Yttrium Oxide Upconverting Phosphors
J. Phys. Chem. B, Vol. 107, No. 7, 2003 1551
addition, very weak features observed between 1079 and 1127
3
3
nm are assigned to the F2 f H4 transition. The assignments
of the present work are given in Table 1 and Figure 9 of this
article.
The two-photon process explains how the 1D2 and 1G4 levels
are populated. The population of the lower energy levels can
be explained by a one-photon process.
Thus, the anti-Stokes emission from cubic Y2O3:Tm3+ under
632.8-nm excitation involves a decay by ca. 2500 cm-1 of the
3
Tm3+ ion down to the F4 state, after initial excitation to the
3F2 state, before the absorption of a second photon of 632.8-
nm light. This two-photon absorption mechanism, which results
in anti-Stokes emission, is different from those occurring when
cubic Y2O3:Er3+ and cubic Y2O3:Eu3+ are excited by 632.8-
nm light because the Er3+ and Eu3+ ions do not decay from
their initially excited states before the absorption of a second
photon.1-3 This would be reflected in the time evolution of these
signals, which is beyond the scope of the present work.
To make the assignments of the emission bands shown in
Figure 2, the emission spectra were studied over the temperature
range of 0 to -190 °C, and the laser power dependence of the
band intensities was investigated.
Figure 5. Emission bands of Tm3+ in Y2O3 in the 740- to 830-nm
region. The spectra are shown in the order of increasing temperature
from bottom to top: -190, -175, -150, -125, -100, -75, -50, -25,
and 0 °C.
Temperature Dependence of Emission Peaks (0 to -190
°C). In Figure 3, the temperature dependence of the 1D2 f 3H4
transition is shown from 0 to -190 °C. It is apparent that the
three small bands at 446.8, 448, and 450.5 nm (see Figure 3)
that disappear at -190 °C are clearly behaving differently from
the four bands above 451 nm, at 452.1, 452.6, 453.1, and 453.9
nm, which increase in intensity as the temperature is lowered
to -190 °C. The obvious explanation is that the three small
bands are hot bands originating from thermal population of
higher-energy Stark sublevels of the 1D2 state, which has 2J +
1 ) 5 components, and this thermal population decreases as
the temperature drops. All of the other bands in Figure 3 behave
Stokes side, the Raman bands for the Y2O3 lattice have been
discussed previously.1
The bands starting at 648.5 nm and extending to 683 nm (see
Figures 4S and 5S in Supporting Information) show three main
types of temperature behavior between 0 and -190 °C. Those
at 648.7, 649.3, 650.5, 651.2, 652.1, 654.8, 662.8, 663.7, 669.3,
and 673 nm are most intense at 0 °C and least intense at -190
°C except for the last of these bands (673 nm), which also shows
some intensity at -125 °C. The bands at 653.6, 656.2, and
660.6, the five spanning 671 to 672.5, and all of those from
673.6 to 683 nm are maximized at -125 °C. The bands at 661.6
and 665.1 nm are still increasing in intensity as the temperature
is lowered toward -190 °C. The bands are assigned to the 1G4
f 3H4 transition (see Figure 2b). The temperature dependence
of these bands is dependent on the relative populations of the
Stark sublevels of the 1G4 state from where the emissions arise.
The final envelope of emission bands is shown in Figure 5.
Two different temperature dependencies are seen for these
bands. One subset sharpens as the temperature approaches -190
°C whereas the other almost disappears at -190 °C. In Figure
5, it is apparent that all of the features from 765 to 825 nm
sharpen as the temperature decreases, and in several cases, the
intensities of the emission bands are much higher at -190 °C
than at 0 °C. In the 745- to 765-nm region, the opposite behavior
is apparent because the intensities of the bands decrease as the
temperature decreases (particularly that at 750.4 nm). In this
case, the different thermal behavior is due to the fact that there
are two different transitions that give rise to the emission bands
1
identically, in keeping with them all arising from the D2 f
3H4 transition.
The emission bands observed between 474 and 500 nm are
presented in Figure 4. The bands in the range of 475-485 nm
sharpen considerably, and that at 480 nm splits on going down
in temperature, but the other bands in the 485-500-nm range
do not split and show very little sharpening. The only transition
1
that can possibly explain these two groups of bands is the G4
f 3H6 transition. At -190 °C, two of the bands not only have
narrowed but also have split into a number of components. These
are the bands around 480 nm (four bands at -190 °C but only
one band at 0 °C) and that at 483 nm (which is a doublet at
-190 °C and almost a singlet at 0 °C). The most likely
explanation for these observations is a change in the Maxwell-
Boltzmann distribution of populations of the Stark sublevels as
the temperature decreases. There is a small triplet at 515.5 nm
at 0 °C that sharpens to five bands (514.4, 514.6, 515.4, 515.6,
and 516.3 nm) at -190 °C (see Figure 3S of Supporting Infor-
mation). This group of bands has a different temperature de-
pendence than the group of bands between 474 and 500 nm
because the former is still increasing in intensity as the tem-
perature is lowered toward -190 °C whereas the latter has maxi-
mum intensity at -50 °C (see Figure 4). This triplet is assigned
3
3
in this region. The main transition is F4 f H6, and a second
transition accounts for the bands between 745 and 760 nm and
is discussed in Laser Power Dependence (see below).
There was no indication in this study for different temperature
behavior of the Tm3+ ions on the C2 and S6 crystallographic
sites of the cubic Y2O3 lattice; therefore, it is not possible to
identify Stark levels of Tm3+ ions on either site. However, there
is a higher probability that the cations are located on the C2
sites because there are 3 times as many as the S6 sites.
The concentration of Tm3+ ions is 1:100 of the total cations.
Because the phosphor material was prepared by homogeneous
precipitation, it is reasonable to assume that clustering of Tm3+
ions has not occurred. It would therefore be reasonable to expect
1
3
to the D2 f H5 transition. In addition, there is a sixth band
around 513 nm, which is apparent at the lower temperatures
and is also assigned to this transition. Again, the most likely
explanation for these observations is a change in the relative
populations of the Stark sublevels as the temperature decreases.
All of the other features between 517 nm and the laser line
on the anti-Stokes side are anti-Stokes Raman bands.1 On the