n-finger (+0.2-0.3 nm), tare da 1050°C pre-anneal don kunna B da jawo crystallinity zuwa 98% (tushen da takardar 26.34% ke bayarwa).
Tsakiyar SiOx (asali 1.5 nm, wurin toshe Ag) — a cikin BC, kayan na iya zama yanki.
Wannan shine babban canji. Tsakiyar a cikin 26.66/26.34 shine tsaftataccen thermal SiO₂, aiki ɗaya (toshe Ag + rage kauri na tsayawa). A ƙarƙashin tsarin interdigitated BC, yankin p-finger yana fuskantar sabbin matsaloli biyu waɗanda cikakken tuntuɓa bai taɓa fuskanta ba.
Matsala A: Bangaren p-TOPCon yana da rauni a yanayin filin. Cajin da aka kafa daga B a mahadar Si/SiO₂ yana da inganci (akasin bangaren n+). SiOx mai tsafta ba ya ba da kariyar filin a bangaren p ba, don haka dole ka aro caji mara kyau daga waje na AlOx.
Matsala B: Sake haduwar karfe J₀,metal a karkashin p-finger shine iyakar ingancin BC. BC mara azurfa a yau yana da Al-contact J₀,metal ~2400-2500 fA/cm², wanda yayi daidai da inganci ~25.9%. Don isa matakin 26.8% na tsarin Ag, dole ka matsa J₀,metal kasa da 400-500 fA/cm².
Tsakiyar SiOx mai tsafta ba zai iya magance A ba. Sauya duka zuwa AlOx/SiOx yana taka wani nakisu.
AlOx/SiOx a matsayin layin tunnel tarko ne, amma a matsayin tsakiya yana iya zama alawa
Hanyoyi biyu suna cakuda a cikin littattafai kuma dole a raba su.
Hanyar 1: AlOx kai tsaye ya maye gurbin SiOx na kasa a matsayin layin tunnel. Aikin Kaur (Solar Energy Materials 2021) shine labarin gargadi a nan:
AlOx da AlOx/SiOx biyu-layi a matsayin layin tunnel suna karewa da kyau fiye da SiOx mai tsafta.
Tsarin: AlOx/SiOx yana kara inganta yaduwar B kuma yana hana yaduwar P. B yana taruwa a cikin "taska" a yankin AlOx sannan ya tura zuwa c-Si — bugu biyu na sake hadewar Auger tare da lahani na B-O guda biyu.
Bugu da kari, Al yana yaduwa daga AlOx zuwa c-Si yana samar da matakan zurfi, don haka SRH recombination ma yana hauhawa.
Don haka SiOx na kasa ba za a iya musanya shi da AlOx ba — wannan shine tushen dalilin da yasa 26.66/26.34 baselines ba sa motsi, kuma shine ainihin dalilin da yasa LONGi HIBC ke gudanar da "P-side tare da a-Si:H (HJT logic), N-side tare da poly (TOPCon logic)." P-side kawai ya kauce wa tsarin SiOx/poly kuma ya bi hanyar HJT amorphous + hydrogen don guje wa tarko na B-SiOx.
Yanayi na 2: AlOx/SiOx yana zaune a matsayin "tsakiya" (babu tunneling, kawai tasirin filin + toshewar Ag). Wannan shine wasan BC kawai:
SiOx na kasa (1.x nm) ya kasance tsarkakakken thermal SiO₂, yana girma pinholes masu karewa.
Sama da poly na ciki (p+, high crystallinity), kasa da poly na waje (p++, heavily doped for metallization) akwai tsarkakakken SiOx na tsakiya (1-1.5 nm).
A cikin yankunan da ba a taɓa su ba (tsakanin yatsu, da kuma cikin yatsu nesa da manna), tsakiyar yana canzawa zuwa wani siririn AlOx (2-4 nm) / SiOx (1 nm) tsari. AlOx mummunan cajin tsayayye ~10¹²-10¹³ cm⁻² yana ba p-yatsa tasirin filin kuma yana matsawa p-TOPCon-gefen Dit ƙasa.
A cikin yankunan tuntuɓar (ƙarƙashin buɗewar manna), tsakiyar yana kiyaye SiOx mai tsabta — don dakatar da AlOx daga barin fashewar B yayin harbi (Ag manna yana harbi a 700-800°C, kuma AlOx yana rashin kwanciyar hankali sama da 550°C yayin da Si-H ke rabuwa).
