Characterizing order in organic and inorganic semiconductors - - PowerPoint PPT Presentation

Characterizing order in organic and inorganic semiconductors

Paulette Clancy Cornell University KDI Order Team NSF Review June 2002

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Cornell KDI Order Team
• Focus and scientific objectives
• Computational tools
• Scientific advances in order-related topics
• Necessity for a team approach
• Summary of current status
• Recommendations to NSF

Overview of the presentation

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Faculty

Paulette Clancy

• Chem. and Biomolec. Engr.

HARD Fernando Escobedo

• Chem. and Biomolec. Engr.

SOFT Edwin Kan Electrical and Computer Eng. HARD George Malliaras Materials Sci. and Engr. SOFT Michael P. Teter Physics HARD Michael O. Thompson Matls. Sci. and Engr. HARD

• Grad Students

Six PhD students; 3 departments; 50% women and minorities Aleksandra Chojnacka* MSE Devashish Choudhary CBE Leonard Harris* CBE Chungho Lee ECE Chin Lung Kuo CBE Ritesh Shetty CBE Michelle Swiggers* MSE

* Denotes women or minority students

• Undergraduate Researchers

25 students over 3 yrs; 50% women and minorities

• Visitors

David Dunlap (UNM, Alb.) SOFT

• External Affiliates

Lawrence Livermore National La., IBM Corp; Intel Corp., Kodak Corp.

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Team Members

Focus:

Explore the relationship between order and properties (structural, electrical, mechanical)

Scientific Objectives:

• Develop a coherent language for the expression of order

(semantic and mathematical)

• Develop computational tools to express this language and to

create partially ordered systems

• Understand and manipulate order-property relationships at

interfaces (e.g. metal/Si; organic/metal; organic/Si)

Gas Crystal Liquid Glass Amorphous

Order 1

Group Focus and Objectives

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Ambiguous terminology of order in semiconductors
• Despite much work on simulating amorphous Si, no convincing best

way to create a realistic description.

• Separate families of empirical intermolecular models: one for hard

and one for soft materials.

• Scale of organic semiconductors a challenge for electronic structure

(ES) packages.

• Multiscale calculations reveal that the accuracy of ES needs to be

higher.

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Challenges

• How do we create a-Si (or a-Ge) models that can be used in

dynamical simulations? [MC and MD routes]

• How well does virtual a-Si resemble real a-Si?
• Is a-Si a unique phase?
• Can we create an electronic structure code with LDA accuracy but

tight-binding speed?

• How does charge transport occur in organic semiconductors?
• How is this mediated by the presence of an interface?
• How does this knowledge translate into working electronic

devices? Creation → → → →Characterization → → → → Knowledge → → → → Application

Initial Project Objectives

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Scope, work plan, and inter-relationships in the project

Develop “Signature Cell” order parameter (FE)

CT

Create + Characterize a-Si, a-Ge (PC)

CA

Nature of order-disorder (locally/globally) in hard semiconductors

I

Understand a-Ge→c-Ge transformation as exemplified in explosive crystallization (MOT,PC)

E CA

Develop Virtual Gibbs ensemble MC code for solids (FE)

CT

Charge Transport in dissimilar systems

I

Create + Characterize metal nanodots on Si (ECK,PC)

E CA

Develop faster and more efficient ab initio tool (MPT)

CT

Understand charge transport in hard/soft heterosystems (All)

I

Use existing ab initio tools ITR : Arias + DFT++ EPSRC : Gillian/Bowler UK CONQUEST Create + Characterize pentacene/Si interfaces (KMC, MD, statics) (FE,PC)

CA

Enhance charge transport in pentacene/Si systems (GM)

E CT CT CA I E

Computational tool Computational analysis Experiments Information Key to symbols

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Meta-project”: Pentacene/Si systems

Understand charge transport in hard/soft hetero systems In-situ X-ray characterization of hard/soft interfaces (MOT)

Creation and characterization of

• rder and charge

transport in pentacene/Si systems

(FE, PC) Electrical characterization of metal nanodots on pentacene (ECK)

Characterization of charge transport in pentacene/X (and other organic semiconductor/X) systems (GM)

Compare to existing codes: DFT++ & CONQUEST New ab initio code (MPT, PC)

CT CA I E

Computational tool Computational analysis Experiments Information E CT E CT CA E Key to symbols

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• “Signature Cell” method to determine order on a local or global
• scale. Richer in information than traditional schemes
• New electronic structure code (Harris-like formalism but with

“tunable” accuracy (at the expense of computational efficiency)

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Computational Tools

Potential exact match at global minimum

= Signature = Configuration

Phase space U(r)

• 1

σ σ σ σ1 σ σ σ σ2 r Pseudo-potential

= Configuration atoms = Signature atoms Find best match

Signature Cell Method (Illustration)

Perform MC moves Rotation moves Breathing moves Superimpose signature Parallel tempering swaps

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

4.05 4.1 4.15 4.2 4.25 4.3 4.35 4.4 4.45 1 2 3 4 5 6 Computational Timesteps (106) Co-ordination no Molecular Dynamics Self-Guided MD Hyper MD

Particle transfer attempt Volume transfer attempt Box I Box II Faked Move on a box

irtual Gibbs ensemble MC method. Allows MC simulation of liquid-solid transitions ccelerated MD schema in a new arena. Transformations of bulk solids;

kinetics.

