Hydrogen passivation of polycrystalline Si thin film solar cells [Elektronische Ressource] / vorgelegt von Benjamin Gorka

Hydrogen Passivationof Polycrystalline Si Thin FilmSolar Cellsvorgelegt vonDiplom-PhysikerBenjamin Gorkaaus BerlinVon der Fakultät IV - Elektrotechnik und Informatikder Technischen Universität Berlinzur Erlangung des akademischen GradesDoktor der NaturwissenschaftenDr. rer. nat.genehmigte DissertationPromotionsausschuss:Vorsitzender: Herr Prof. Dr. Christian BoitBerichter: Herr Prof. Dr. Bernd RechBerichter: Herr Prof. Dr. Uwe Rau (Forschungszentrum Jülich)Tag der wissenschaftlichen Aussprache: 17.12.2010Berlin, 2010D 832ContentsZusammenfassung 6Abstract 71 Introduction 92 Fundamentals 132.1 Hydrogen Diffusion and Passivation in Si . . . . . . . . . . . . . 132.1.1 Hydrogen Diffusion in Monocrystalline Si . . . . . . . . . 142.1.2 Trap-limited Hydrogen Diffusion in Silicon . . . . . . . . 162.1.3 Passivation of Defects . . . . . . . . . . . . . . . . . . . 182.2 Sources for Hydrogenation of Si . . . . . . . . . . . . . . . . . . 192.2.1 Molecular H Source . . . . . . . . . . . . . . . . . . . . 1922.2.2 Plasma Sources . . . . . . . . . . . . . . . . . . . . . . . 202.2.3 Solid Source (Firing) . . . . . . . . . . . . . . . . . . . . 212.3 Hydrogen Passivation of Poly-Si Thin Film Solar Cells: State-of-the-Art . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Experimental Procedures 253.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . 253.1.1 PolycrystallineSiSolarCellsformedbySPCofPE-CVDgrown a-Si:H . .
Publié le : vendredi 1 janvier 2010
Lecture(s) : 64
Tags :
Source : D-NB.INFO/101182471X/34
Nombre de pages : 141
Voir plus Voir moins

Hydrogen Passivation
of Polycrystalline Si Thin Film
Solar Cells
vorgelegt von
Diplom-Physiker
Benjamin Gorka
aus Berlin
Von der Fakultät IV - Elektrotechnik und Informatik
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
Doktor der Naturwissenschaften
Dr. rer. nat.
genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Herr Prof. Dr. Christian Boit
Berichter: Herr Prof. Dr. Bernd Rech
Berichter: Herr Prof. Dr. Uwe Rau (Forschungszentrum Jülich)
Tag der wissenschaftlichen Aussprache: 17.12.2010
Berlin, 2010
D 832Contents
Zusammenfassung 6
Abstract 7
1 Introduction 9
2 Fundamentals 13
2.1 Hydrogen Diffusion and Passivation in Si . . . . . . . . . . . . . 13
2.1.1 Hydrogen Diffusion in Monocrystalline Si . . . . . . . . . 14
2.1.2 Trap-limited Hydrogen Diffusion in Silicon . . . . . . . . 16
2.1.3 Passivation of Defects . . . . . . . . . . . . . . . . . . . 18
2.2 Sources for Hydrogenation of Si . . . . . . . . . . . . . . . . . . 19
2.2.1 Molecular H Source . . . . . . . . . . . . . . . . . . . . 19
2
2.2.2 Plasma Sources . . . . . . . . . . . . . . . . . . . . . . . 20
2.2.3 Solid Source (Firing) . . . . . . . . . . . . . . . . . . . . 21
2.3 Hydrogen Passivation of Poly-Si Thin Film Solar Cells: State-
of-the-Art . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3 Experimental Procedures 25
3.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1.1 PolycrystallineSiSolarCellsformedbySPCofPE-CVD
grown a-Si:H . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1.2 Polycrystalline Si Solar Cells formed by SPC of E-Beam
Evaporated a-Si . . . . . . . . . . . . . . . . . . . . . . . 27
