Design of Molecular Rectifiers

Shriram Shivaraman

School of ECE, Cornell University

Molecular Electronics or “Moletronics”

• Computation using molecules• Replacement devices and interconnects• Key feature : Few molecules per device

Why do we care?

Main issues with conventional scaling:

• Rising costs of conventional fabrication

~ $200 billion by year 2015

• Physical limitations - Leakage currents, Doping non-uniformity

Advantages of Molecules

• Small and identical units

• Bottom up fabrication: Self-assembly by functionalization

• Discrete energy levels – A design handle

• Special properties e.g. flexible substrates and low-cost printing, sensors etc.

Some outstanding issues

• Lack of suitable production methods: Interfacing techniques

• Inherent disorder because of self-assembly: Defect-tolerant architectures

• Speed, Stability, Reproducibility

About this work

Design of molecular rectifiers

Molecular Rectifier

• Aviram and Ratner in 1974

• Donor-spacer-acceptor configuration

• X = e- donating e.g.

-NH2, -OH, -CH3 etc.

• Y = e- accepting e.g.

-NO2, -CN, -CHO etc.

• R = insulating aliphatic group (barrier)

J.C. Ellenbogen et al, Proc. IEEE, Vol. 88, No. 3, March 2000

( ) ( )LUMO LUMO LUMOE E Donor E Acceptor

Working of the Rectifier

J.C. Ellenbogen et al, Proc. IEEE, Vol. 88, No. 3, March 2000

Design of a Rectifier

• Promote charge localization on either side of the barrier : high ΔELUMO

• Shortest aliphatic chain allowing planarity: dimethylene group –CH2CH2-

• Optimal geometries have parallel rings: assumed to be enforced by embedding medium

Candidate Rectifiers

X = -CH3 x 2

Y = -CN x 2

X = -OCH3 x 2

Y = -CN x 2

In-plane Out-of-plane

A B

C D

Method

• Geometries optimized with Gaussian 03

• Ab-initio HF/STO 3-21G basis set calculation

• HOMO/LUMO calculated using Koopman’s Theroem

• Orbitals plotted using Molekel to visualize localization

Results and Discussion:In-plane –CH3 (A)

HOMO -8.99 eV (-9.11 eV)

LUMO2 2.34 eV (2.36 eV)

LUMO1 1.68 eV (1.74 eV)

LUMO3 3.74 eV (3.79 eV)

Results and Discussion:Out-of-plane –CH3 (B)

HOMO -9.03 eV (-8.99 eV)

LUMO2 2.30 eV (2.22 eV)

LUMO1 1.69 eV (1.59 eV)

LUMO3 3.78 eV (3.74 eV)

Results and Discussion:In-plane –OCH3 (C)

HOMO -8.55 eV (-9.23 eV)

LUMO2 2.31 eV (2.17 eV)

LUMO1 1.65 eV (1.52 eV)

LUMO3 3.90 eV (3.49 eV)

Results and Discussion:Out-of-plane –OCH3 (D)

HOMO -8.58 eV (-9.24 eV)

LUMO2 2.28 eV (2.12 eV)

LUMO1 1.67 eV (1.50 eV)

LUMO3 3.88 eV (3.74 eV)

Comparison of ΔELUMO

Molecule Calculated ΔELUMO ΔELUMO [1]

A 2.06 eV 2.05 eV

B 2.09 eV 2.15 eV

C 2.25 eV 1.97 eV

D 2.21 eV 1.99 eV

[1] J.C. Ellenbogen et al, Proc. IEEE, Vol. 88, No. 3, March 2000

Conclusions

• Both molecules A and C have significant intrinsic potential drops (> 2 V)

• They show robustness to out-of-plane rotation

• C seems to have higher built-in voltage from the simulations

Final thoughts

• Koopman’s theorem doesn’t take into account relaxation energies.

• Though that maybe overcome, HF method doesn’t take into account electron correlation.

• DFT and other semi-empirical methods like OVGF(AM1) maybe used. But, they might not always give better results.

Experiments are the only means of knowledge at our disposal. The

rest is poetry, imagination.

-Max Planck