The orientation dependence of the photochemical activity of α-Fe2O3 —
supplemental material
Yisi Zhu, Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador
Carnegie Mellon University, Department of Materials Science and Engineering,
Pittsburgh, PA 15213
Supplemental Material
1. Hematite Structural Information
α-Fe2O3 adopts the corundum crystal structure that consists of layers of close-packed
oxygen anions (O2-) stacked along the c-axis, with iron cations (Fe3+) filling two-thirds of
the octahedral interstitial sites.1 This layered structure is apparent in Fig. S1(a). The
different ideal low-index bulk-truncated surface planes of hematite are non-polar when
they exhibit terminations having the same stoichiometry as the crystal (they are neutral)
and are polar when they exhibit terminations having the with stoichiometries that differ
from the Fe2O3 formula. Surface with excess oxygen (iron) are negative (positive).
The prismatic (01̅10) and (11̅00) planes are non-polar, while the prismatic (12̅10)
and the basal (0001) surfaces are both polar. The bulk-truncated (0001) is shown (viewed
from the side) in Fig. S1(a), with a negatively charged oxygen (positively charged iron)
plane shown as the upper (lower) surface. The bulk-truncated rhombohedral (11̅02) plane
can also be either Fe or O terminated (see Fig. S1(b)), which indicates it could be polar.
However, the iron layers are almost on the same plane as one of the adjacent oxygen
layers (0.35 Å difference), which renders this plane essentially non-polar. In Fig. S1(b),
the same (11̅02) oxygen plane is shown with the upper (lower) surface representing the
1
“The orientation dependence of the photochemical activity of α-Fe2O3 — supplemental material,” Yisi Zhu,
Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador, Journal of the American Ceramic Society,
online, 2015.
polar (nearly non-polar) version. As described in the main document, the real surfaces of
hematite differ considerably from these ideal versions, but their terminations still vary
with orientation and they can be charged.
(a)
(b)
Figure 1. Schematics of hematite viewed along (a) [21̅1̅0] and (b) [11̅01̅]. The dashed
black lines indicate (a) (0001) and (b) (11̅02) surfaces. Oxygen terminated planes are
depicted on the upper surfaces. The lower surfaces have (a) pure iron termination and
(b) mixed termination. The larger red (smaller brown) spheres represent oxygen (iron)
and grey lines represent bonds between them.
2. Kelvin-Probe Force Microscopy
Kelvin-Probe Force Microscopy (KFM) is a high throughput a scanning probe
analytical method that yields a measure of this contact potential difference; such
measurements should therefore allow correlations to be made between the local surface
potential and reactivity. The origin of the contact potential difference can be explained
using the schematics in Fig. 1 of the main document. Fig. 1(a) shows a schematic of the
relevant energies associated with the tip / sample interaction when the surface is neutral
(the bands remain flat) and the work function of the sample, ϕS, is greater than the work
function of the tip, ϕT (ϕS for hematite ≈ 5.5 – 5.6 eV 2,3 and ϕT for the Pt/Cr coated probe
2
“The orientation dependence of the photochemical activity of α-Fe2O3 — supplemental material,” Yisi Zhu,
Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador, Journal of the American Ceramic Society,
online, 2015.
is ≈ 5.1 eV 4). The schematics are drawn with the Fermi levels being equal (they are
grounded) and the contact potential difference (ECPD) represented by the discontinuity in
the reference level (EREF): ECPD = ϕS – ϕT. When the surfaces are charged positively
(negatively), as in Fig. 2(b) (Fig. 2(c)), a negative (positive) space charge develops in the
semiconductor and the bands bend downward (upward) at the surface by ESC, which is
defined as positive for downward bending and leads to ECPD = (ϕS – ESC) – ϕT. Smaller
work function materials and more positively charged surfaces will result in a less positive
(more negative) values of ECPD.
Experimentally, this contact potential difference results in a Coulomb force between
the tip and sample. To enhance the Coulomb interaction, an AC-bias is applied to the tip,
causing it to oscillate, which is detected in the instrument. A DC-bias is then applied to the
tip until the Coulomb force is fully compensated, at which point the oscillations cease. This
dc-bias is defined as the measured surface potential VSP and can be equated to the materials
work function and surface charge (all being relative to those of the tip). VSP is related to the
contact potential difference (shown in Fig. 1) as: VSP = – a × ECPD / e, where a is a
constant that relates to the scan settings (0 < a < 1). Smaller sample-tip distances and larger
AFM probe vibration amplitudes will increase the value of a. Because VSP has the opposite
sign to ECPD, smaller work function materials and more positively charged surfaces will
result in a more positive (less negative) values of VSP. It should be noted that VSP is what is
reported and discussed throughout the main text.
We used a Au-Pt standard sample to calibrate the KFM for every experiment. Because
the work function of Au (5.1 eV) is smaller than that of Pt (5.7 eV),4 VSP should be more
positive for Au than for Pt.5,6 However, we only observed this in 30 % of the experiments.
3
“The orientation dependence of the photochemical activity of α-Fe2O3 — supplemental material,” Yisi Zhu,
Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador, Journal of the American Ceramic Society,
online, 2015.
