Page 1
Rare-earth doped fluorides for silicon solar Rare-earth doped fluorides for silicon solar cell efficiency enhancementcell efficiency enhancement
Diana Serrano Garcia
A.Braud, P.Camy, J-L.Doualan, A.Benayad, V.Menard, R.Moncorge
CIMAP, University of Caen, France
Page 2
Summary
• Limitations of solar energy conversion
• Downconversion: Quantum cutting
• Experimental results and models
• Conclusion
• Quantum cutting with Rare-earth doped fluorides
Page 3
Photovoltaics (PV)
- Conversion of solar energy into electricity
Silicon as the most famous semiconductor for solar cell development*
Si doped n
Si doped p+²
-
- Electron-hole pairs creation
*R. Singh; Journal of Nanophotonics, 3 (2009)
Page 4
I) Silicon limitations for solar spectrum conversion
Short wavelenghts
BC
BV
Si Gap =1,12eV
E
- Energy loss due to
carrier thermalization
- Absorption of photons
h>1,12 eV
Energy lost
Page 5
CB
VB
1,12eV
E
- Silicon transparent for
h<1,12 eV .
Low efficiency for Silicon solar cells (a-Si 9%,c-Si 25%*)
Silicon limitations for solar spectrum conversion
Long wavelenghts
*M.Green, Progress in Photovoltaics : Research and Applications 17, 320 (2009)
Page 6
Stacking of semiconductors with
different bandgaps
SC 2 SC 1SC 3
Decreasing bandgap
Sem
icond. 1
Sem
icond. 2
Sem
icond. 3
Solution I: multi junction solar cells
The larger bandgap at the
surface of the device
BUT:
expensive and difficult to produce
Aerospace Applications
High efficiency
(World record 40%*)
*M.Green, Progress in Photovoltaics : Research and Applications 17, 320 (2009)
Page 7
Si 2 e-
Quantum cutting
Solution II: Frequency Conversion
hν
hν
hν
Efficiency enhancement
Quantum CuttingLow cost production
*T. Trupke, M. Green; Journal of applied physics 92, 3 (2002) 1668
Ideal converter 36,6%*
c-Si Solar Cell
Si Si
Quantum cutting layer
Page 8
Downconversion: Quantum cutting by energy transfer
Donor Acceptor
E
D0 A0
D1
D2
A1
Acceptor
A0
A1
D2D1 and A0A1D1D0 and A0A1
Donor excitationD0D2
Acceptor relaxationA1A0
2 photons emission
Energy transfer 1 Energy transfer 2
hv/2 hv/2
hv
Getting 2 photons from 1 photon?
1
2
From 1 photon we get 2 photons
Page 9
Quantum cutting with rare-earth doped fluorides:
►Why Yb?
E~10000 cm-1
~1,2eV ~ Si Gap
Yb3+
53 H
43 F
43 H
Pr3+
23 F
63 H
33 F
41G
03 P
21 D
1
272 F
252 F
23 P
61 I
Yb3+
272 F
252 F
Donor Acceptor
Pr3+/Yb3+ system
E(3P0 – 1G4)~E(2F5/2 – 2F7/2)
E(1G4 – 3H4)~E(2F5/2 – 2F7/2)Acceptor
E
►Why Pr ?
