A supply chain technology roadmap

Would and ITRS-type PV

roadmap make sense for the
industry?

Dr. Christopher Case

The Linde Group

3rd PV Fab Managers Forum

9 March 2009 - Dresden, Germany
All roadmaps are wrong –
some roadmaps are useful

• Discussion items
– A bit of roadmap history
– Roadmapping and innovation
– The International Roadmap for
Semiconductors (ITRS)
– PV – the parallel universe
– Gases and chemical challenges
– Supply chain roadmap straw man

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

1
Roadmap(ping) history

• Developed and first used in the 1970s by

Motorola
• By 1980 - in common use by Motorola, TI,
Intel and Japanese firms such as Hitachi
and Toshiba

Early Motorola product roadmap

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

2
Generic roadmap

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

3
The ITRS sponsors and objective

• Sponsored by the five leading chip

manufacturing regions:
– Europe, Japan, Korea, Taiwan, and the United
States
• Objective:
– ensure cost-effective advancements in the
performance of the integrated circuit and the
products that employ such devices, continuing
the health and success of the industry
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
4
ITRS participation demographics

Over 1300 volunteer participants

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

5
ITRS methodology

ITRS methodology

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

6
 2008 Barrier/Nucleation/Resistivity
Year of Production 2007 2008 2009 2010 2011 2012 2013 2014 2015

MPU/ASIC Metal 1 ½ Pitch

68 59 52 45 40 36 32 28 25
(nm)(contacted)

Barrier cladding thickness

4.8 4.3 3.7 3.3 2.9 2.6 2.4 2.1 1.9
Metal 1 (nm)
Conductor effective resistivity
3.51 3.63 3.8 4.08 4.30 4.53 4.83 5.2 5.58
(µΩ-cm) Cu Metal 1

Year of Production 2016 2017 2018 2019 2020 2021 2022

MPU/ASIC Metal 1 ½ Pitch 22 20 18 16 14 13 11

(nm)(contacted)
Barrier cladding thickness
1.7 1.5 1.3` 1.2 1.1 1.0 0.9
Metal 1 (nm)
Conductor effective resistivity
6.01 6.33 6.7 7.34 8.19 8.51 9.84
(µΩ-cm) Cu Metal 1

• ALD barrier processes and metal capping layers for Cu are

lagging in introduction
• Resistivity increases due to scattering and impact of liners
•No known practical solutions
Courtesy: C Case, presentation ITRS 2008 Winter Conference – 9 December 2008 Seoul. ROK
Roadmap challenges

• Overcoming “red brick wall”

• Path dependency: how to balance on-and off-roadmap
innovation
• Caution of becoming too unwieldy and prescriptive
• “Roadmap” may not be the best metaphor - implies
certainty
• Maintaining voluntary participation
• Increasing cost of roadmapping
• Tends to emphasize incremental innovation
• Participation mix between suppliers and regions mixed

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

8
Why is the roadmap successful?

• Roadmap has adapted and evolved along with the industry

it serves.
– Driven by common purpose
– Technology specific but adaptable
• Process is public
– Collaborative
– Organised
• Strong industry commitment ensures that it is kept up-to-
date and reflects the most complete collective knowledge of
technology needs.
• Healthy “beat-the-Roadmap” competitive behavior helps
reinforce success.
• Evolves with the industry – now includes ESH and emerging
materials
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
9
Semiconductor innovation

• ITRS roadmap emerged from the recognition that

the productivity curve, known as Moore’s Law
could only be maintained with sustained
technology efforts
• Often offered as a model or template for other
roadmap efforts
• Semiconductor technology innovation is usually
characterized as being evolutionary
• Roadmapping serves this type of organized
innovation well
• Collecting a great amount of knowledge in great
amount of detail
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
10
PV – the parallel universe

• Similar technology age

• Crystalline and thin film improvements
evolutionary
• Productivity enhancements from tool
suppliers and automation
• 20% + CAGR
• Consortia and university technology pipeline

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

11
No materials roadmap

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

12
Supply Chain challenges – order of success

• Semiconductor – success, roadmap driven

• Flat panel displays – success, no roadmap,
leveraged learning from semiconductor
• MEMS – sufficient, not roadmap driven
• Solid state lighting – entering scaling phase
which will challenge suppliers
• Printable electronics - entering scaling
phase which will challenge suppliers
Where can we adapt existing learning and
standards to PV?
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
13
PV Supply Chain Roadmap Opportunities

• Bulk solar cell fabrication or thin film PV cell fabrication

– Chemical and gas general specifications
• Defect specifications
• Purity specifications
– Chemical hazards
– Utilization
• Silicon and polysilicon supply
• Wafer manufacturing and slicing
• Substrate and cover glass specifications
– TCO
• Module assembly and balance of systems
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
14
 Silane demand forecast

