Review of Photovoltaic Energy Production Using CdTe …sites.ieee.org/scv-photovoltaic/files/slides/2009-March...Tim Gessert General PV Energy Talk, Date Overview US Energy 101 Thin-Film

Timothy A. Gessert

Principal Scientist, Group Manager - Polycrystalline Thin Film PV Devices National Center for Photovoltaics, NREL

[email protected], 303-384-6451

Research supported by United States Department of Energy Contract No. DE-AC36-99GO10337

Review of Photovoltaic Energy Production Using CdTe Thin-Film Modules

Tim Gessert General PV Energy Talk, Date

Overview

US Energy 101 Thin-Film PV Market and Industry

Operation of Thin-Film CdTe PV Devices


2004 US Total Energy 3.3 TW

(99.6 Quadrillion BTU)

3.3 TW Power

Nuclear

8%

Petroleum

41%

Coal

22%

Natural Gas

23%

Renewabl

e

US Energy 101 2004 US Renewable Energy 0.2 TW

(6.1 Quaddrillion BTU)

Alcohol5%

Wood33%

Wind2%

Waste (LFG)9%

Solar1%

Geothermal6%

Hydro44% Hydro

AlcoholSolarGeothermalWaste (LFG)WindWood

2004 US Solar Energy 0.002 TW (2 GW)(0.063 Quadrillion BTU)

1998 - 91% - Hot Water 6%-Thermal 3%-PV

2004 US PV Module Production ~150 MW/Year

2004 Cumulative World Production3,844 MW

Silicon83%

Thin Films17%

SiliconThin Films

2004 US Total Energy Consumption by Fuel Type 99.6 Quadrillion BTU Energy (~100 Quads)

(~100x1015 BTU, 28x1012 KWH) 3.3 TW (Constant) Power

(2025 Projected - 133.2 Quads)

2004 US Renewable Energy 6.1 Quadrillion BTU (0.2 TW Power)

2004 US Solar Energy 0.063 Quadrillion BTU - 0.002 TW Power

(1998 91% Hot Water 6% Thermal 3% PV)

2004 US PV Module Production ~150 MW/Year

2004 Cumulative World Production ~4000 MW (4 GW, 0.004 TW)

Source: 2004 Annual Energy Review Energy Information Administration

www.eia.doe.gov

Thin-Film PV produced ~0.0003% of 2004 US Energy Consumption

Renewable 6%


US Energy 101

Source: Annual Energy Outlook 2005 - With Projections to 2025 Energy Information Administration

www.eia.doe.gov

Historical and Projected US Energy Consumption by Major Fuel Type - to 2025

Total Energy Consumption projected to increase by 1.4%/year Petroleum consumption projected at 1.5% year

Electricity consumption projected to increase by 1.8%/year

Qua

drill

ion

BTU


US Energy 101

Source: Energy Information Administration www.eia.doe.gov

US Energy Consumption by Sector (2004) Electricity Sector is largest and coal generates most of electricity

TransportElectricityResidentialIndustrial

Industrial 22% Transport

27%

Electricity 39%

Residential 11%

NuclearOtherHydroPetroleumOther GasesNatural GasOther RenewCoal

Coal 50%

Nuclear 20%

Natural Gas 18%

Hydro. 7%

Petroleum 3%

Total US Energy Consumption (100 Quads)

Electric Power Generation By Fuel Type

Other Renewables 2%

The Dilemma: The US needs to plan to double its electricity consumption in ~35 years (~1.8% growth)

- while acknowledging issues associated with burning more coal, natural gas, and petroleum - and acknowledging issues associated with building more dams and nuclear power plants


Solar Land Area Requirements For 2004 US Electricity Requirements

Sources: K. Zweibel, “PV Past the Tipping Point” (www.nrel.gov/ncpv/thin_film) PVWATTS Calculator, Version 1 (http://www.nrel.gov/rredc/pvwatts), 0° Tilt

Photovoltaics

Phoenix, AZ 2100

KWH/m2-yr

Boulder, CO 1900 KWH/m2-yr

Madison, WI 1600 KWH/m2-yr

Caribou, ME 1550 KWH/m2-yr

Nebraska ~1800 KWH/m2

Note: Munich, Germany ~1100 KWH/m2


US Energy 101

World Growth in PV Module Production Source: PV News, Paul Maycock, Vol. 27, No. 3, April 2008

