Theory of MX Chain Compounds - Material Designs and New Physical Properties in MX-and MMX-Chain Compounds - page 134

calculated for the dp model. The appropriate parameter set ðS 1 ; S 2 Þ¼ð 0 : 2 ; 0 : 072 Þ

gives the effective width of 2 meV.

8.3.3 Photoinduced Metallic States

As was first demonstrated by Iwai et al., strong photoirradiation on these materials
can convert them into a type of metallic states [ 40 ]. As has been discussed already,
one-photon absorbed states are well described as a bound or unbound
doublon-holon pair. Then, the next question is about the nature of the strongly
excited states. In more detail, weak or intermediate excitation also causes some
photoinduced effect, which is characterized as a photoinduced absorption within
the gap region. However, this is not a so-called Drude peak, which is usually
centered at 0, being a clear marker of the metallic state. We therefore know

that there is some threshold for the formation of this metallic state in these
compounds. This makes a contrast to the carrier-number dependency known for
the one-dimensional Hubbard model, in which the singularity exists only at the half
filling.
To understand this discrepancy, one key is the existence of the e-l interactions.
There is a possibility that the e-l interactions exist to some extent and work as an
inhibitor to the metallic state. However, as we have already discussed, the e-l
interaction that works between the halogen stretching and the metal d z 2 orbital is
suppressed effectively in the limit of low carrier density. Much more detailed
analyses about the time scales of polaron formations, actual associated modes,
and the excitation-density dependency will be necessary for better understanding.
Another key will be the result by Takahashi et al., which traced numerically the
real-time evolution of photoexcited states within the Hubbard model [ 41 ]. Their
results show some critical intensity of photoexcitation realizing the metallization
and seems to be consistent with the observation. This result is therefore attractive,
since it suggests a different situation in the photoexcited states from the doped
cases, more specifically, a trend that the nature of the excited state is changed
drastically beyond a certain amount of excitations. Unfortunately, the numerical
calculation is limited to a small system size and does not allow a definite conclu-
sion. A proof based on analytical treatments will be further required to finish this
argument.

o ¼

8.4 Bridging CDW and Mott Insulators

So far, we have argued the properties of the CDW insulator and the Mott insulator
from each side, although we already discussed them from the universal point of
view, i.e., the phase diagram. The phase diagram describes a phase transition
between the two phases. In this sense, we implicitly expect a conversion between