1. Low Alpha Materials and Metrology in the IC Industry
Current Status and Future Requirements
Dr. Brett M. Clark, Honeywell
Jeff Wilkinson, Medtronic
July 25, 2005
Thank you Dr. Cushman for the invitation and introduction.
My name is Brett Clark, I’m a development scientist with the Electronic Materials division of Honeywell.
We develop many high purity materials for the semiconductor industry.
In this talk I will discuss some of the the materials used in the IC industry and some of the challenges facing the the industry both in terms of material purity characteristics and the ability to measure those characteristics.

2. Outline Soft errors & IC development Current Status
Materials
Types
Purity / Activity requirements
Metrology
Alpha emissivity
Instrument capability
Future Requirements
Material specifications
Changes on the horizon
Purity /Activity requirements become more stringent
Improvements
Snapshot of what I see in the IC industry currently and some possible paths in the future.
Discuss instrumentation challenges to meeting industry roadmaps and what will be required in the future.

3. Integrated Circuits and Soft Error Upsets
Energetic particles depositing energy/charge in critical nodes
Radioisotopes in component materials
Cosmic ray contributions
Microelectronic design trends
Flip Chip
Smaller geometries/decreasing line widths
Increased vulnerability to soft error mechanisms
No expert on soft errors, defer to Jeff Wilkinson Colleague for more specifics
Soft errors = significant issue in the future

4. Material purity critical to reliability
IC Materials Overview
High purity metals and alloys
Cu, Al, Ta, W, Ti, Pb, Sn, Ag, Ru
Range in purity from 99.99% to %
Alpha Activity Requirements
0.02 hr‑1cm‑2 : early 1990’s
0.01 hr‑1cm‑2 : late 1990’s
0.002 hr‑1cm‑2: 2001
hr‑1cm‑2:2006
Primary alpha emitters
210Pb in Pb/Sn solders
U & Th
No expert on soft errors, defer to Jeff Wilkinson Colleague for more specifics
Material purity critical to reliability

5. Current Material Challenges: Contamination
Any material with U and/or Th above 1 ppb
Some 99.99% materials fail
Sn, In
Possible contaminants
Refractory materials and ceramics
Abrasives
Atmospheric dust and debris
Commercial metals & alloys
All stages of manufacturing & processing must be controlled & monitored
Quickly gain an appreciation for how ubiquitous U& Th are in the terrestrial environment
Assume any material is an alpha contamination
hazard until proven otherwise

6. Current Instrumentation Capability
Industry uses Gas Proportional Counters
Area :1000 cm2
Geometry: 2
Background
2-3 cph optimal, 4-6 nominal
No energy spectroscopy capability
Limited ability to identify contamination sources
0.002 analysis requires ~7 days counting time for 20% RSD
Ordela 8600
Alpha Sciences 1950
No expert on soft errors, defer to Jeff Wilkinson Colleague for more specifics
Current instrumentation incapable of timely analysis

7. Precision/Counting Time trade off

8. Counting time determined by detector size

9. Alpha counting rates relative to background
Solution: Increase signal and/or decrease noise

10. Background effect on counting time

11. Material Requirements - Future
0.001 hr‑1cm‑2 material in manufacturing
Technology roadmaps: hr‑1cm‑2 in 2006
2 ’s from 1 square meter area per hour!
With current measurement technology this would take 228 days to measure with 20% relative precision
Where is the specification bottom?
Design dependent
Where influences other than cosmic are insignificant
Measurement times compatible with manufacturing schedules
Instrumentation improvement required

12. Instrumentation for the Future
Measure increasingly low activity levels in timely manner without loss of precision
Increase signal/noise ratio
Increase detector area without increasing background
Optimum solution S/N requires background<0.5 cph
Current progress
Large area Si detectors
Cr-39 hole counting
Low background facilities
Eliminate cosmic component
Enable higher purity materials in detector construction
Decrease detector background
Assess detector quality
Validate production methods and materials
Chicken and egg - how to build low background detector without out measuring activity of components low enough
Significant opportunity for improvement

13. Large area - Zero background - High precision
Best case scenario
Large area - Zero background - High precision

14. Conclusions
IC improvements continue to require lower activity materials
As material activity decreases, contamination risk increases
Current instrumentation has significant limitations
Long analysis times
No spectroscopic capability
Future activity specifications require more than incremental improvement in instrument capability
Ultra Low background instruments will be necessary to meet these specifications