390 likes | 560 Vues
Charge Transport and Related Phenomena in Organic Devices Noam Rappaport, Yevgeni Preezant, Yehoram Bar, Yohai Roichman, Nir Tessler Microelectronic & Nanoelectronic Centers, Electrical Engineering Department, Technion, Haifa 32000, Israel Oren Tal, Yossi Rosenwaks,
E N D
Charge Transport and Related Phenomena in Organic Devices Noam Rappaport, Yevgeni Preezant, Yehoram Bar, Yohai Roichman, Nir Tessler Microelectronic & Nanoelectronic Centers, Electrical Engineering Department, Technion, Haifa 32000, Israel Oren Tal, Yossi Rosenwaks, Dept. of Physical Electronics, Faculty of Engineering, Tel Aviv University, Tel Aviv 69978, Israel Calvin K. Chan and Antoine Kahn Dept. of Electrical Engineering, Princeton University, Princeton NJ 08544, USA
Charge Transport • Short Introduction (or the things we tend to neglect) • Simplified Device-Oriented Approach • FETs (charge density & Electric Field) • PN Diodes • Thin film device HighlightInconsistencies
e x Hopping conduction model • Conjugated segments “States” • Charge conduction non coherent hopping
What are the important factors? • What is the statistics of energy-distribution? • What is the statistics of distance-distribution? • Is it important to note that we are dealing with molecular SC? Do we need to use the concept of polaron? • Energy difference • Distance • Similarity of the Molecular structures
Detailed Equilibrium Ej Ei Anderson:
Anderson (Miller Abrahams) E Polaron Picture Q Configuration co-ordinate
20% 100% 20% H. Bassler, Phys. Stat. Solid. (b), 175,15, (1993) Morphology or TopologySpatial (Off Diagonal) Disorder
Mean Medium Approximation Energy X0-DX X0 X0+DX Physical picture is GREATLY relaxed to allow for Charge Density and Electric Field effects is a single model Y. Roichman & Nir Tessler 2003
Mean Medium Approximation(at low charge density) Dashed line – Fit using s/kT 3 C=2.5*10-4 S=2 4 5 6 7 Y. Roichman, et. al., Phys. Stat. Solidi a-201 (6), 1246-1262 (2004)
mPk “Unified” percolation models predict much higher density dependence: Mean Medium Approximation s=4kT s=5kT s=7kT
Limitations of the MMA • Assumes equilibrium • Can not model intrinsically non-equilibrium density of states (as exponential). • Can not model a sample with preferential paths (percolation). • Assumes that the sample is uniform on the length scale defined by the distance between contacts. But it actually shares the assumptions used with any device model
Role of MW Choose material that shows less mixed phases. Shaked et. al., Adv mat, 15,913, 2003
Eliminate parasitic currents Close topology
Transistors • Mobility = ? • Need to account for: • It is density dependent (varies along the channel) • Real DOS is not single Gaussian (density dependence is “unknown”) • Develop a method for a general density dependence -8 VDS= -4 -4 -3 -2 -2 -1 -1 O. Katz, Y. Roichman, et. al, Semicond. Sci. Technol. 20, 90-94 (2005) N. Tessler & Y. Roichman, Organic Elect., 2005.
m Field dependent at 2-3x103V/cm. A longer length scale, as in correlation, is required • Extracting m • Use low VDS • Do not use mPk K= 0.38 -4 -1 But the polymer is MEH-PPV
Can we use MMAto describe LEDs too Can we use Semiconductor Device model to describe 100nm thick device?
To model LEDs we need to be able to predict the charge density distribution inside the device Current continuity Eq. charge density distribution inside a device is governed by D/m Generalized Einstein-Relation: Y. Roichman and N. Tessler, Applied Physics Letters 80, 1948 (2002).
Simple expression to fit them all For the MMA model: N. Tessler & Y. Roichman, Organic Elect., 2005.
Use General Einstein Relation to Model Junctions • Semiconductor / Semiconductor (PN diode) • Metal / Semiconductor (contact)
Organic /Organic Junction Exponential DOS P N mnK Ideality Factor:
Energy Energy x x o o Distance, Distance, x x UBand D D eff eff , , max max = U -Ex F D D + + + + U U U U image image F F Model the contact to LED as a transport problem
Transport in contact region & bulk is modeled using semiconductor equations Equilibrium at the contact interface defines the charge density on the organic side Model the contact to LED as a transport problem Gaussian nature m and D (or D/m) are functions of density Y. Preezant and N. Tessler, JAP 93 (4), 2059-2064 (2003). Y. Roichman, et. al., Phys. Stat. Solidi a-201 (6), 1246-1262 (2004)
Model the contact to LED as a transport problem Results: We could reproduce effects of – barrier temperature …. BUT – each experiment required a different physical set of parameters to make the fit quantitative. Take home message: 1. The Device Model doesn’t work well 2. The contact region and bulk may be governed by a different picture Arkhipov – non equilibrium at the contact leads to injection that is limited by hops into a Gaussian DOS (1nm insulating gap will make it valid). Baldo – The metal enhances disorder at the contact region only V. I. Arkhipov, et. al., Phys. Rev. B 59 (11), 7514-7520 (1999). B. N. Limketkai and M. A. Baldo, Phys. Rev. B 71, 085207 (2005
~10nm Are there Concerns Regarding LEDs 100nm Filaments ?
A linear part a dominant mobility Time of flight measurement (excitation = step function) m Current Time But step function is better suited for the understanding of devices as it has the same steady state!
Transient measurement(in thin, 300nm, films) Low excitation density Photocurrent Hard to find a linear slope
Mobility Distribution Saturated Pathways Unsaturated Pathways • Thick Films: V.I. Arkhipov, E.V. Emelianova, G.J. Adriaenssens, H. Bässler, • J. of Non-Crystalline Solids 299-302 (2002) • 2. Thin Films (experimental): R. Österbackaet al., Synthetic Metals 139,811-813, 2003
Fitting with Model Transient Fit to Measurement Mobility Distribution Function
~10nm Are there Concerns Regarding LEDs 100nm Filaments ?
Motion on a 3D grid Note: the “long jumps” are due to the cyclic conditions at Y & Z axis
Monte Carlo(Hopping in Gaussian DOS) Monte-Carlo Drift-diffusion
Conclusion • Are all types of devices “seeing” the same microscopic physical picture? (don’t think so) • New description for dispersive transport (Filaments, Stability?)
Spatial (Off Diagonal) Disorder C=3x10-4 (cm/Vs)0.5
6 6 5 5 5 5 4 4 4 4 4 4 3 3 3 3 3 3 3 2 2 2 2 2 2 2 1 1 1 1 0 0
Equilibrium conditions (existence of a Fermi level + constant temperature) Generalized Einstein-Relation (Ashcroft, solid state physics) To model LEDs we need to be able to predict the charge density distribution inside the device Current continuity Eq. charge density distribution inside a device is governed by D/m Y. Roichman and N. Tessler, Applied Physics Letters 80, 1948 (2002).