Organic Electronics

(ELEC324)

Carrier concentrations and

Transport processes in Disordered
Organic semiconductors
 Recap from last lecture
 Organic materials posses conducting properties due to conjugation i.e.
overlap of unhybridised pz orbitals and presence of alternating double and
single bond, which results in the formation of HOMO and LUMO ( - *)
– i.e. equivalent valence and conduction bands respectively

 In disordered organic semiconductors, the density of states (DOS) consists

of discrete energy levels, commonly described by a Gaussian distribution
which can be approximated to an exponential, assuming the transport level
is within the Gaussian ‘tail’

 Alternative function i.e. Laplace distribution is also proposed for describing

the DOS which fits the exponential dependency over a wider energy range

ELEC324 - [2018]
 Overview on today’s lecture
 Determination of the carrier (electron or hole) concentration
 Number of occupied states per unit volume at a given energy i.e. how
many of the existing states are filled up with electrons at a given energy

 Proposed transport mechanisms

 Dependency of mobility with carrier concentration – also known as the
Universal Mobility Law (UML)
 Dependency of mobility on field and temperature

ELEC324 - [2018]
 Carrier Distribution Functions
 In order to understand the transport mechanisms in organic semiconductors, it is
essential to determine the concentration of the carriers i.e. electrons or holes
present within the LUMO and HOMO levels

 However we first need to understand how such carriers are distributed within
the discrete energy levels, now we know the number of available states at
each energy i.e. DOS as discussed in the last lecture

 From Quantum/Classical Mechanics, the occupancy of such discrete states by

particles can be described by probability distribution functions1 such as
Fermi-Dirac (FD) and Maxwell-Boltzmann’s (MB) statistics, depending on type
of particle or system

 Fermi-Dirac is the statistical distribution of elementary particles which are

indistinguishable from each other i.e. particles such as fermions e.g. electrons
or protons, and follows Pauli’s Exclusion Principle, whilst Maxwell-Boltzmann’s
statistics describes statistical distribution of material particles over various energy
states and applicable to distinguishable particles (by mass, spin or trajectory)

 Note the type of statistics used influences the possible states of the system
ELEC324 - [2018]
 MB vs. FD
 In Maxwell-Boltzmann statistics, the particles
are distinguishable: State 1 State 2
 Consider two particles A and B, and two single AB empty
Possibility 1
states i.e. state 1 and state 2, which particles B
A Possibility 2
may occupy then the possibility of such
particles occupying these two states is as B A Possibility 3
shown in the diagram. There is a total of 4 empty AB Possibility 4
states in the system. Half of the states have the
particles bunched in the same state whilst other
half have particles in separate states

 In Fermi-Dirac statistics, the particles are

indistinguishable: State 1 State 2
 Both particles may be labelled A, however they A A Possibility 1
have opposite spin. From Pauli’s Exclusion
Principle (PEP), no more than one particle can
be in any one single state
 In this case, the state of the system is as in the
diagram, where there is a total of only one
state of the system as a whole. The states are
not bunched up by particles due to Pauli’s
Exclusive Principle
ELEC324 - [2018]
 a. Fermi-Dirac (FD) Distribution
 Fermi-Dirac distribution gives the probability that a state
f FD E  
at energy E in thermal equilibrium, is occupied by an 1
electron. The expression is given on the side, where k is  E  EF 
1  exp  
the Boltzmann constant, T is the absolute temperature  kT 
and EF is the Fermi level where the probability fFD (E) of
finding a filled energy state is half (0.5)

 Such a system is also characterised by its temperature for example:

(EF) (EF)

At 0 K, there are no electrons above valence At high temperature, some electrons can reach
band since none have energy above EF and the conduction band and contribute to current
there are no available states in the band gap ELEC324 - [2018]
 b. Maxwell Boltzmann (MB) Distribution
 For disordered and doped systems, the   E  EF 
 exp    E  EF
preferred distribution function is the f MB E     kT 
Maxwell-Boltzmann statistics given by  1 E  EF

the expression on the side

 Here the focus is mainly on the carriers

above the Fermi level

 Below the Fermi level EF, the carriers are

assumed to be trapped and immobile,
thus do not contribute to the conduction
(which is attributed to the free rather than
trapped carriers)

