Reviewers' comments:
Reviewer #1 (Remarks to the Author):
The manuscript here presents time
-resolved two
-photon photoemission studies of Cu2O
photoelectrons. The manuscript is interesting. However, I was confused by the overall message. On
the one hand the authors attempt to motivate the work to demonstrate that
the technique is useful for
studying the surface carrier dynamics of photoelectrons. In that regard they determine the effect of
the carrier dynamics between a bare Cu2O surface and one with a Pt overalayer catalyst. On the other
hand the authors discuss the important of Cu2O. An present the work to understand the carrier
dynamics at the surface of this important semiconductor. Since there are significant differences in the
response for the bare and Pt overlay sample this is something to learn about how the
se
photoelectrons are working. But I didn't get a clean message as to what the Pt is doing.
Thus, the biggest issue with this manuscript is that there is no insight for how one might develop a
Cu2O photoelectrode that is not impacted by the defects. Furth
ermore, the measurements are done in
ultrahigh vacuum
- which seems to lesson the impact of the measurements and the measurement
technique somewhat. While the authors do make an attempt to justify why these measurements are
still valid.
The discussion an
d presentation of Fig. 2 is confusing. What determines the zero energy scale here? Is
the Fermi energy not determined by the system for these measurements? Why is the CB at a lower
energy than the VBM and shouldn't the Fermi energy be between the CB and VBM. I see that the DB is
at a lower energy than the CB which would make sense. But it's unclear from this plot how the energy
diagrams of Fig. 4 are constructed. Maybe this is well known for the tr
-2PPE community (but this is
probably small).
Could the au
thors change the excitation wavelength. This would help determine if the rise time in Fig.
3 is dues to diffusion of carriers within the bulk to the surface. How do the authors determine the CB
electrons first relax to DB and then transport to the surface
. Couldn't CB diffuse to the surface and
then relax to the DB? The authors rule this out due to the fast decay of surface CB to the DB and then
the subsequent arrival of DB electron density. However, if the relaxation of surface CB electron to the
DB is fast then one might expect a similar behavior. As CB electrons arrive at the surface (in
coherently) they quickly relax to the DB.
The main conclusion is that photoelectrons relax to bulk defect states prior to reaching surfaces and
thus loose their potential very fast interior to the material. Does this mean that Cu2O cannot be fixed?
If the defect are due to Cu vacancies is there a way to decrease their number and see how that
impacts the results.
Overall. I like the manuscript
- but not sure that this
should be published in Nature communication.
The impact of the manuscript seems rather low. Maybe the authors should use the results to improve
the operation of the photoelectrons.
Reviewer #2 (Remarks to the Author):
In the manuscript ‘Femtosecond time
-resolved two
-photon photoemission studies of ultrafast carrier
relaxation in Cu2O photoelectrodes’ M. Borgwardt et al. report on a time
-resolved 2PPE study of the
carrier dynamics in a pristine Cu2O single crystal and after deposition of a small amount
of Pt. From
their data they conclude on Cu bulk vacancy states acting as decay channels for charge carriers
photoexcited into the conduction band and being responsible for the reduced photovoltages observed
for this material system. The experiments have b
een performed thoroughly and the data were
analyzed carefully. The main text itself misses sometimes relevant details, which are partly ‘hidden’ in
the supplementary information. The interpretation stays at qualitative level but is convincing at least
rega
rding the major conclusions.
As the key outcome of the study the authors conclude that the ‘modest photovoltages that can be
obtained from these Cu2O samples are not primarily due to recombination induced by surface states,
but primarily from bulk recomb
ination processes into bulk defect states.’ This seems to me an
interesting finding, but I am not really convinced that this is enough to justify publication in Nature
Communications. I have to admit, however, that experts in the field of solar
-driven wate
r splitting
processes may come to a different conclusion.