Babu bayanan jama'a na BC giciye-sashe akan wannan "zoned tsakiya" tukuna, amma sarkar tunani tana riƙe. a-SiOx:H/AlOx:H a 6nm/12nm akan n-type gaban-fuskar passivation ya riga ya gudanar da τeff 5.1 ms ba tare da anneal ba, wanda ke nuna AlOx ba ya taɓa c-Si kai tsaye ba (tare da a-SiOx ko SiOx a tsakanin) yana barin tasirin filin da sinadarai passivation su haɗa kai. Yankin BC p-yatsa mara tuntuɓa shine daidai wannan yanayin: ƙasan SiOx yana raba c-Si, tsakiyar AlOx/SiOx yana raba ciki poly, AlOx ba ya taɓa c-Si, kuma hanyar B watsawa an toshe ta da shinge biyu, ƙasan SiOx da ciki poly — mafi aminci fiye da AlOx a matsayin kai tsaye tunnel Layer.
Aikace-aikacen Samfur
Matakai uku masu goyan baya don p-yatsa localization (ƙarin 26.34 zuwa 27.6)
| Motsa | 26.34% tushe | BC p-finger gano wuri karuwa | Matsalar da aka magance |
|---|---|---|---|
| SiOx na kasa | Cikakken tuntuɓar 1.8 nm (p-TOPCon gefe) | 1.8→2.0 nm a ƙarƙashin p-finger (anti-B pinning), 1.3-1.5 nm a ƙarƙashin n-finger | B segregation yana lalata SiOx |
| Tsakiyar SiOx | Pure SiOx 1 nm (in-situ) | Zoned: non-contact AlOx(3nm)/SiOx(1nm), contact pure SiOx 1 nm | Rashin ƙarfi p-gefe filin tasiri |
| Waje poly | 90 nm p++, alkali-metal manna | Laser LCO buɗewa a ƙarƙashin p-finger, yana buɗe AlOx/SiNx kawai ba tare da cutar poly/SiOx ba (iVoc tabbatar da barga) | J₀,metal 2400→400 |
| Anneal | 1050°C pre-anneal, 98% crystallinity | Haka, amma p/n finger co-anneal taga dole ne a daidaita — P-gefe 880-920, B-gefe 1050, raba a ~950-980°C dogon anneal | Rikicin thermal budget |
| Metallization | Alkali-metal Ag manna (4wt% Na₂O/K₂O) | Al lamba marar azurfa ko manna Ag-Al gauraye, p-finger 700°C firing J₀,metal ~2400, n-finger ~2500 yanayi iri ɗaya | Marar azurfa + danna J₀,metal zuwa 400 |
Wannan bayanan BC marar azurfa yana da muhimmanci: bayan LCO (laser contact opening) iVoc da kyar ya faɗi kuma Raman bai nuna alamar amorphous ba, yana tabbatar da "bude taga ba tare da karya poly/SiOx passivation ba" abu ne mai yiwuwa — wannan shine tushen tsari na "kiyaye tsakiyar SiOx mai tsabta a wuraren tuntuɓar + taga manna LCO" a ƙarƙashin BC p-finger. Amma J₀,metal 2400 fA/cm² kawai yana nuna 25.9% inganci. Wannan gibin 0.9% abs zuwa tsarin Ag 26.8% gaba ɗaya ana cinye shi ta hanyar sake haɗuwa ta tuntuɓa. Matsalar tsallakewa ta gaba ba ita ce Layer passivation ba, ita ce metallization.
To menene bayan baya ke fama da shi a wannan tsallakewar ƙarshe zuwa 27%
Jera dukkan ƙarni huɗu — Das 25.4 → JinKo 26.66 → JinKo 26.34 → Golden Stone 27.62 / LONGi 28.13:
Ƙarni na 25.4%: bayan baya ya yi yaƙi a kan "halin pinhole" na SiOx na ƙasa (passivation vs sake haɗuwa), LPCVD mataki biyu oxidation yana girma 10¹²-class passivating pinholes.
Ƙarni na 26.66%: bayan baya ya yi yaƙi a kan SiOx/poly Layer biyu + tsakiyar SiOx Ag block + laser thinning, warware lalacewar metallization da parasitic absorption.
26.34% samarwa: baya p-TOPCon cikakken lamba biyu-Layi + alkali-karfe manna + 1050°C pre-anneal, cin a p-gefe B rabuwa da karfe.
27.6%+ samarwa (BC): baya p/n interdigitated, kasa SiOx daidaita ta polarity (p-gefe 2.0 / n-gefe 1.3) + yanki tsakiya (ba-lamba AlOx/SiOx don filin tasiri, lamba tsarki SiOx) + LCO bude ba tare da cutar da passivation + azurfa-free/low-Ag karfe danna J₀,karfe zuwa 400.