(and previous volumes)

0

50

100

150

200

250

300

MWp/yr

2002 2004 2006 2008

US a-SiUS CdTeUS CIGSUSWorld

% G

row

th in

PV

Mod

ule

Prod

uctio

n

Could we deploy 3.3 TW of PV in US - In my lifetime? -

Assumptions: 35% Annual Exponential Growth in Production 2004 US Installed PV ~100 MW (=0.0001 TW)

Installed PV = ~ Twice Present Production Capacity

US PV Production Deployed (=Twice Capacity) 2004 0.00005 TW 0.0001 TW 2006 0.0001 TW 0.0002 2008 0.0002 TW 0.0004 2010 0.0004 TW 0.0008 2012 0.0008 TW 0.0016 2014 0.0016 TW 0.0032 2016 0.0032 TW 0.0064 2018 0.0064 TW 2020 0.0128 TW 2022 0.0256 TW ~20 GW/Year DOE 2050 TF Goal 2024  0.0512 TW 2026 0.1024 TW 2028 0.2028 TW ~0.47 TWp = Total 2004 US Nuclear 2030 0.4096 TW 2032 0.8192 TW 2034  1.6384 TW 2036 2.2768 TW 2038  6.5536 TW ~13 TWp = ~4 TW Constant Power

(I’ll be only 78 years old!)

0

10

20

30

40

50

60

1 2 3 4 5 6 7 8

Series1

35%

2000 2001 2002 2003 2004 2005 2006 2007


Applications for Web-Based PV

How large is the PV Roof Shingle Market? US Residential Roof Area = 163 x109ft2 = ~16x109 m2 US Commercial Roof Area = 50 x109 ft2 = ~5x109 m2 Shading Factor Residential = 78% Commercial = 50%

Roof Area Appropriate for PV US Residential = 3.6x109 m2 Commercial = 2.5x109 m2

How much energy would this produce? Assume 1800 KWH/m2/year (US Average) 10% efficient PV modules

Residential + Commercial = 1.1x1012 KWH/year Or 10%of U.S. Electricity Consumption in 2004

Amount of peak module power ~6x1011 Watts - 240 times the 2006 World Production of PV!!

Source for Roof Areas and Shading Factors: Ron Judkoff, Director Buildings and Thermal Systems, NREL

Rigid Glass

Flexible

Note:

US BIPV market alone is 240X more pies - not just 240X the 2% piece!


US Produced vs. End-Used Energy (Exajoules) 2/3-3/4 of energy used to produce electricity is ultimately “Rejected”

BIPV application embodies very low rejection factor, so 10% is much higher


Background

Example - Size of PV Market for Commercial Glass

Source: Brent Nelson, NREL

Area of PV Installed

1.00E+05

1.00E+06

1.00E+07

1.00E+08

1.00E+09

1.00E+10

1.00E+11

1.00E+12

1.00E+13

1980 2000 2020 2040 2060 2080 2100

Year

Area (

sq

. m

ete

rs)

20 TW-yr 10% efficient

Actual

Area to Install

USA Roofing Production

Roofs in USA

World Glass Production

World Paved Roads

PV Use of Glass Could Exceed 2005

World Glass Production By ~2020!


The Best One-of-a-Kind Laboratory Cell Efficiencies for Thin Films

(Standard Conditions)

Effic

ienc

y (%

)

CuInGaSe2

CdTe Amorphous silicon (stabilized)

Univ. of Maine Boeing

Boeing

Boeing

Boeing

ARCO

AMETEK

NREL

Boeing EuroCIS

Univ. of So. Florida

Univ. of So. FL

BP Solar

Kodak

Kodak Monosolar

Matsushita

12

8

4

0 2000 1995 1990 1985 1980 1975

United Solar

16

20

NREL

2005

RCA

ECD

NREL

Photon Energy


Some CIGS-alloy, CdTe, & a-Si Laboratory Cell NREL-Confirmed Record Efficiencies

Area (cm2)

CIGSe

CIGSe

CIGS

CIAS

CdTe

CdTe

a-Si

Area (cm2)