Total PV Market Silane Demand

16000
C-Si TFS
14000
Silane Market (TPA)

12000

10000

8000

6000

4000

2000

0
2006 2007 2008 2009 2010 2011 2012

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

15
Key gases –
typical supply modes

Gas Supply Mode for large scale (>350 MW)

3000kg / 6000kg ISO modules

Silane
Pipeline to local production for >1GW

Cleaning gas NF3: 4000kg / 8000kg ISO modules

Fluorine: On-site F2 generator

Hydrogen On-site generator (Natural gas based)

with compressed or liquid H2 backup
Nitrogen On-site generator with liquid N2 back up

Dopants 0.5% in H2 blend supplied in cylinders

On-site blending from high purity source
3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
16
 Key gases –
critical planning considerations
Gas Safety Global Lead time Space Fab Ramp
Supply Required Implications
Availability
Silane 333 333 333 333 33
NF3 33 33 3 3 33
Fluorine 33 333 333 3
Hydrogen 333 333 333 33 333
Nitrogen 3 3 333 33 333
Dopants 333 3 3
Helium 3 3

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

17
 Thin film silicon – evolution of critical
material consumption and supply modes
NM3/hr 900 4,000 8,000 12,000

N2

Mode Bulk liquid (3-6mths) Small N2 gen. (9-12mths) Large N2 gen. (12-18mths)

NM3/hr 300 1,400 2,700 4,000

H2

Mode Tube trailers / Small SMR Medium SMR Large SMR

T/yr 50 250 500 750

SiH4
Dedicated small scale prod. Dedicated small scale prod.
Mode
ISO (6T) (~24 months) (~24 months)

T/yr 64 320 640 1,025

Cleaning
Gas NF3 ISO (8T)
Mode Small on-site F2 Modular on-site F2 Modular on-site F2
65MW 330MW 660MW 1,000MW

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

18
Crystalline silicon – evolution of critical
material consumption and supply modes

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

19
 Dopant mixtures - cylinder delivery
logistics are an issue for larger fabs
100% dopant gas cabinet

Blender
Mixer

Buffer Chamber
Blender

PH3
Hydrogen gas feed

•Typical delivery is in cylinders as 0.5% •On-site blending with H2 reduces

PH3 or TMB mixtures in H2 cylinder handling and logistics
•65 MWp > 400 cylinders / yr • 1GWp < 40 cylinders / yr
•1GWp > 6,000 cylinders / yr

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

20
Key gases – typical purities

Gas Role Purity

Silane SiH4 (kg/yr) Active layer Si 99.9997%
deposition
Nitrogen trifluoride NF3 Chamber clean 99.9%
(kg/yr)
Hydrogen H2 (Nm3/hr) Active layer Si 99.999%
deposition
Nitrogen N2 Pump purge and 99.999%
(Nm3/hr) chamber vent
TMB 0.5% in H2 (Nm3/yr) p-dopant 99.5%

PH3 0.5% in H2 (Nm3/yr) n-dopant 99.9998%

Argon Ar (Nm3/yr) Inert gas for sputter 99.9997%

Methane CH4 (Nm3/yr) Co-dopant 99.97%

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany
21
 2008 - Unchanged
Fig 3
Production Ramp-up Model and Technology Cycle Timing
100M
Development Production 200K
10M
Volume (Parts/Month)

Volume (Wafers/Month)
20K

1M
2K
100K Alpha Beta Production
Tool Tool Tool 200

10K First Two

First Companies 20
1K Conf. Reaching
Papers Production 2

-24 -12 0 12 24
Source: 2005 ITRS - Exec. Summary Fig 3 Months 22
ITRS 2008 Update – April, Konigswinter, Germany
Summary - materials roadmap straw man
• To ensure
– lowest cost per watt
– availability and security of supply
– safety
– best usage of industry resources
• Roadmap technical specifications including
– purity
– utilization per wafer/substrate
– delivery and packaging technology
– new material requirements for TCO
– sustainability metrics – GWP potential, manufacturing
energy

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

23
All roadmaps are wrong

a photovoltaic materials supply

chain roadmap would be useful

3rd PV Fab Managers Forum - 9 March 2009 - Dresden, Germany

24
Linde provides essential infrastructure and materials for
the global photovoltaic industry

The Linde Group is a leading gases and engineering company with almost 52,000 employees working in around
100 countries worldwide. In FY2008 it achieved sales of EUR 12.7 billion.

In photovoltaics capabilities include:

• Global supply of critical materials and services
• Innovation in gas and chemical based solutions
• Turnkey engineering
• Lowering emissions
• Enabling environmentally sustainable manufacturing
• Reducing manufacturing costs and improving yields

Linde’s PV customer experience is leading with crystalline and

thin film Si customers in every major manufacturing geography.

TF Line supplier WW total Linde total

OEM “A” 14 8

OEM “B” 10 6

OEM “C” 6 3

The content of this presentation is confidential and should not be distributed to a third party without the prior authorization of Linde
25