Ɗaya counter-intuitive kira: sama 27%, a baya "filin tasiri + biyu-Layi combo," AlOx ba zai iya shiga kasa (tunnel wuri busa B yaduwa), kawai a tsakiya ba-lamba yanki. Wannan shi ne babban bambanci daga PERC zamanin "AlOx kai tsaye a kan c-Si don aro korau caji." Tsarin BC faruwa ba ka da sarari don "yanki tsakiya" (interdigitated geometry kawo kansa zoning). Cikakken lamba TOPCon kawai ba zai iya wasa wannan. Wannan kuma ya bayyana dalilin da ya sa 26.66/26.34 ba su taba sanya AlOx a tsakiya ba — su cikakken lamba ne, tsakiya dole ya zama duka tsayawa Layer + Ag toshe, kuma tsarki SiOx ya fi aminci.
Abin da BC matukin layin zai iya gwada farko
Matsar da p-yatsa kasa SiOx daga 1.5 zuwa 1.8-2.0 nm — ƙara 30-50% LPCVD 620°C O₂ lokaci, fara duba B rabuwa ECV profile.
Gudanar da samfuran tsakiya na "zabi biyu": Rukunin A tsarkakken SiOx 1.5 nm (26.66 baseline), Rukunin B mara lamba ALD AlOx 3nm / SiOx 1nm tare da lamba tsarkakken SiOx 1.5 nm (samun daidaiton taga LCO).
Daidaita laser LCO don karɓar AlOx — AlOx ya fi kula da 355 nm fiye da SiOx, don haka girman photon 4.8 eV na deep-UV "buɗe SiNx/AlOx kawai ba tare da cutar poly/SiOx ba" yana buƙatar sake daidaitawa don allurai na gefen B da gefen N daban.
Canza metallization zuwa Ag-Al gauraye manna ko cikakken Al, kulle firing a 700°C — baseline n/p J₀,metal 2400-2500 yana buƙatar glass frit + yanayin firing + yawan tsarin lamba duk a daidaita su wuri ɗaya don kai 400-500.
Tambayoyi a buɗe ga ƙungiyar metallization na BC
Yiwuwar masana'anta na "zoned middle" a ƙarƙashin p-finger — shin ALD AlOx kawai a yankin mara lamba yana buƙatar mask, ko kuna ajiye fuska cikakke kuma ku bar LCO ya share AlOx na yankin lamba? Na farko yana ƙara matakin mask (BC ya riga ya rikita), na biyu yana haɗarin laser LCO ya share tsarin AlOx/SiOx kuma ya cutar da SiOx na ƙasa da ke kare pinholes (kawai share AlOx/SiNx aka tabbatar ba shi da lahani ga poly/SiOx; tsarin AlOx/SiOx ba a tabbatar ba).
Har ila yau, kashi 27.62% na Golden Stone ya yi amfani da "gradient B-doped a-Si + O-doped a-SiOx front passivation," tare da gefen P akan dabaru na HJT maimakon TOPCon. Shin HIBC (gefen P-HJT + gefen N-TOPCon) zai kai 28% da wuri fiye da cikakken-TOPCon BC (duka p-finger da n-finger poly/SiOx)? Layin 28.13% na LONGi yana kama da HIBC ya ci nasara, amma saukin na'urar na cikakken-TOPCon BC (poly a kan duka N da P gefe, tare da anneal) na iya dawowa da nasara akan farashi na dogon lokaci. Wadannan hanyoyi biyu na BC za su yi karo a gaba akan cinikin "passivation ceiling vs process complexity."
Takardu hudu a jere (Das 25.4 → JinKo 26.66 → JinKo 26.34 → Golden Stone 27.62 / LONGi 28.13), kuma dabaru na uku na tsararraki na TOPCon daga 25%→26%→27%+ sun cika: passivating pinholes → double-layer Ag block + thinning → BC interdigitation + zoned middle + field-effect combo.
Ra'ayin Ooitech
Tunanin zoned-middle yana da wayo, amma a gaskiya yaƙin mafi wuya yana kan layin, ba a kan zanen ba. Matsa J₀,metal daga 2400 zuwa ƙasa zuwa 400 yana nufin dosing laser LCO naka, frit ɗin paste, da kuma tsarin firing dole su riƙe haƙuri a kowane yatsa, wafer bayan wafer — kuma wannan matsala ce ta maimaitawa a layin samar da module kamar yadda take a kimiyyar cell. Ko masana'antar ta fara kan cikakken TOPCon BC ko HIBC, taga tsari yana raguwa ko ta yaya, don haka metrology da kula da zafi a kan backside stack sun zama masu tsaron ƙofa na gaske. Mutanen da ke gina ko haɓaka layukan samar da module za su iya ɗauko yawancin waɗannan trade-offs na firing da lamination a tashar YouTube ta Ooitech a www.youtube.com/ooitech.