0.410

0.402

0.409

–

1.03

–

-

VOC (V)

0.697

0.67

0.83

0.621

0.845

0.814

-

JSC (mA/cm2)

35.1

35.1

20.9

36.0

25.9

23.56

-

FF (%)

79.52

78.78

69.13

75.50

75.51

73.25

73.25

Efficiency (%)

20.0

18.5

12.0

16.9

16.5

14.0

12.1

Comments CIGSe/CdS/Cell

NREL, 3-stage process CIGSe/ZnS (O,OH) NREL, Nakada et al

Cu(In,Ga)S2/CdS Dhere, FSEC

Cu(In,Al)Se2/CdS IEC, Eg = 1.15eV

CTO/ZTO/CdS/CdTe NREL, CSS

ZnO/CdS/CdTe/Metal U. of Toledo, sputtered

United Solar, Stabilized Efficiency

Sources: Updated from R. Noufi and K. Zweibel, Proc. 4th WCPEC, Waikola, Hawaii, 5/2006, Photon International, October 2004 Updated


Compan y Device Aperture Area (cm 2 )

Effi c ienc y (%)

Power ( W) Date

G l oba l So l ar CIGS 8390 10.2* 88.9* 05/05

She ll So l ar CIGSS 7376 11.7* 86.1* 10/05

W ürth So l ar CIGS 6500 13.0 84.6 06/04

F i rst So l ar CdTe (6623) 12.6* 84.36 * 08/08

She ll So l ar G m bH CIGSS 4938 13.1 64.8 05/03

Antec So l ar CdTe 6633 7.3 52.3 06/04

She ll So l ar CIGSS 3626 12.8* 46.5* 03/03

United Solar a-Si 4519 7.9 35.7 06/97

Polycrystalline Thin Film PV Modules (Standard conditions, Aperture-area, *NREL Confirmed)

Ranked by Power

Sources: R. Noufi and K. Zweibel, Proc. 4th WCPEC, Waikola, Hawaii, 5/2006, Photon International October 2004, Updated.


Demonstrated Efficiency

Perceived Production Advantage

Perceived Materials

Abundance/ Low Toxicity

a-Si Strength

CdTe Strength

CIS Strength

Present Strengths of Each Thin-film PV Technology

Thoughts: Each technology has different advantages

Its not clear which (if any) advantage will yield a long-term product advantage This situation could remain true for many years!


CdTe Thin-Film Solar Cells

Process Direction


TCO and “Buffer” Layer

CdTe bandgap

TCO Layer Alternatives: SnO2:F (First Solar, AVA, others, US) In2O3:Sn/SnO (Antec Solar, Germany) Cd2SnO4 (Primestar Solar, US)

“Buffer” Layer Alternatives Undoped SnO2 Zn2SnO4

Data from X. Wu et. al, Proc. 28th Photovoltaics Specialists Conf. pp. 470-474 (2000))

Buffer

Loss of ~1.5 mA/cm2

for Commercial TCO

Model Parameters: Mobility

~30 cm2/V-sec Carrier Concentration

~7x1020 cm-3

Effect of TCO on PV Module Performance

CdTe PV Module (~850 nm Bandgap) Loss of ~1.5 mA/cm2

for Commercial Glass 100

80

60

40

20

0

Unsc

aled

QE,

T+R

, A (

%)

900800700600500400300Wavelength (nm)

~3 mA/cm2

TCO Absorption

T+R, Soda-Lime Glass -1000 ppm Fe2O3 T+R, Soda-Lime Glass - 100 ppm Fe2O3 T+R, 7059 Technical Glass Absorption, Commercial TCO

Drude Model 7e20 cm-3, 30 cm2/V-sec QE, Commercial CdTe Device QE, NREL CdTe Device

~4 mA/cm2

0.4

0.3

0.2

0.1

0.0

Abs

orpt

ance

(%)

2400 1900 1400 900 400 Wavelength (nm)

n (1/cm 3 ) 1 x 10 21

5 x 10 20

1 x 10 20

5 x 10 19

Visible

µ = 100 cm2V-1s-1

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Abs

orpt

ance

(%)