Computational Tools

Evolution of a quenched liquid Silicon using various MD schemes: Hyper MD and Self Guided MD (Stillinger Weber potential, Isothermal Isobaric simultaion)

• Created a new route to a-Si and a-Ge. Accelerated MD makes

‘quench and anneal’ viable.

• Evidence for a-Si as a unique phase; prediction of the structural

lifetime of (as yet undiscovered) glassy Si.

• Created an order map for post-explosively crystallized Ge

Information about order-property relations

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

0.4 0.6 0.8 1.0

h

• 1 [µm
• 1]

550 600 650 700 750 800

Tmin, Tcolumnar, Tend [K] hx

• 1

no explosive crystallization (film too thin or substrate temperature too low) T

m i n

= 4 7 7 K + h

• 1

3 2 6 µ µ µ µ m K Tcolumnar = 6 4 6 K + h-1 1 2 4 µ µ µ µ m K Tend = 826 K s c a l l

• p

e d

• r

t r a n s i t i

• n

a l m

• r

p h

• l
• g

y columnar morphology (high-temperature plateau)

hend

• 1

Morphology diagram for Germanium

Vbatt

Vo~ 1/Rsample

Rsample

Vo time

t=0

Explosive Crystallization (Illustration)

laser liquid amorphous crystalline Cr

Steady state conditions: Tcl > Tal vcl = val

• Surprising evidence for the width of the mediating liquid layer in

explosive crystallization and its implications for modeling.

• Demonstrated that alignment in pentacene films can be induced

by surface alignment layers used in liquid crystals, leading to pronounced increase on transport properties

Information about order-property relations continued…

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Charge transport in organic-inorganic junction Pentacene-Silicon interface

Information about order-property relations continued…

• Use physical and electrical methods to characterize self assembly of

metal nano-crystals on semiconductors and insulators.

Collaboration with ab-initio calculation to build realistic models.

• Demonstrate improved nonvolatile memory device characteristics

by using metal nanocrystals

• Demonstrate new carrier injection schemes 2 to 3 orders of

magnitude better than conventional contact schemes by using metal nanocrystal interfaces

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

CMOS Device Nano-crystal growth

• f Au on SiO2

• Fertile ground for hard/soft cross-talk.

– (Lively) weekly meetings of the whole team crucial to understand each

• ther’s vocabulary, approaches, challenges.

– No such coordinated effort on hard-soft interfaces possible without this award. – KDI award has redirected the research focus of all the ‘hard’ faculty

• Freedom to explore new skills

– Two computational specialists created experimental components to their research – Two experimentalists added theoretical aspects to their research

• Enhanced the education of students

– Grads and undergrads. exposed to the richness of these inter-related projects – “Grad sharing” by several faculty made easy – Enhanced gender diversity provides important educational benefits – Industrial/gov. labs involvement offers internship opportunities.

Importance of a Team Effort

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Enhanced faculty mentoring

– Research focus makes mentoring of young faculty natural rather than contrived (proposal writing, advising students, working with industry) – Important for this team (half were brand new)

• Leverages interest by industrial concerns and government labs

– New contacts with LLNL, Intel, IBM – Deepened relations with IBM, Kodak, and others

• Broadened experimental/computational interactions

For our topic, linking experiments to computational efforts isn’t important, it’s everything!

Importance of a Team Effort (continued)

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Allows a meaningful Outreach Program. Here’s what we

learned…. College students:

• Involving under-rep. groups in research proven to direct students’

careers in research (85% of our UG researchers went to grad. school

in science and engineering)

• Linking engineering to society is a key strategy in retaining under-

represented groups in engineering. See recent article in Wall St. J.

• Tell good women and minority grad. students to consider a faculty

career and mentor them towards this goal. Simple but effective.

Importance of a Team Effort (Continued)

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

Outreach Program (continued)… Grade school students:

• Surveying where women “leak” from the system helped focus our

efforts

• You can’t neglect any grades in making science appealing, but

high school is the key right now.

• College students are great role models
• Target all the constituents (students, teachers, guidance counselors)
• Use the Media effectively (web, newspapers)

11% 25%PhD 50% High School 25% BS

Women in Physical Science and Engineering at Cornell

Faculty

Importance of a Team Effort (Continued)

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Significant advances in understanding how to create, characterize,

and provide information about the link between order and property

• relationships. A position paper on this will be drafted at the conclusion of

this award (12/02).

• Innovative developments of new computational tools and effective

synthesis of new and existing computational tools.

• Production of a cohesive effective research team with unique

hard/soft research skills.

• Significant educational successes (2PhDs awarded, 5 in training; ~15

publications so far; richer education; 50% under-rep. groups).

• Important advances in Outreach to under-represented groups

– 50% undergrad. women involved – New High School honors program in Phys. Sci./Engr. under development

Summary Research Educational

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group

• Make 5-year awards standard for group efforts (3 years is too short;

review after 3 years)

• Allow 1 extra page per PI for ITR proposals (like NIRTs).
• Harvest the collective experiences of this assembled audience

w.r.t. effective and ineffective Outreach strategies and make web- accessible.

• Continue to support affirmative action. Enlist our help to make the

case to the government if necessary.

• Continue to offer team-based research opportunities. The benefits

far outweigh the administrative efforts to keep teams successful.

Recommendations to NSF

Cornell Engineering Knowledge and Distributed Intelligence Group Cornell Engineering Knowledge and Distributed Intelligence Group