In case the manuscript is considered for publication in Nature communications, the following points
should be answered by the authors and considered in a revised version:
(1) The authors state i
n their conclusions that their ‘observations indicate that tr
-2PPE can provide a
powerful experimental method to simultaneously unravel the energetics and dynamics of photoexcited
electrons as they arrive at surfaces or internal interfaces of semiconductor
s. Tr
-2PPE thus provides
explicit insight into the origin and mechanism of photovoltage losses in photovoltaic or photocatalytic
materials and can offer crucial insights at an early stage of material development.’ This is definitely
correct, but in my opin
ion not actually new or surpising. Tr
-2PPE experiments have been performed
over more than 30 years and a multitude of experiments on quite different material systems from
many groups were published in the past, also including experiments addressing relevan
t processes in
photovoltaic systems. I actually was very surprised to see that the authors refer only to a very limited
number of tr2PPE studies (in the supplemental information, only). Even more, these citations are in
my opinion not really of relevance f
or the key outcome of the study. I strongly recommend checking
for instance tr2PPE studies by the Zhu group on inorganic
-organic perovskites. As far as I know also
the Perfetti group and the Fauster group (D. Niesner) have published some work on this stuff
. In a
wider context also work by the Wolf group on charge carrier solvation in ice overlayers and work by
the Harris group and the Höfer group on charge
-transfer processes across interfaces could be of
interest. Finally, I remember some work by the Aeschlimann group on surface charge accumulation in
GaAs due to carrier diffusion from the bulk. I am pretty sure that at least part of this work could be
helpful in the interpretation of the submitted study.
(2) It becomes not clear what amount of Pt was finally deposited onto the reconstructed Cu2O surface
that was investigated by tr2PPE. Is the surface fully covered with Pt or are there open Cu2O areas in
between the Pt islands that form upon deposition as verified by AFM and SEM. In the main text, the
autho
rs refer to a non-
conformal structure of the Pt adlayer, which may hint to open CuO2 areas.
However, in the supplementary information (section 6) the authors write: ‘X
-ray photoelectron
spectroscopy (XPS) data for Pt films of increasing thickness revealed
a non
-conformal coverage for the
thinnest films (0.5 nm) while a continuous Pt layer was obtained for longer deposition times (Fig.
S8b). Based on these results, 1 nm Pt films were used for all of experiments discussed herein.’ What
does this finally mean
for the samples discussed in the main text? The information is in my opinion
important as the presence of open Cu2O areas could be of relevance for the interpretation of the
data.
(3) The authors directly conclude from the bulk
-sensitive photoluminescence measurements onto the
origin of the dominating signal recorded using the surface sensitive 2PPE method. I would suggest
being more conservative with such statements as it is well known that the surface electronic structure
can differ significantly from t
he bulk electronic structure as the bulk termination can results in a
surface reconstruction and also the formation of quite different types of defects.
(4) In line 112 the authors state that the Cu2O
-derived features on the Pt covered sample primarily
originate from a combination of directly emitted electrons from the Cu2O substrate (I assume this
refers to Cu2O areas not covered by Pt) in conjunction with electrons probed through the Pt top layer.
I have to admit that I do not completely understand how
the authors come to this conclusion from
their discussion on the electron mean free path in the same paragraph. Particularly I am wondering
how the authors can exclude that the Cu2O
-derived features on the Pt covered sample exclusively
arise from directly
emitted electrons from the Cu2O substrate?
(5) The conclusion that the characteristic signal rise in the band
-gap feature arises from carrier
diffusion to the surface in the defect band seems to me very reasonable. I was surprised that the
observed times
cales were not related to the time
-resolved PL data on the (much longer) defect
depopulation times. Furthermore, the authors observe and mention in the main text a two
-component
signal rise with a pronounced fluence dependence of the fast rise
-time. What c
an we learn from these
observations? This point is not discussed at all in the rest of the manuscript.
(6) What mechanism channels excited electrons at the surface into sites associated with a high
dislocation density (line 214)? Shouldn’t such type of p
rocess result in the formation of an in plane
surface potential gradient which at some point will stop the channeling process.
(7) The author mention thermally assisted hopping within the VCu defect band. What energies are we
talking about? Here, experim
ents performed at different temperatures would be nice and would
strongly support the interpretation in terms of bulk to surface diffusion.
(8) In the abstract the authors claim that their data suggest that the Pt adlayer reduced the surface
defect state
s. Later in the manuscript (line 147) they write that ‘...either the defect states disappeared
after deposition of Pt, or that the Pt provided an alternative pathway for the photoexcited electrons
that prevented filling of defect states.’ From the following
discussion in the manuscript I learned that
none of these statements is finally correct: The defects did not disappear but become hidden
underneath the Pt
-island. These defect states become still populated (at least as implied by Fig. 4(b)).
However, in th
e presence of the Pt overlayer they get much faster depopulated in comparison to the
reconstructed surface due to coupling to platinum electronic states.