 The probability of finding a state occupied

by an electron decreases exponentially
with increase in energy above the Fermi
level EF

ELEC324 - [2018]
 Carrier (Electron) Concentration
 Assuming an n-type semiconductor, the total carrier (electron) concentration
can be determined by integrating density of states (DOS) function with the
distribution function, with respect to energy as below:


ntotal  

f ( E ) N ( E ) dE

 The total carrier concentration includes the trapped charges (i.e. below Fermi
level EF) and free charges (above Fermi level EF). However we are more
interested in free carriers i.e. those contributing to current thus the limits of
integration can be defined from EF

 Assuming the density of states (DOS) to be given by the Laplace distribution and
the Maxwell-Boltzmann statistics for the distribution of the carriers, we can
determine the free carrier concentration nfree as in equation below:


n free  
EF
f MB ( E ) N '( E ) dE

ELEC324 - [2018]
 ‘Free’ Carrier (Electron) Concentration
 The integration describes an area under which 
the two distribution functions i.e. fMB (E) and
N’(E) overlap, and within an energy range from
n free  
EF
f MB ( E ) N '( E ) dE

EF to infinity. Note the limits can be changed

 Both functions are exponential

however one rises i.e. N’(E) whilst
the other falls i.e. f(E) rapidly i.e.
At high energies, there are more
empty states however fewer
carriers, thus no conduction free
carriers
 For conduction, there needs to be
an optimum number of unoccupied ET
states and sufficient carriers
present within a few kTs from EF
i.e. where the transport level ET trapped
exists. For simplicity the transport carriers
level can be considered the edge
of the LUMO level (or conduction
band in silicon) for n-type
ELEC324 - [2018]
 Free Carrier (Electron) Concentration
 Substituting for the Laplace distribution and Maxwell-Boltzmann, we have

   N '  0   E 

  E  EF
n free  
EF
exp  
  kT
 
   kTc
exp   dE
 c 
kT

 Upon solving the integration, we have,

To E 
n free  N '  0  exp  F 
Tc  kTc 

1 1 1
where  
To T Tc

To is the characteristic temperature for the distribution of carriers

TC is the characteristic temperature for the distribution of traps (disordered material)
and T is the absolute temperature i.e. 300 K

Exercise: Derive the above expression, and compare the carrier concentration using the
two different DOS distributions i.e. Exponential and Laplace
ELEC324 - [2018]
 Charge Transport Mechanisms
 Unlike band-like transport in conventional crystalline silicon, the transport
mechanisms in disordered organic semiconductors is due to hopping of the
carriers between the discrete states

 Most hopping models for organic materials originate from the Miller-Abraham
jump rate model2. The transition of carrier from an occupied site i to
unoccupied site j is given as:

  E j  Ei 
 2 rij   exp    , if E j  Ei State empty
vij  vo exp      kT 
   
 1 , if Ei  E j State filled

where o describes the attempt to escape frequency, α is a parameter dependent on the quantum
mechanical overlap between the states, and rij is the distance between the sites i and j with
energies Ei and Ej respectively.

 Analysis on the distance the carriers can hop is extended in the Variable range
hopping (VRH) model, which suggests that carriers may hop over large
distances with low activation energy rather than small distances with high
activation energy
ELEC324 - [2018]
 Charge Transport in Conjugated Polymer

Zero field

Applied field

Hopping transport Hopping transport

ELEC324 - [2018]
 Other Hopping Transport models
Some of the models based on hopping of carriers between states:

1. Disorder model3
 The model (simulated using Monte Carlo) takes account of both spatial and energetic
disorder assuming Gaussian DOS

log e    1 T 2 log e    F
1
2
and

where T is temperature and F is the field

2. Multi-trapping and release models5

 Charge transport occurs via hopping between states above a critical energy.
Below the critical energy, the states act traps
 Charges are released from the traps and excited to a transport level as result
of increase in temperature
 Ea 
  o  exp  
 kT 
where µo is the mobility above a transport level, α is the ratio of the effective density of states at
the transport level to concentration of traps and Ea is the mean energy ELEC324 - [2018]
 Other Hopping Transport models
3. Polaron models5
 Based on the formation of polarons – a self-localised charge or coupling of the charge
with lattice deformation associated with the charge
 Polaronic motion take place via succession of random jumps in which each electron
hop to neighboring site
 Polaron hopping was studied further by introducing the coincidence concept where the
energy states of predecessor site coincided with the successor site

4. Bipolaron models6
 Also known as resonance tunneling model developed by Zuppiroli for doped disordered
polymers of doping levels greater than 1018 cm-3
 Electron transport is co-related either due to adiabatic jumps within polaronic clusters or
non-adiabatic jumps from cluster to cluster
 Clusters are formed due to fluctuation in dopant concentration
 Dopant behave as a potential and also as a tunneling bridge between neighboring
chains

ELEC324 - [2018]
 Other Hopping Transport models
5. Universal Mobility Law
 In organic semiconductors, mobility is strongly dependent on the carrier concentration -
an empirical relationship known as the Universal Mobility Law given as:

  K nm

Where K is a constant and m is a constant associated to the Meyer-Neldel energy

given as,

Tc
m 1
T

 Further discussion on this empirical relationship will be given in the next lecture

 In terms of the transport mechanism, an increase in concentration results in

increase in mobility and subsequently enhanced conduction

ELEC324 - [2018]
 Significance of VRH on UML assuming Laplace DOS

Increase in carrier concentration

At much high energy, even results in increase in mobility
though there are more states, due enhance hopping of carriers
the density of carriers is low
as the EF is much lower, and
thus no conduction

LUMO (assuming n-type)

ET Transport energy ET is assumed to

be close to EF i.e. in inorganic
semiconductor, this would be at the
edge of conduction band
Electron
energy

 Initial carrier introduced by doping/field-effect occupy the lower state where the hopping
distances are larger thus the carriers are trapped and immobile
 Further increase in concentration results in higher level states being occupied where the
hopping distance are smaller hence increase in mobility ELEC324 - [2018]
 Summary
 Fermi-Dirac and Maxwell Boltzmann statistics gives the probability of occupancy
of states by particles

 The concentration of the free and/or trapped carriers can be determined using
the integration of the density of states with distribution functions

 Hopping mechanisms is believed to be the common transport mechanism in

conjugated polymers. Additional effects such as the presence of polarons and
dopants can alter the mechanism

 The mobility of the charge carriers depends on the temperature, field and carrier
concentration

ELEC324 - [2018]
 Revision Questions
1. Describe the difference between Fermi-Dirac and Maxwell-Boltzmann statistics

2. Derive the expression for carrier concentration assuming Exponential

dependency of the DOS and Maxwell-Boltzmann statistics on occupancy

3. Describe the Universal Mobility Law in Organic semiconductors. How does this
dependency compare with inorganic semiconductors i.e. Silicon

Next Lecture

Patterning processes

ELEC324 - [2018]
 References
1. S. M. Sze, “Physics of Semiconductor Devices”, 3rd Edn. Wiley, NJ 2007.

2. A. Miller and E. Abraham, Phys. Rev., vol. 120 (3), pp. 745, 1960

3. N. F. Mott and E. A. Davies, “Electronic Processes in Non-crystalline Materials”, 2nd Edn,

Oxford University Press, London, 1979

4. G. Horowitz, J. Mater. Chem., vol. 9, pp. 2021, 1999 and J. Phys. III 5, pp. 355, 1995

5. D. Emin and T. Holstein, Ann. Phys., vol. 53, pp. 439, 1969

6. L. Zuppiroli et. al., Phys. Rev. B, vol. 50 (8), pp. 5196, 1994

ELEC324 - [2018]