3.1.3 Rapid Thermal Annealing . . . . . . . . . . . . . . . . . 28
3.1.4 Hydrogenation . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 Plasma Characterization . . . . . . . . . . . . . . . . . . . . . . 34
3.3 Sample . . . . . . . . . . . . . . . . . . . . . . 34
3.3.1 Raman - Phonon Scattering . . . . . . . . . . . . . . . . 34
3.3.2 Electron Spin Resonance . . . . . . . . . . . . . . . . . . 35
3.3.3 Open Circuit Voltage and Suns-Voc . . . . . . . . . . . . 35
3Contents
4 Plasma Process Optimization 39
4.1 Influence of H Plasma Treatment on Device Performance . . . . 39
4.2 of Plasma Conditions . . . . . . . . . . . . . . . . . . 42
4.3 Plasma Simulation . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.1 Plasma Model . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.2 Simulation Results . . . . . . . . . . . . . . . . . . . . . 49
4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5 Dynamic of the Hydrogen Passivation 59
5.1 Time and Temperature Dependence of Hydrogen Passivation . . 59
5.2 Optimum Temperature . . . . . . . . . . . . . . . . . . . . . . . 63
5.3 H Out-Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6 Interaction of Passivation with Material Properties 77
6.1 Influence of Defect Annealing on Hydrogen Passivation . . . . . 77
6.1.1 Influence of Defect Annealing on the Dynamics of HP . . 78
6.1.2 Variation of Defect . . . . . . . . . . . . . . . 82
6.2 Passivation of Poly-Si Films with Different Structural Properties 83
6.2.1 Structural Properties After Electron Beam Evaporation
and Crystallization . . . . . . . . . . . . . . . . . . . . . 84
6.2.2 Hydrogen Passivation of Defects (Si dangling bonds) . . 89
6.2.3PassivationofSolarCellswithDifferentStruc-
tural Properties . . . . . . . . . . . . . . . . . . . . . . . 91
6.2.4 InterplayofDefectAnnealingwithPassivationandStruc-
tural Properties . . . . . . . . . . . . . . . . . . . . . . . 93
6.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
7 Discussion of Electronic Solar Cell Performance 103
8 Discussion 111
9 Conclusion 119
Acknowledgement 121
List of Publications 123 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Conferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4Contents
Nomenclature 127
Bibliography 131
5Zusammenfassung
Die Wasserstoffpassivierung ist ein zentraler Prozess in der Herstellung von
polykristallinen Si (poly-Si) Dünnschichtsolarzellen. Im Rahmen der Arbeit
wurde ein RF-Parallelplattenreaktor für die Wasserstoffbehandlung eingesetzt.
Schwerpunkte der Untersuchungen waren (i) die Rolle von Plasmaparametern
(wie Druck, Elektrodenabstand und Leistung), (ii) die Dynamik der Wasser-
stoffbehandlung und (iii) das Zusammenspiel aus Passivierung und Materialei-
genschaften. Die Charakterisierung erfolgte anhand von Messungen der Leer-
laufspannung V an poly-Si Referenzzellen.OC
Durch Messungen der Plasma-Durchschlagspannung V wurden optimalebrk
Bedingungen für die Passivierung gefunden. Beste Ergebnisse bei Elek-
trodenabständen erzielt, bei denenV für den jeweiligen Druck ein Minimumbrk
hatte. Es wurden Plasmasimulationen durchgeführt, die nahe legen, dass dies
einerMinimierungderIonenenergieentspricht.EineErhöhungdesWasserstoff-
gehaltes im Plasma führte dagegen zu keiner Verbesserung der Passivierung.