(9) The authors state that the reconstructed surface has a non
-uniform defect distribution (line 143
)
(and conclude from this that the ‘long
-lived signal can consistently be ascribed to sites associated with
a high dislocation density.’) I missed somehow, where this information is coming from. Maybe this
point is somewhere hidden in the supplementary inf
ormation. In any case at least a reference must be
given.
(10) It seems to be very important that a reconstructed Cu2O surface instead of a non
-reconstructed
surface is used in the experiment as this point is mentioned several times throughout the manusc
ript.
I am not familiar with the details of the water
-splitting capabilities of Cu2O and I think it would be
helpful for the non-
expert to spend a few words in the main text or the supplementary information on
this point.
Reviewer #3 (Remarks to the Au
thor):
This work measures the relaxation dynamics of photoexcited electrons in Cu2O pure and with a Pt
cover. The performed 2
-photo
-photoelectron measurements are well designed and executed. The
manuscript describes the undertaken experiments and analyse
s clearly. I also summarizes the results
clearly. The difference between pure and Pt
-covered material is strikingly different and I think that the
authors are right with their analysis. The defect sites on the surface become slowly filled with bulk
electro
ns. it is quite fascinating to see this in real time.
I only have one very minor comment: in the abstract the authors write about energetically
"unfavorable" states. It is not upfront clear what that should mean. The authors may want to chose a
more clari
fying description.
Otherwise, publication of this work is recommended.
Clemens Burda
Point
-by-point response to the referees’ comments:
Reviewer #1 (Remarks to the Author):
The manuscript here presents time-
resolved two-
photon photoemission studies of Cu2O
photoelectrons. The manuscript is interesting. However, I was confused by the overall
message. On the one hand the authors attempt to motivate the work to demonstrate
that
the technique is useful for studying the surface carrier dynamics of photoelectrons.
In that regard they determine the effect of the carrier dynamics between a bare Cu2O
surface and one with a Pt overlayer catalyst. On the other hand the authors discuss the
important of Cu2O. An present the work to understand the carrier dynamics at the
surface of this important semiconductor. Since there are significant differences in the
response for the bare and Pt overlay sample this is something to learn about how thes
e
photoelectrons are working. But I didn't get a clean message as to what the Pt is doing.
Thus, the biggest issue with this manuscript is that there is no insight for how one might
develop a Cu2O photoelectrode that is not impacted by the defects
We thank the reviewer for carefully reading and evaluating the manuscript
. We agree
with the reviewer’s opinion in that the manuscript needs a clearer focus on its central
outcome. We have adapted the manuscript accordingly
(see changes to abstract below
and in response to the comments hereinafter)
and underline in more detail the
importance
of photovoltage losses in
Cu
2
O electrodes originating from deep defect
state
s. However, we note that the manuscript was not intended to
investigate
specific
aspects of material engineering, but rather unraveling limit
ing
factors and providing
deeper insight into this loss mechanism. We therefore believe that our results will be
particularly helpful in future Cu
2
O material development broadening the emphasis of
current research from interface engineering towards a more profound and controlled
synthesis of high-
quality bulk material.
Associated changes to the abstract:
“By referencing ultrafast energy resolved surface sensitive spectroscopy
to bulk data we
identify the full bulk to surface transport dynamics for excited electrons rapidly localized
within an intrinsic deep continuous defect band ranging from the whole crystal volume to
the surface. No evidence of bulk electrons reaching the s
urface at the conduction band
level is found resulting into a substantial loss of their energy through ultrafast trapping.“
Furthermore, the measurements are done in ultrahigh vacuum -
which seems to lesson
the impact of the measurements and the measurement technique somewhat. While the
authors do make an attempt to justify why these measurements are still valid.
As cross check, al
l tr
-2PPE, PL and tr
-PL measurements were also conducted directly
after crystal cleavage without further surface treatment (as
-received) resulting into
qualitatively similar signals compared to the reconstructed samples
demonstrating the
general insensitivity to specific surface preparation methods. However, signal intensities
and background contributions with tr
-2PPE were dif
ferent and varied depending on
respective sample and chosen spot on the sample surface. Such variations are not
surprising due to the high surface sensitivity, so that we decided to include only the
results from the well
-defined and highly reproducible sur
face that were not hampered by
these difficulties.
