118
6 Advanced Models for Practical Devices
the penetration of electrons into the semiconductor does not cause any charging at
the interface. However, the position of the CNL is usually different from that of the
Fermi level, resulting in the generation of interface charge. This charge causes the
potential gap in Fig. 6.4, which is the origin of nonideality. One can understand
that the MIGS is intrinsic. The CNL and the S parameter will be discussed in more
detail in Sect. 6.3.
6.2.2 Disorder-Induced Gap States (DIGS) Model
The concept of the DIGS model is that disorder in the atomic arrangement is introduced at the interface upon interface formation, which breaks the periodicity and
generates gap states in the band gap. In some cases, disorder might be introduced
unintentionally, i.e., a disordered arrangement in the thermodynamic equilibrium
state. In other cases, disorder can be introduced intentionally, i.e., an artificial disordered arrangement in the nonequilibrium state. Therefore, the DIGS can be either
intrinsic or extrinsic.
In the interface-bonding picture without an atomic viewpoint, the polarized chemical bonding [7] or Coulomb potential [8] at the interface is considered. The polarized chemical bonding results in the formation of an electric dipole through interface
bonding, and a similar equation for nonideality to that in the MIGS model is satisfied
[7]. In this picture, however, interface bonding can be artificially induced, so the
resulting nonideality can be either intrinsic or extrinsic. The Coulomb potential at
an interface of one atom thickness was initially proposed for a Si–insulator interface, not a metal–semiconductor interface, and the interface was regarded as being
in thermodynamic equilibrium [8]. The authors of Ref. [8] showed that the Coulomb
potential is linearly dependent on the atomic distance at the interface for the series
of alkali-earth metals, indicating that the charge of the dipole is the same and only
atomic distance is different for different alkali-earth metals. Since Ref. [8] concerns
the Si–oxide interface, no discussion on the S parameter was given.
6.2.3 Interface-Induced Gap States (IFIGS) Model
Regardless of whether the gap states originate from the MIGS or DIGS, it is important
to know which part is intrinsic or extrinsic for the control of the SBH including the
ohmic contact. In the IFIGS model, the origin of the intrinsic part of the induced
gap [6, 9] can be either a MIGS or a DIGS. By comparing metal–semiconductor
interfaces with a non-ionic semiconductor such as Si or Ge and those with an ionic
semiconductor such as ZnS, it is suggested that if the MIGS decay length is large,
the DIGS part is negligible [6].
6 Advanced Models for Practical Devices
the penetration of electrons into the semiconductor does not cause any charging at
the interface. However, the position of the CNL is usually different from that of the
Fermi level, resulting in the generation of interface charge. This charge causes the
potential gap in Fig. 6.4, which is the origin of nonideality. One can understand
that the MIGS is intrinsic. The CNL and the S parameter will be discussed in more
detail in Sect. 6.3.
6.2.2 Disorder-Induced Gap States (DIGS) Model
The concept of the DIGS model is that disorder in the atomic arrangement is introduced at the interface upon interface formation, which breaks the periodicity and
generates gap states in the band gap. In some cases, disorder might be introduced
unintentionally, i.e., a disordered arrangement in the thermodynamic equilibrium
state. In other cases, disorder can be introduced intentionally, i.e., an artificial disordered arrangement in the nonequilibrium state. Therefore, the DIGS can be either
intrinsic or extrinsic.
In the interface-bonding picture without an atomic viewpoint, the polarized chemical bonding [7] or Coulomb potential [8] at the interface is considered. The polarized chemical bonding results in the formation of an electric dipole through interface
bonding, and a similar equation for nonideality to that in the MIGS model is satisfied
[7]. In this picture, however, interface bonding can be artificially induced, so the
resulting nonideality can be either intrinsic or extrinsic. The Coulomb potential at
an interface of one atom thickness was initially proposed for a Si–insulator interface, not a metal–semiconductor interface, and the interface was regarded as being
in thermodynamic equilibrium [8]. The authors of Ref. [8] showed that the Coulomb
potential is linearly dependent on the atomic distance at the interface for the series
of alkali-earth metals, indicating that the charge of the dipole is the same and only
atomic distance is different for different alkali-earth metals. Since Ref. [8] concerns
the Si–oxide interface, no discussion on the S parameter was given.
6.2.3 Interface-Induced Gap States (IFIGS) Model
Regardless of whether the gap states originate from the MIGS or DIGS, it is important
to know which part is intrinsic or extrinsic for the control of the SBH including the
ohmic contact. In the IFIGS model, the origin of the intrinsic part of the induced
gap [6, 9] can be either a MIGS or a DIGS. By comparing metal–semiconductor
interfaces with a non-ionic semiconductor such as Si or Ge and those with an ionic
semiconductor such as ZnS, it is suggested that if the MIGS decay length is large,
the DIGS part is negligible [6].