2400 1900 1400 900 400

Wavelength (nm)

n = 5 x 1020 cm-3

µ (cm 2 /V-s) 50 100 500 1000

€

µ =qτm *

Effect of TCO on PV Module Performance

General Conclusion – High TCO Mobility is a Good Thing


Thickness: ~300 nm (Industry) , ~50-100 nm (Research)

Solution Growth (NREL, BP Solar) CdSO4, NH4OH, N2H4CS (Thiorea), H2O

Close-Space Sublimation (Antec, SSI, AVA) Gas-Phase Transport (First Solar) Sputtering CdS:O (NREL) CdZnS CdS:In

Future Trend – (Near) Elimination of CdS (Photon Energy/Golden Photon)

CdS Layer

With ZTO Buffer Layer

Buffer

CdTe Layer (Congruent Sublimation)


Gas-Phase-Transport (GPT, VTD )

(Univ. Delaware [IEC] Design Shown)

4”x6” Pre-Heater, 600°C 2”x6” Post-Heater 600°C

4”x4” Substrate

Heater/Enclosure (Hot-Press boron nitride (BN) with borate binder)

Ta Wire Confined in Boron Nitride

Non-Heated Region Constrains Deposit

CdTe

Halogen Lamp Heaters

CdTe Source

Substrate

Halogen Lamp

Halogen Lamp

Close-Space Sublimation (CSS) (Present NREL Design)

CdTe Layer (Research System Designs)

Source: D. Rose et. al., Prog. In Photovoltaics; Res. and Appl. 7, 3331-340 (1999)


CdS

ZnTe :Cu

2µm

CdTe CdS

ZnTe:Cu

CdTe Back Contact

All-Dry Contact Process Glass/SnO2:F/CdS/CdTe/ZnTe:Cu/Ti Device (Ti removed)

Historic Back Contact Te-enriched interface Produced by chemical Etching processes

Sources: Gessert et. al., 31th IEEE PVSC, pp. 291-294. Romero et. al., MRS Proc. 719 F8.40.1 (2002)


How do Thin-Film CdTe PV Junctions Form??

Motivation

40

20

0

-20

Curr

ent

Dens

ity (

mA

/cm

2 )

1.00.80.60.40.20.0-0.2Voltage (V)

612A-1, 0Å 614A-2, 500Å 615A-4, 1000Å' 616A-1L, 2000Å 613A-2, 3000Å 611A-2, 4200Å 617A-3, 5000Å


Lower Temp.

Downstream Processes

The ZnTe:Cu/Ti Contact Process (All-Dry, High-Temperature [~300°C])

ZnTe:Cu/Ti Contact Process

Higher Temp.

Upstream Processes

Photolithography

CdTe

ZnTe:CuTi

Glass

Some Research Advantages Precise control of junction performance

High device stability Can achieve very low Cu incorporation

Easy to make large, identical sample sets


Some Insight from Past Studies

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~320°C (32V, Optimum) ~340°C (34V) ~360°C (36V)

Cu Diffusion Less Than Optimum Cu Diffusion Greater Than Optimum

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~240°C (24V) ~280°C (28V) ~320°C ( 32V, Optimum)

40

30

20

10

0

-10

-20

Curre

nt D

ensit

y (m

A c

m-2

)

1.00.80.60.40.20.0Voltage (Volts)

ZnTe:Cu ThicknessDevices Contacted at 360°C 1 µm 0.5 µm 0.2 µm 0.1 µm 0.04 µm

Effect of Contacting Temperature

Effect of ZnTe:Cu

Thickness

Insight It’s not as simple

as just accounting for Cu from the contact!