On the other hand, our results suggest that even in contact with a
solvent the main loss channel is
ascribed to the decay of charge carriers into deep bulk
defect states and thus would
not alter the final conclusions
.
Associated changes to the main text:
Page 4:
“Artificially synthesized Cu
2
O crystals were chosen and an extensive surface
reconstruction procedure was employed that ensures high reproducibility accounting for
the surface sensitivity of the tr
-2PPE technique. Investigations conducted at crystals
without any surface treatment (as
-received) yielded qualitatively similar results and are
not shown.”
The discussion and presentation of Fig. 2 is confusing. What determines the zero
energy scale here? Is the Fermi energy not determined by the system for these
measurements? Why is the CB
at a lower energy than the VBM and shouldn't the Fermi
energy be between the CB and VBM. I see that the DB is at a lower energy than the CB
which would make sense. But it's unclear from this plot how the energy diagrams of Fig.
4 are constructed. Maybe this is well known for the tr
-2PPE community (but this is
probably small).
We agree that the discussion of the used energy scale in Fig.
2 needs clarification.
The
photoemission spectra are presented in dependence on the directly
measured electron
kinetic energy. It is important to note, that t
he spectra can be understood
as a
superposition of two different ionization processes –
one involving two probe photons
(equally to solely applying the probe beam; black background curve) and, secondly, a
mixed transition involving one pump and one probe photon. The transformation of the
kinetic energy scale into
a single common binding energy scale is therefore not
achievable since the different combinations of pump and probe photons result into
different binding energy scales. For the same reason the VBM originating from
two
probe photons occurs at higher absolute value on the electron kinetic energy scale than
the CB and DB originating from a mixed transition involving one pump und one probe
photon. For further clarification we refer to section 1 in the original version of the SI.
Associated changes to the main
text:
Page
6:
“We point out that the recorded spectra consist of a combination from two different
ionization processes involving different photoionization energies and originate from
initially occupied and unoccupied states. For
example,
the stationary background (Fig.
2, black line) arises from a transition comprising two probe photons, whereas the actual
2PPE signal is composed of a mixed transition involving one pump and one probe
photon. Hence, the definition of a single binding energy scale representing all signals
is
not feasible and the directly measured electron kinetic energy scale was chosen.
However, each spectral component can be referenced to the systems internal Fermi
energy with E
F
=0 and the calibration of the binding energy scale was performed by
measurement of the Fermi edge of a Cu reference sample (see Supplementary section
1).
“
We further adapted Fig.
2 to avoid confusion and to simplify the general presentation of
the measured results. We deleted the indicator for the VBM and E
F
.
Could the authors change the excitation wavelength. This would help determine if the
rise time in Fig. 3 is dues to diffusion of carriers within the bulk to the surface. How do
the authors determine the CB electrons first relax to DB and then transport to the
surface . Coul
dn't CB diffuse to the surface and then relax to the DB? The authors rule
this out due to the fast decay of surface CB to the DB and then the subsequent arrival of
DB electron density. However, if the relaxation of surface CB electron to the DB is fast
then one might expect a similar behavior. As CB electrons arrive at the surface (in
coherently) they quickly relax to the DB.
The
used
frequency conversion of the Ti:Sapphire laser system via non-
collinear optical
parametric amplifiers (NOPA)
is highly limited
in the spectral range that can be achieved
with
reasonable
pulse energy
output
. The limited range would
resul
t only
into
small
variations in absorptions dept
h that might be difficult to compare
on a
qualitative level.
We further
point
to
the fact that
the
dynamics of
2PPE signal is particularly sensitive to
the charge carrier dynamics in the space-
charge region that might considerably deviate
from the bulk
. A fluence dependent rise of the 2PPE signal was found that could
potentially be explained by dynamic changes in surface band alignment and the space
charge
region.
However, due to this number of potential limitations great care are must
be taken to draw clear conclusions and we decided to not address this topic in more
detail without further investigation.
We agree with the reviewer’s opinion that solely the 2PPE
results
do not provide an
unambiguous picture of the carrier recombination. For this reason, we complemented
the surfaces sensitive method with the bulk information we gained by
photoluminescence measurements.