Resonant Energy Transfer
1
2
Page 10
Host matrix: Fluorides
- Low phonon energy
•High fluorescence quantum yield
- Large transparency range
Differences: RE3+ doping
Short distance between ions
Very efficient energy transfer
KY3F10 Uniform distribution of dopants
CaF2 Formation of complexes (clusters)
Page 11
0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0-0,25
0,00
0,25
0,50
0,75
1,00
1,25
1,50
1,75
2,00
I Pr/I
Yb
%Yb
IPr/IYb
CaF2:Pr/Yb First energy transfer Pr Yb
Intensity ratio as a function of Yb concentration Yb
03
Pr
I
PI
Pr excitation
Increase of first transfer Pr (3P0)Yb (2F5/2) with Yb concentration
03
Pr PI
YbI53 H
43 F
43 H
Pr3+
23 F
63 H
33 F
41G
03 P
21 D
1
272 F
252 F
23 P
61 I
DonorYb3+
Acceptor
Transfer
Ytterbium emission under Pr (3P0) excitation
• Energy transfer from Pr to Yb
(%Pr constant at 0.5%)
Experimental Results
Page 12
0 80 160
0,1
1
3 µs
17 µs
45 µs
82 µs
CaF2: 0,5% Pr_4% Yb
CaF2: 0,5% Pr_2% Yb
CaF2: 0,5% Pr_1% Yb
CaF2: 0,5% Pr_0,5% Yb
Inte
nsi
té (
U.A
.)
t(µs)
3P0 fluorescence decay in CaF2 and KY3F10
Decrease of (3P0 ) with Yb concentration for both hosts
Experimental Results
PrPr
11
11)(
Yb
sW
Pr
Pr1
Pr1
1 1
Yb
W
W
Energy transfer rate :
Energy transfer efficiency :
40 80 120
0,1
1
1 µs
22 µs
3 µs
Inte
nsi
té (
U.A
.)
t (µs)
KY3F
10: 0,5% Pr 20%Yb
KY3F
10: 0,5% Pr 10%Yb
KY3F
10: 0,5% Pr 1%Yb
Page 13
0 5 10 15 200
20
40
60
80
100
R
ende
men
t (%
)
% Yb
KY3F
10: Pr,Yb
CaF2: Pr,Yb
Experimental Results
97% eficiency with 4% Yb in CaF2
96% eficiency with 20% Yb in KY3F10
Transfer in CaF2 more efficient than transfer in KY3F10
Eff
icie
ncy
(%)
Page 14
Modeling I: Classical model
3 energy transfer
0
34
15/2
2
5/22
0
5/22
PnGnβFnαFτ
1
dt
Fdn
0
3
03
0
320
35/2
24
1
41
0
41
PnPτ
BRPnFnαGnβ
Gτ
1
dt
Gdn
Pr0
3
03
0
03
NσPnPτ
1
dt
Pdn
- Possible interaction
between all Pr and Yb ions
- Uniform ion distribution53 H
43 F
43 H
23 F
63 H
33 F
41G
03 P
21 D
1
272 F
252 F
61 I
272 F
252 F
Page 15
41
0
4
03
0
63132
5/22
0
22
2
Gτ
P
Pτ
PBRPβPα
Fτ
PPwp
dt
dP
03
0
531
41
0
3
5/22
0
21
1
Pτ
PBR
Gτ
P
Fτ
PPwp
dt
dP
6 possible states
3H4
2F5/2
3P0
1G4
2F7/2
Pr3+ Yb3+
5/22
0
4
03
0
532323
41
0
33
Fτ
P
Pτ
PBRPβPαPβ
Gτ
P
dt
dP
03
0
6325
5/22
0
4
41
0
44
Pτ
PBRPγ
Fτ
P
Gτ
P
dt
dP
5/22
0
6
03
0
532
03
0
53151
5
Fτ
P
Pτ
PBR
Pτ
PBRPγPwp
dt
dP
5/22
0
6
03
0
632
03
0
6312
6
Fτ
P
Pτ
PBR
Pτ
PBRPwp
dt
dP
Limited interaction within the pair
P1+P2+P3+P4+P5+P6 =1
Each state has a
probability Pi
Modeling II: Pair model
Page 16
Conclusion
• Possible QC with CaF2:Pr/Yb and KY3F10: Pr/Yb
• Transfer in CaF2 more efficient!
Implementation in Si solar cells
• Very efficient Pr Yb first transfer (97%)
Conclusions
New models (Three or more ions model??)
Page 17
THANK YOU
for your attention!!