Gessert, et. al, 4th World Conf. PV Energy Conversion, pp.432-435


(100 kHz, 10 mV, -8.0 to +0.6 Sweep)

Comparison of C-V Analysis

1013

1014

1015

1016

1017

1018

N a-N

d (cm

-3)

43210 Depletion Width (um)

ZnTe:Cu ThicknessContact Temperature 360°C 100 KHz, 10 mV Sweep -8.0 V - +0.6 Volts 1 µm 0.5 µm 0.2 µm 0.1 µm

0 V Bias10,000 Å

0 V Bias 5,000 Å

0 V Bias 2,000 Å

0 V Bias 1,000 Å

1013

1014

1015

1016

1017

1018

NA-

ND (c

m-3

)

4321Depletion Width (µm)

ZnTe:Cu Depositon Temperature 240°C 300°C 320°C 340°C 360°C

340°C 320°C

240°C

360°C

300°C

Vary Contact Temperature Vary ZnTe:Cu Thickness




1017

1018

1019

1020

1021

Cu

Con

cent

ratio

n (c

m-3)

43210-1Depth (microns)

ZnTe:Cu CdTe CdS

1.0 µm, 360°C 0.5 µm, 360°C 0.2 µm, 360°C Previous, 1.0 µm, 360° Previous, 1.0 µm, 320°

1017

1018

1019

1020

1021

Cu

Con

cent

ratio

n (c

m-3

)

543210

Depth (µm)

ZnTe:Cu CdTe CdS

240°C 300°C 320°C 340°C 360°C

Vary Contact Temperature Excessive Cu

Optimum Cu

Insufficient Cu

Vary Contact Thickness Excessive Cu ?? Optimum Cu

Insufficient Cu


High-Resolution Compositional

Analysis (SIMS)

Gessert, et. al, 33rd IEEE Photovoltaics Specialists Conf., Paper No. 14



100

80

60

40

20

0

Qua

ntum

Effi

cien

cy (%

)

900800700600500400Wavelength (nm)

24 Volt Heater Insufficient Cu

32 Volt Heater Optimum Cu

36 Volt Heater Excessive Cu

1-Sun Bias+0.4 Volts

QE Comparison Suggests Narrowing Junction

for Excessive Cu Devices

Red-Light Bias QE Confirms Photoconductive CdS

Only For “Excessive” Cu Devices

120

100

80

60

40

20

0App

aren

t Q

uant

um E

ffic

ient

y (%

)

900800700600500400Wavelength (nm)

Excessive Cu_Blue Excessive Cu_Red Excessive Cu_Dark Excessive Cu_White

Optimum Cu_Blue Optimum Cu_Red Optimum Cu _Dark Optimum Cu_White

Insufficient Cu_Blue Insufficient Cu_Red Insufficient Cu_Dark Insufficient Cu_White

Uncorrected for Actual Jsc



1013

1014

1015

1016

1017

1018

N A (

cm-3

)


1e141e11

2e131e17

4e131e16

8e131e15

1e141e14

1e141e13

SCAPS Simulation 100 kHz, Sweep -0.7V to +0.1 Volt

NA Nd NA ND CC46_2e13_1e17 CC46_1e14_1e14 CC46_4e13_1e16 CC46_1e14_1e13 CC46_8e13_1e15 CC46_1e14_1e11

CdTe NACdS ND

4th World Conf. PV Energy Conversion, pp. 432-435 SCAPS1D Simulation

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~320°C (32V, Optimum) ~340°C (34V) ~360°C (36V)

40

30

20

10

0

-10

-20

Curr

ent

Dens

ity (

mA

cm

-2)

1.00.80.60.40.20.0Voltage (Volts)

SCAPS Modeled LIV/DIV CdTe NA_CdS ND

LIV_2e13_1e17, 844 mV, 22.9, 83.1% LIV_1e14_1e14, 893 mV, 24.0, 72.4% LIV_1e14_1e12, 894 mV, 23.9, 64.7% LIV_1e14_1e11, 887 mV, 23.8, 60.8% DIV46_1e14_1e14 DIV46_1e14_1e12 DIV46_1e14_1e11

1013

1014

1015

1016

1017

1018

NA-

ND (c

m-3

)


ZnTe:Cu Depositon Temperature 240°C 300°C 320°C 340°C 360°C

340°C 320°C

240°C

360°C

300°C



Constant Temperature ~300°C Vary ZnTe:Cu Thickness

Constant Thickness ~0.5 µm Vary ZnTe:Cu Temperature

Cu Cu


But…Assumes Minority Carrier Lifetime Does Not Change


Minority Carrier Lifetime

Problem - Previous TRPL studies indicate minority carrier lifetime decreases with Cu

Metzger, Albin, Levi, Sheldon, Li, Keyes, Ahrenkeil, JAP 94(6) (2003)

Wu, Zhou, Duda, Yan, Teeter, Asher, Metzger, Demtsu, Wei, Noufi,

TSF, 515, 5798 (2007)

Demtsu, Albin, Sites, Metzger, Duda,

TSF, 516 p. 2251 (2008)



Why a Problem?