As stationary PL reveals the main recombination
channel occurs via defect
-mediated radiative recombination originat
ing
from V
Cu
vacancies. We could further show
that any surface modification via Pt
-deposition does
not change
the signal intensity and
the ratio between the free exciton signal and defect
PL
. In contrast
the Pt
-deposition
modified the surface lifetimes of both species,
conduction band and defect band electrons. From both findings
we draw the
conclusion
that the characteristic signal rise in the band-
gap feature arises from carrier diffusion to
the surface in the defect band.
In response to this point we shifted and extended the
discussion about the PL findings and their
correlation with the 2PPE results
.
Associated changes to the main
text:
Page
8:
“In order to gain additional information about the origin of the filling of these defect
levels located within the band gap photoluminescence (PL) measurements were
conducted exhibiting primarily bulk sensitivity and, therefore, represent a well-
suited
complementary method to the surface sensitive 2PPE technique. The experiments were
conducted at the same stages of sample preparation (reconstructed, Pt
-deposited)
without breaking UHV conditions and applying identical pump laser pulse conditions as
utilized in the 2PPE measurements (Supplementary section 5). Both stationary and
time
-resolved photoluminescence measurements (PL) strongly suggest that Cu
vacancies act as the dominant defect type in the investigated material, and the
photoemission energy is i
n close accord with the energetic position of the defect state in
the band gap. The different surface treatments did not alter the PL signals, suggesting a
relatively high concentration of Cu vacancies in the bulk and that the surfaces
contribution as well
as influence to the charge recombination is negligible. We point out
that the found energetic match does not allow to draw direct conclusions about the type
and density of defects at the surfaces. Therefore, additional surface-
sensitive
experiments (LEED,
XPS) were performed to obtain information about the
stoichiometry, structure, and energetics (band bending, band alignment) of the material
(see Supplementary section 6).
The main conclusion is that photoelectrons relax to bulk defect states prior to reaching
surfaces and thus loose their potential very fast interior to the material. Does this mean
that Cu
2
O cannot be fixed? If the defect are due to Cu vacancies is there a way to
decrease their number and see how that impacts the results.
The outcome of
the study highlights the importance of bulk recombination processes in
Cu
2
O. Hence, in addition to known issues in Cu
2
O, such as a mismatch between the
electronic band alignment
at junctions
and defect states at interfaces, bulk defect
s in the
Cu
2
O may
substantially limit the obtainable voltage in Cu
2
O devices.
High quality Cu
2
O
sheets or nano assemblies are therefore required to obtain efficient photocathodes. In
recent studies (
Refs.
46,
49,
50)
there have been significantly
increased conversion
efficiencies reported, however,
without a clear correlation of these improvements with
specific bulk properties.
The manuscript therefore underlines the importance of material
development in C
u
2
O electrodes.
We discuss these
potential approaches in the main
text
but further analysis
of their suitability and impact on
the presented results is beyond
the scope of the study.
Overall. I like the manuscript -
but not sure that this should be published
in Nature
communication. The impact of the manuscr
ipt seems rather low. Maybe the
authors should use the results to improve the operation of the photoelectrons
.
We thank the reviewer for the comments and critical assessment of the work. We have
addressed the concerns in the modifications shown above.
Reviewer #2 (Remarks to the Author):
In the manuscript ‘Femtosecond time-
resolved two-
photon photoemission studies of
ultrafast carrier relaxation in Cu2O photoelectrodes’ M. Borgwardt et al. report on a
time
-resolved 2PPE study of the carrier dynamics in
a pristine Cu2O single crystal and
after deposition of a small amount of Pt. From their data they conclude on Cu bulk
vacancy states acting as decay channels for charge carriers photoexcited into the
conduction band and being responsible for the reduced photovoltages observed for this
material system. The experiments have been performed thoroughly and the data were
analyzed carefully. The main text itself misses sometimes relevant details, which are
partly ‘hidden’ in the supplementary information. The int
erpretation stays at qualitative
level but is convincing at least regarding the major conclusions.
As the key outcome of the study the authors conclude that the ‘modest photovoltages
that can be obtained from these Cu
2
O samples are not primarily due to recombination
induced by surface states, but primarily from bulk recombination processes into bulk
defect states.’ This seems to me an interesting finding, but I am not really convinced
that this is enough to justify publication in
Nature
Communications. I have to admit,
however, that experts in the field of solar
-driven water splitting processes may come to
a different conclusion.