Untersuchungen zur Dynamik zeigten, dass eine Wasserstoffbehandlung bei
geringen Temperaturen (≤ 400°C) mehrere Stunden dauert. Dagegen kann
diese bei erhöhten Temperaturen von 500°C bis 600°C in nur 10 min (Pla-
teauzeit) erfolgreich durchgeführt werden. Anhaltende Behandlungen führten
zu einer Verschlechterung von V , vor allem ober- und unterhalb des be-OC
obachteten Optimums (<500°C und >600°C). Als alternatives Verfahren zur
Herstellung von poly-Si Schichten wurde die Elektronenstrahlverdampfung un-
tersucht. Unterschiedliche Materialeigenschaften wurden durch Variation der
DepositionstemperaturT = 200−700°C eingestellt und mithilfe von Ramandep
und ESR untersucht. Große Körner wurden nach Festphasenkristallisation von
amorphem Si, das bei 300°C abgeschieden wurde, erreicht. Die Anzahl offener
Si-Bindungen konnte mittels Passivierung um etwa eine Größenordnung re-
16 −3duziert werden. Die niedrigste Konzentration von 2.5· 10 cm wurde für
poly-Si mit den größten Körnern gefunden und steht im Einklang zu den be-
sten Solarzellenergebnissen (nach RTA und Passivierung).
Die Wasserstoffpassivierung von poly-Si Filmen konnte bei hohen Tempera-
turen von 500°C bis 600°C mittels Plasmabehandlung erfolgreich durchgeführt
werden. Jedoch scheinen im Laufe der Plasmabehandlung auch neue Defekte
zu entstehen, die mit andauernder Passivierung zu einer Verschlechterung von
V führen. Dieser Effekt sollte minimiert werden. Um hohe Leerlaufspannun-OC
gen oberhalb von 450 mV zu erreichen, wird zunehmend auch eine niedrige
Rekombination an Grenzflächen wichtig.
6Abstract
Hydrogen passivation is a key process step in the fabrication of polycrystalline
Si (poly-Si) thin film solar cells. In this work a parallel plate rf plasma setup
was used for the hydrogen passivation treatment. The main topics that have
been investigated are (i) the role of plasma parameters (like hydrogen pressure,
electrode gap and plasma power), (ii) the dynamics of the h treatment
and (iii) passivation of poly-Si with different material properties. Passiva-
tion was characterized by measuring the open-circuit voltage V of poly-SiOC
reference samples.
Optimum passivation conditions were found by measurements of the break-
down voltage V of the plasma for different pressures p and electrode gapsbrk
d. For each pressure, the best passivation was achieved at a gap d that corre-
sponded to the minimum in V . Plasma simulations were carried out, whichbrk
indicate that best V corresponds to a minimum in ion energy. V was notOC OC
improved by a larger H flux. Investigations of the passivation dynamic showed
that a plasma treatment in the lower temperature range (≤ 400°C) is slow and
takes several hours for the V to saturate. Fast passivation can be success-OC
fully achieved at elevated temperatures around 500°C to 600°C with a plateau
time of 10 min. It was found that prolonged hydrogenation leads to a loss in
V , whichislesspronouncedwithintheobservedoptimumtemperaturerangeOC
(500°C− 600°C). Electron beam evaporation has been investigated as an al-
ternative method to fabricate poly-Si absorbers. The material properties have
been tuned by alteration of substrate temperature T = 200− 700°C anddep
were characterized by Raman, ESR and V measurements. Largest grainsOC
were obtained after solid phase crystallization (SPC) of a-Si, deposited in the
temperaturerangeof300°C.ThedefectconcentrationofSidanglingbondswas
lowered by passivation by about one order of magnitude. The lowest dangling
16 −3bond concentration of 2.5· 10 cm after passivation was found for poly-Si
with largest grains and coincides with best solar cell results, obtained after
rapid thermal annealing and hydrogen passivation.
Hydrogen passivation of poly-Si films was successfully achieved with a par-
allel plate rf H plasma treatment at elevated temperatures around 500°C to
600°C. Yet it seems thatt induced defect generation causes a loss in
V with prolonged passivation time and should be minimized. In order toOC
achieve high open circuit voltages larger than 450 mV, in addition to hydro-
gen passivation, low recombination at the interfaces becomes more and more
important.
781 Introduction
Solar cells are photovoltaic (PV) devices that can convert sunlight directly into
electricity, providing a clean and decentralized renewable energy source. The
PV market in the year 2009 was still dominated by Si wafer based solar cell
technology, which amounted to 81% of the overall market [1]. However, the
trend in PV goes towards thin film technologies, in order to reduce production
costs and material consumption. Polycrystalline Si (poly-Si) thin films on for-
eign low cost substrates such as glass offer the potential of high cell efficiencies
[2, 3, 4] combined with the advantages of thin film technology and vast Si
abundance.
