good means to determine the average thickness of the nanocrystal shells of different materials.
11.6.2
Valence Band Photoemission
Colvin et al. [104] were the first to study the valence band photoemission of CdS
quantum dots. To avoid charging of the samples, they attached CdS nanocrystals
on metal surfaces such as gold (for 38 A ˚ diameter dots) or aluminum (for 24 A ˚ and
70 A ˚ diameter dots). The quantum dots were anchored to the metal surfaces using
bifunctional organic groups. The experiments were carried out on a synchrotron
source with variable photon energies in the range 20–70 eV. They did not observe
any change in the spectra due to the presence of different metals. The metal surfaces are deep inside and the escape depth of electrons for this range of photon
energies varies between 10 and 30 A ˚ . In this work, it was found that the top of
the valence band of the dots varies as a function of the size in accordance with the
theoretical predictions based on the finite potential-EMA model that includes the
polarization term as well.
Van Buuren et al. [105] performed photoemission and X-ray absorption experiments on Si nanocrystals to determine the TVB and BCB shifts, respectively, as a
function of size. The Si nanocrystals were grown in situ at 1700
C in an Ar gas
buffer of 112 mTorr followed by hydrogen exposure to passivate the surface. The
resolution of the photoemission and absorption measurements carried out on a
synchrotron radiation source were 0.25 eV and 0.05 eV, respectively. They observed
a valence band to conduction band shift ratio of 2:1 for all sizes of Si nanocrystals.
This is in agreement with various calculations reported for Si nanocrystals [106].
Soft X-ray emission spectroscopy (SXE) and soft X-ray absorption spectroscopy
(SXA) are techniques that also can map out the states in the valence and conduction bands, respectively. Using SXE and SXA, Lu ¨ning et al. [107] studied
four different sizes of CdS nanocrystallites namely, cadmium thiolate Cd 8 (SR) 16
Cd 17 S 4 (SR) 26 (13.5 A ˚ ), Cd 32 S 14 (SR) 36 (17.5 A ˚ ) and a CdS nanoparticle of 40 A ˚ diameter. Apart from the band gap variation, they also observed a systematic decrease in the band width of the upper part of the valence band, which is mainly
comprised of S 3p states; the width decreases from 4.8 eV for CdS bulk to 3.5 eV
for Cd thiolate. However, the lower part of the valence band which is derived from
the more localized S 3s and Cd 4d states, is hardly affected by size quantization. To
determine the shifts in the TVB and BCB, the S 2p core level photoemission spectra for all the nanocrystals were used to calibrate the energy scale internally. The
shifts were compared with the IP-EMA, finite potential EMA (FP-EMA) and the
sp
3 s* TB model. The FP-EMA gives a good description, with the finite potential
barrier adjusted to obtain the best fit to the experimental results but the other two
models do not provide good agreement with the experimental results.
Another method to avoid charging of the semiconductor samples is to mix the
samples with a conducting matrix such as graphite. Nanda et al. [108] measured
11.6 Photoemission Studies 399
11.6.2
Valence Band Photoemission
Colvin et al. [104] were the first to study the valence band photoemission of CdS
quantum dots. To avoid charging of the samples, they attached CdS nanocrystals
on metal surfaces such as gold (for 38 A ˚ diameter dots) or aluminum (for 24 A ˚ and
70 A ˚ diameter dots). The quantum dots were anchored to the metal surfaces using
bifunctional organic groups. The experiments were carried out on a synchrotron
source with variable photon energies in the range 20–70 eV. They did not observe
any change in the spectra due to the presence of different metals. The metal surfaces are deep inside and the escape depth of electrons for this range of photon
energies varies between 10 and 30 A ˚ . In this work, it was found that the top of
the valence band of the dots varies as a function of the size in accordance with the
theoretical predictions based on the finite potential-EMA model that includes the
polarization term as well.
Van Buuren et al. [105] performed photoemission and X-ray absorption experiments on Si nanocrystals to determine the TVB and BCB shifts, respectively, as a
function of size. The Si nanocrystals were grown in situ at 1700
C in an Ar gas
buffer of 112 mTorr followed by hydrogen exposure to passivate the surface. The
resolution of the photoemission and absorption measurements carried out on a
synchrotron radiation source were 0.25 eV and 0.05 eV, respectively. They observed
a valence band to conduction band shift ratio of 2:1 for all sizes of Si nanocrystals.
This is in agreement with various calculations reported for Si nanocrystals [106].
Soft X-ray emission spectroscopy (SXE) and soft X-ray absorption spectroscopy
(SXA) are techniques that also can map out the states in the valence and conduction bands, respectively. Using SXE and SXA, Lu ¨ning et al. [107] studied
four different sizes of CdS nanocrystallites namely, cadmium thiolate Cd 8 (SR) 16
Cd 17 S 4 (SR) 26 (13.5 A ˚ ), Cd 32 S 14 (SR) 36 (17.5 A ˚ ) and a CdS nanoparticle of 40 A ˚ diameter. Apart from the band gap variation, they also observed a systematic decrease in the band width of the upper part of the valence band, which is mainly
comprised of S 3p states; the width decreases from 4.8 eV for CdS bulk to 3.5 eV
for Cd thiolate. However, the lower part of the valence band which is derived from
the more localized S 3s and Cd 4d states, is hardly affected by size quantization. To
determine the shifts in the TVB and BCB, the S 2p core level photoemission spectra for all the nanocrystals were used to calibrate the energy scale internally. The
shifts were compared with the IP-EMA, finite potential EMA (FP-EMA) and the
sp
3 s* TB model. The FP-EMA gives a good description, with the finite potential
barrier adjusted to obtain the best fit to the experimental results but the other two
models do not provide good agreement with the experimental results.
Another method to avoid charging of the semiconductor samples is to mix the
samples with a conducting matrix such as graphite. Nanda et al. [108] measured
11.6 Photoemission Studies 399