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~240°C (24V) ~280°C (28V) ~320°C ( 32V, Optimum)

Glass

Glass

CdTe If both carrier lifetime and space charge width decreases with Cu incorporation, why does voltage-dependant collection decrease?




Time Resolved Photoluinescence (TRPL)

Glass CdTe

650 nm Excitation

Two Power Levels 250 µW and 2500 µW

820 nm PL Measurement Room Temperature



TRPL Study of Cu Diffusion (As a Function of Contact Temperature)

1

10

100

1000

PL c

ount

s

50n(s)403020100Time

UV700 (25°C) UC699 (200°C) UC697 (240°C)

UC694 (280°C) UC702 (300°C) UV696 (320°C)

UC695 (340°C) UC698 (360°C)

1

10

100

1000

PL c

ount

s

50n(s)403020100Time

UC700 (25°C) UC699 (200°C) UC697 (240°C)

UC694 (280°C) UC702 (300°C) UC696 (320°C)

UC695 (340°C) UC698 (360°C)

250 µW Power 2500 µW Power



Time Resolved Photoluinescence (TRPL) (As a Function of Contact Temperature)

10000

8000

6000

4000

2000

0

Min

ority

Car

rier

Life

time

(ps)

350300250200150100500Approximate Substrate Temperature During Contact (°C)

Insufficient Cu Excessive Cu

Optimum Cu

t1 2.5 mW t1 0.25 mW t2 2.5 mW t2 0.25 mW




Low-Temperature Photoluinescence (LTPL) Increasing Contact Temperature produces peak that that has been associated with a defect complex Cui-OTe , and an unidentified peak at ~1.53 eV

Reference for Cui-OTe

Corwine, Sites, Gessert, Metzger, and Duda, Appl. Phys Lett. 86 (1) 221909

(2005)

103

104

105

106

PL (

Arb

. Uni

ts)

1.601.551.501.451.401.35Energy (eV)

1.463 eV

~1.53 eV

4.25K, 632.8 nm, 1 mW PL from Glass Side

UC700 (25°C) UC697 (240°C) UC694 (280°C) UC696 (320°C) UC698 (360°C)

Cui-O Te

Gessert, et. al, 2008 EMRS Meeting, Strassbourg, France To Be Published Thin Solid Films, 2009.



Time Resolved Photoluinescence (TRPL) (As a Function of Contact Thickness)

1500

1000

500

0Min

ority

Car

rier

Life

time

(ps)

1000080006000400020000ZnTe:Cu Thickness (Å)

NREL Tau 1 (2.5 mW, 360°C) CSU Tau 1 (2.5 mW, 320°C) VTD Tau 1 (2.5 mW, 320°C) VTD (5000 Å, 25°C) Wu (No Cu) Demtsu (No Cu)



What Does it all Mean??

1500

1000

500

0Min

ority

Car

rier

Life

time

(ps)

1000080006000400020000ZnTe:Cu Thickness (Å)

NREL Tau 1 (2.5 mW, 360°C) CSU Tau 1 (2.5 mW, 320°C) VTD Tau 1 (2.5 mW, 320°C) VTD (5000 Å, 25°C) Wu (No Cu) Demtsu (No Cu)

40

30

20

10

0

-10

-20

Curre

nt D

ensit

y (m

A c

m-2)

1.00.80.60.40.20.0Voltage (Volts)

ZnTe:Cu ThicknessDevices Contacted at 360°C 1 µm 0.5 µm 0.2 µm 0.1 µm 0.04 µm

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~240°C (24V) ~280°C (28V) ~320°C ( 32V, Optimum)

30

20

10

0

-10

Curr

ent

Dens

ity (

mA

cm

2 )

1.00.80.60.40.20.0-0.2Voltage (Volts)

~320°C (32V, Optimum) ~340°C (34V) ~360°C (36V)