First investigations about low cost polycrystalline Si (poly-Si) thin film solar
cells on foreign substrates date well back into the mid 70’s [2, 5]. In the recent
years noticeable progress has been made. In 2006, first solar modules based on
poly-Si on glass substrate have entered into commercial mass production [6, 7].
One of the key process steps relies on defect passivation of the poly-Si layers
by a hydrogen treatment, which can potentially multiplicate the efficiency of
a poly-Si solar cell [8, 9].
Crystalline Si can be categorized by grain size. Here the following definition
will be used:
• monocrystalline Si (c-Si) for single-crystal,
• multicrystalline Si (mc-Si) with an average grain size of > 100 μm
• polycrystalline Si (poly-Si) with an average grain size between 100 μm
and 0.1 μm
• microcrystalline (μc-Si) with crystallites smaller than 0.1 μm, usually
embedded in an amorphous network.
A poly-Si thin film solar cell consists of fully crystalline poly-Si absorber layer.
Thecompletestackwithatypicalthicknessbetweenonetoseveralmicrometers
cannot support itself and needs to be grown onto a foreign substrate. The
average grain size is between 0.1μm and 100μm and hence much smaller than
91 Introduction
thegrainsizeofmulticrystallineSi. Grainboundariesaresourceofdefectsthat
deteriorate the efficiency of the solar cell. One approach to improve the device
quality focuses on increasing the average grain size and thereby reducing grain
boundary induced defects. Large grained poly-Si (~10 μm) can be achieved
through a method based on aluminum induced crystallization (AIC) of a thin
amorphous Si layer [10]. Even larger grained (~100 μm) layers were obtained
by laser crystallization of amorphous Si with a scanning cw laser beam [11].
Both methods produce a poly-Si thin film that in a next step can be used
as a seed layer for subsequent epitaxial growth of the absorber [12, 13, 14].
Nevertheless, the most successful method to produce poly-Si absorber layers is
based on solid phase crystallization (SPC). Amorphous Si layers are annealed
in a furnace at typically 600°C for several hours until a completely crystallized
poly-Si film is formed (grain size: 1-3 μm). Matsuyama et al. (Sanyo) could
demonstrate a cell efficiency of 9.2% [15]. CSG Solar recently achieved a mini
module efficiency of 10.4% [4] and is to date the first and only company that
went into industrial production of poly-Si thin film solar cells [7].
Poly-Si films exhibit numerous extended and point defects mainly in the
form of Si dangling bonds. These dangling bonds form electrically active de-
fect states within the bandgap, which act as recombination centers for charge
carriers and thus deteriorate the device quality. Hydrogen has the ability to
diffuse in Si and in particular to bond to dangling bonds, resulting in a pas-
sivation of the defects [16]. However, due to the high process temperatures
almost no hydrogen remains in the poly-Si layers. Therefore hydrogen needs
to be introduced by post-deposition exposure to atomic hydrogen.
First studies about hydrogen in Si date back to the 1950s, when Van Wierin-
gen and Warmoltz determined diffusion coefficient and activation energy for
hydrogen diffusion in single crystal silicon [17]. Great interest in hydrogen
arose with the discovery in 1976 that incorporation of H strongly improves
the properties of amorphous Si [18]. Amorphous Si lacks any long range order
and exhibits a high degree of unpaired Si bonds. Hydrogen can passivate the
unpaired Si dangling bonds which otherwise form defect states in the band gap
of a-Si and act as recombination centers for the charge carriers [19].
Asimilarbeneficialeffecthasalsobeenfoundforgrain-boundarypassivation
by H in poly-Si [20, 21, 16]. First studies, performed in 1979 by Seager et al.
on hydrogen plasma treatments of grain boundaries in mc-Si, demonstrated
a significant reduction of potential barrier and density of states due to the
plasma treatment [21]. In the following year the authors could report on
10

Soyez le premier à déposer un commentaire !

17/1000 caractères maximum.