In the remaining 70 %, Pt had a more positive VSP than did Au. The reasons for this were
not resolved, but could arise from variations in the tip, machine set up, or environmental
conditions.7-10 Whenever the VSP values of Au and Pt in the control sample were reversed
(in order and in sign), so were the VSP values from the Fe2O3 grains. In the main document,
we report only data from datasets exhibiting inverted (reversed) control values, because
these represented the majority (70 %) of experiments. The values reported in the main
document were corrected (sign inverted) such that they agree with the expectations
(presented above for ECPD and VSP for the controls). It should be noted that the data from
the Fe2O3 grains from the 30 % of the experiments in which the potentials of the control
sample had the expected order are consistent with the more extensive data set presented in
the main document.
3. Anisotropic Band Structure
It is possible that anisotropic light absorption, relative to the momentum of the
photogenerated carriers, affects the overall activity, as proposed for SrTiO 3.11 A
calculated band dispersion diagram for hematite Fe2O3 is shown in Fig. 10.12 In this
energy level diagram, Γ represents the center of the Brillouin zone, Γ → Α represents
states with momentum along the c-axis (parallel [0001]), Γ → Μ represents states with
momentum perpendicular to the (11̅02) plane (parallel to a prismatic axis), and Γ→ Κ
represents states with momentum perpendicular to the (11̅00) orientation, traveling in a
direction inclined by 30° to the prismatic axis, along a close packed oxygen direction.
This band structure also gives an explanation for the low activity of (0001) faces of bulk
hematite. The direct band gap for electrons traveling perpendicular to this face is on the
order of 0.3-0.5 eV larger than for electrons traveling toward the prismatic and
4
“The orientation dependence of the photochemical activity of α-Fe2O3 — supplemental material,” Yisi Zhu,
Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador, Journal of the American Ceramic Society,
online, 2015.
rhombohedral faces. The larger band gap is expected to result in fewer charge carriers
generated in that direction than for the other faces. However, the band dispersion shows
little difference in the band gap when comparing the shaded regions, which includes to
the (12̅10), (11̅00), (11̅02) directions. Similarly, the bulk conductivity is maximized in
directions perpendicular to [0001] (i.e., in the basal plane). However, we observe the
highest activity to occur for the (11̅02) plane, whose normal is approximately midway
between the [0001] and the basal plane directions. Again, while the anisotropy of this
bulk property supports the poorly active nature of the (0001) basal plane, it cannot
explain the higher activity of the (12̅10) plane compared to the prismatic planes.
Figure S2. A calculated12 electronic band structure for hematite, with the bands rigidly
shifted to reflect the experimentally observed band gap. The shaded region represents the
momentum states where vertical transitions are possible on absorption of blue light.
5
“The orientation dependence of the photochemical activity of α-Fe2O3 — supplemental material,” Yisi Zhu,
Andrew M. Schultz, Gregory S. Rohrer, and Paul A. Salvador, Journal of the American Ceramic Society,
online, 2015.
4. References
1. K. Sivula, F. le Formal, and M. Grätzel, "Solar Water Splitting: Progress using
Hematite (α-Fe2O3) Photoelectrodes," Chemsuschem, 4 [4] 432-49 (2011).
2. C. M. Eggleston, A. J. Shankle, A. J. Moyer, I. Cesar, and M. Grätzel, "Anisotropic
photocatalytic properties of hematite." Aquat. Sci., 71 [2] 151-159 (2009).
3. E. R. Batista, and R. A. Friesner, "A self-consistent charge-embedding methodology
for ab initio quantum chemical cluster modeling of ionic solids and surfaces:
Application to the (001) surface of hematite (α-Fe2O3)." J. Phys. Chem. B, 106
[33] 8136-8141 (2002).
4. H. B. Michaelson, "The work function of the elements and its periodicity." J. Appl.
Phys., 48 [11] 4729-4733 (1977).
5. M. Nonnenmacher, M. P. O’Boyle, and H. K. Wickramasinghe, "Kelvin probe force
microscopy," Appl. Phys. Lett., 58 [25] 2921-23 (1991).
6. S. Mugo and J. Yuan, "Influence of Surface Adsorption on Work Function
Measurements on Gold-Platinum Interface Using Scanning Kelvin Probe
Microscopy," J. Phys.: Conf. Ser., 371 [1] 012030:1-4 (2012).
7. M. Rohwerder, and F. Turcu, "High-resolution Kelvin probe microscopy in corrosion
science: Scanning Kelvin probe force microscopy (SKPFM) versus classical
scanning Kelvin probe (SKP)." Electrochimi. Acta, 53 [2] 290-299 (2007).
8. T. Hochwitz, A. K. Henning, C. Levey, C. Daghlian, and J. Slinkman, "Capacitive
effects on quantitative dopant profiling with scanned electrostatic force
microscopes." J. Vac. Sci. Technol. B, 14 [1] 457-462 (1996).
9. A. K. Henning, T. Hochwitz, J. Slinkman, J. Never, S. Hoffmann, P. Kaszuba, and C.
Daghlian, "Two dimensional surface dopant profiling in silicon using scanning
Kelvin probe microscopy." J. Appl. Phys., 77 [5] 1888-1896 (1995).
10. D. Ziegler, and A. Stemmer. "Force gradient sensitive detection in lift-mode Kelvin
probe force microscopy." Nanotechnology, 22 [7] 075501 (2011).
11. J. L. Giocondi, P. A. Salvador, and G. S. Rohrer, "The origin of photochemical
anisotropy in SrTiO3," Top. Catal., 44 [4] 529-33 (2007).
12. M. N. Huda, A. Walsh, Y. Yan, S.-H. Wei, and M. M. Al-Jassim, "Electronic,
structural, and magnetic effects of 3d transition metals in hematite," J. Appl.
Phys., 107 [12] 123712:1-6 (2010).
6