Constant Cu, Vary Temp Constant Temp, Vary Cu Thickness 2000

1500

1000

500

0

Min

ority

Car

rier L

ifetim

e (p

s)

350300250200150100500Approximate Substrate Temperature During Contact (°C)

Insufficient Cu Excessive Cu

Optimum Cu

t1 2.5 mW


Conclusions

Some CdTe PV Devices Conclusions • Net acceptor level increases during contacting, producing optimum Wdepletion

• Minority carrier lifetime can increase and/or decrease during contacting • Primarily depends on temperature during Cu diffusion • Also depends on amount of Cu diffused • Reason(s) remain uncertain

• For the ZnTe:Cu/Ti contact, the longest lifetimes occur at contact temperature of ~280-320°C

• This may explain various contact-process functionalities!

• Coincident with formation of 1.463 eV peak (~4K) • Anti-correlated with ~1.53 eV Pl peak? (~4K)

• Strategy for optimization of Cu-containing contacts: • Understand effect of temperature on lifetime • Adjust Cu to optimize depletion width and Voc (and limit Cu diffusion into CdS)


Life-Cycle & Mineral Resource

Considerations For Thin Film PV

PV Life Cycle Comparisons - Energy Pay Back Energy Payback Time (EPBT) Considers Energy Consumed During:

Extracting, refining, and purifying raw materials Production, associated materials, and transportation of modules

Installation, balance of systems (BOS) components, and use of modules Module disposal or recycling

Thin FilmsRibbonSi

PolySi

MonoSi a-Si CIS CdTe

ModeledEfficiency 11.5%1 13.2%1

14.0%3 14.0%1 6%2 10%3 9%1

8%3

EPBT(Years)1 1.7 2.2 2.7 - - 1.1

EPBT(Years)2 - 2.1-3.7 - 1.1-3.0 - -

EPBT(Years)3 5 2.5 1.7

1. V. Fthenakis and H. Kim, Proc. 2006 European MRS Meeting, Nice France (6/06) 2. PV FAQs, U.S. Dept. Energy Pub. No. DOE/GO-102004-2040 (and internal references, 12/04). 3. M. Raugie et. al., 20th European PVSC, Barcelona, Spain (6/05)

Ref 1&3: www.nrel.gov/ncpv/thin_film/publications_news Ref 2, www.nrel.gov/ncpv

PV Life Cycle Comparisons - Cadmium Emissions Emissions During Electricity Production from typical

US coal-fired power plant plant (per GWhr = 3600x109 Joules)

2-7 grams Cd (+140 grams collected) 1000 tons CO2, 8 tons SO2, 3 tons NOx, 0.4 tons particulate

Cd emissions for oil-fired power plants 12-14 higher than coal Cd emission also associated with natural gas and nuclear plants

Source: V. Fthenakis and H. Kim, Proc. 2006 European MRS Meeting, Nice France (6/06) www.nrel.gov/ncpv/thin_film/publications_news

PV Life Cycle Comparisons - EcoToxicity

Total Eco-Toxicity Potential

Frameless Polycrystalline Si Modules Frameless CdTe Modules

Glass Si Wafers Electricity

Source: M. Raugie et. al., 20th European PVSC, Barcelona, Spain (6/05) www.nrel.gov/ncpv/thin_film/publications_news

EVA

Other

Mineral Resource Considerations for Thin Films (Si, a-Si, CuInGaSe2, CdTe)

Source: PV FAQs, DOE/GO-102004-1834 (1/04), www.nrel.gov/ncpv

Some Other Conclusions

• Understanding CdTe total eco-toxicity continues to evolve

• Performance of CdTe PV modules would benefit immediately from improved glass and TCO products

• Fastest actual pathways to reduced CdTe PV eco-toxicity 1.  Reduce amount of glass used in module 2.  Reduce electricity used to produce module 3.  Reduce other material inputs (primarily EVA)

• Fastest pathway to reduced CdTe PV political toxicity 1.  Reduce amount of CdTe and Pb-solder used in module


Conclusions

Thank You!

Documents

Review of Photovoltaic Energy Production Using CdTe …sites.ieee.org/scv-photovoltaic/files/slides/2009-March...Tim Gessert General PV Energy Talk, Date Overview US Energy 101 Thin-Film