Chemistry Reference
In-Depth Information
Fig. 5.7 . With increase of x ph from 0.12 to 0.5, the photoinduced bleaching observed
around the CT band is enhanced, while the spectral intensity in the IR region
decreases slightly as seen in Fig. 5.7 . These results suggest that the negative values
of
0.5) may be attributable
to the miscount of the spectral weight in the IR region below the lower energy
bound (0.12 eV) of the measured spectral range.
Here, let us comment on the validity of the KK analyses on the TR spectra. When
applying the KK analyses on the TR spectra, there are two important effects to be
considered [ 56 , 59 ] (1) carrier concentration changes depending on the distance
from the sample surface and (2) absorption depths of the probe light l r and the pump
light l p are different. It is, therefore, necessary to check carefully the validity of the
analyses. In the experimental results of the Ni-Br chain compound, the spectral
shape of the TR due to the midgap absorption observed for x ph <
D N eff above 1.5 eV for x ph >
0.1 (especially for x ph ¼
0.012 is almost
unchanged and, therefore, will not be considerably affected by those two effects.
However, it is reasonable to consider that the absolute values of the TR and
s
are
460 ˚ ). To evaluate the two effects on
the Drude-like reflection band observed for x ph >
somewhat underestimated, since l r > l p (
¼
0.1, we postulated a metallic
state expressed by a simple Drude model with the thickness l t (200-2,000 ˚ ) on the
surface of the Ni-Br chain compound and simulated the R and
s
spectra. The result
1,000 ˚ , spectral shape and absolute
of the simulation demonstrates that for l t >
value of R and
are independent of l t . In the Ni-Br chain compound, the thickness
of photoinduced metallic state (region with carrier concentration
s
>
0.1) exceeds
1,000 ˚
for x ph >
0.2. So we can consider that the Drude-like reflection band
observed for x ph >
0.2 will not be considerably influenced by the two effects. For
the intermediate excitation density (0.02
0.2), it might be necessary to
take account of some errors in the analysis. Nevertheless, the observed systematic
changes of
< x ph <
) and the approximate holding of the sum rule over the wide
range of x ph ensure that the analysis using the KK analyses presented here reflects
well the photoinduced changes of the electronic state.
In Fig. 5.8a ,
D N eff (
o
0.1 ps are plotted as a
function of x ph . D N eff (1.0 eV), which represents the total spectral weight trans-
ferred from the CT band to the inner-gap region, saturates for x ph > 0.04. In
contrast,
D N eff (1 eV) and
D N eff (0.2 eV) at t d ¼
D N eff (0.2 eV), i.e., the spectral weight accumulated in the lower energy
region between 0.12 and 0.2 eV, increases almost linearly with x ph up to x ph ¼
0.1.
The steady increase of
0.04 is attributable to the growth of
the Drude weight. Let us compare the x ph dependence of
D N eff (0.2 eV) for x ph >
D N eff with the chemical-
doping-density ( x ) dependence of N eff in the 2D cuprates previously reported,
which is shown in Fig. 5.8b [ 50 , 54 ]. In Fig. 5.8b , the solid and open circles show
the x dependence of N eff at 1.5 eV and that for the Drude component in the IR
region, respectively. The observed clear resemblance of the x ph dependence of
D N eff in the Ni-Br chain compound with the x dependence of N eff in the 2D cuprate
demonstrates that the Mott transition is driven by the photocarrier doping in the
Ni-Br chain compound.
A remarkable aspect of the photoinduced Mott transition is the influence of the
electron-hole asymmetry on the transient optical spectra. Being distinct from the
chemical doping case, photoexcitation generates both electrons and holes in
Search WWH ::




Custom Search