In case the manuscript is considered for publication in
Nature
communications, the
following points should be answered by the authors and considered in a revised version:
We thank the reviewer for carefully reading and evaluating the manuscript. We agree
with the reviewer’s opinion in that the manuscript needs a clearer focus on its central
outcome and highlighting of its impact on future Cu
2
O material development. We have
adapted the manuscript accordingly (see changes to abstract and in response to the
comments hereinafter)
. Below is our point by point response to
all comments.
(1) The authors state in their conclusions that their ‘observations indicate that tr
-2PPE
can provide a powerful experimental method to simultaneously unravel the energetics
and dynamics of photoexcited electrons as they arrive at surfaces or internal interfaces
of semiconductors. Tr
-2PPE thus provides explicit insight into the origin and mechanism
of photovoltage losses in photovoltaic or photocatalytic materials and can offer crucial
insights at an early stage of material development.’ This is definitely correct, but in my
opinion not actually new or s
urpr
ising. Tr
-2PPE experiments have been performed over
more than 30 years and a multitude of experiments on quite different material systems
from many groups were published in the past, also including experiments addressing
relevant processes in photovolt
aic systems. I actually was very surprised to see that the
authors refer only to a very limited number of tr2PPE studies (in the supplemental
information, only). Even more,
these citations are in my opinion not really of relevance
for the key outcome of the study. I strongly recommend checking for instance tr2PPE
studies by the Zhu group on inorganic
-organic perovskites. As far as I know also the
Perfetti group and the Fauster group (D. Niesner) have published some work on this
stuff. In a wider context als
o work by the Wolf group on charge carrier solvation in ice
overlayers and work by the Harris group and the Höfer group on charge-
transfer
processes across interfaces could be of interest. Finally, I remember some work by the
Aeschlimann group on surface c
harge accumulation in GaAs due to carrier diffusion
from the bulk. I am pretty sure that at least part of this work could be helpful in the
interpretation of the submitted study.
We thank the reviewer for pointing us to these well
-known and highly relevant studies.
We have included the suggested references in the manuscript.
Associated changes to the main
text:
Page 3:
“2PPE spectroscopy has been previously successfully applied
to study electron
dynamics at surfaces and charge transfer processes across interfaces of metals and
semiconductor model systems
22-
24
. Only recently has its capabilities been utilized to
study
processes in
emerging photovoltaic materials systems such as hot electron
relaxation dynamics in hybrid metal
-organic perovskite semiconductors
25-
27
. We have
now extended these efforts
to the group of metal oxide semiconductors by including
Cu
2
O – one of the most prom
ising
metal oxide candidates for solar water splitting
28
.”
Page 10:
“Similar findings have been reported for tr
-2PPE studies on GaAs (100) surfaces where
a rising electron population on a ps timescale has been assigned to scattering of
electrons into low-
energy states in the band-
bending region
34
.”
(
2) It becomes not clear what amount of Pt was finally deposited onto the reconstructed
Cu2O surface that was investigated by tr2PPE. Is the surface fully covered with Pt or
are there open Cu2O areas in between the Pt islands that form upon deposition as
verified by AFM and SEM. In the main text, the authors refer to a non-
conformal
structure of the Pt adlayer, which may hint to open CuO2 areas. However, in the
supplementary information (section 6) the aut
hors write: ‘X
-ray photoelectron
spectroscopy (XPS) data for Pt films of increasing thickness revealed a non-
conformal
coverage for the thinnest films (0.5 nm) while a continuous Pt layer was obtained for
longer deposition times (Fig. S8b). Based on these results, 1 nm Pt films were used for
all of experiments discussed herein.’ What does this finally mean for the samples
discussed in the main text? The information is in my opinion important as the presence
of open Cu2O areas could be of relevance for the i
nterpretation of
the data.
We thank the reviewer for pointing this out.
Our XPS data and its theoretical modeling
indicated a non-
conformal growth for Pt deposition via UHV evaporation of films with
thicknesses < 1nm. For films with theoretical thickness equ
ivalent to 1nm and larger –
as
determined by monitoring the rate of deposition using a quartz crystal microbalance
(QCM) in UHV
– the fitting procedure of the XP
S data revealed a decrease in the size
effects (line shape asymmetry, shift of the Pt 4f binding energy, linewidth broadening)
associated with non-
conformal growth. However, AFM and SEM verified the non-
conformal growth of Pt for the 1
nm thick layers as used in the samples for tr
-2PPE
measurements.
Associated changes to the S
upplementary text
:
Page 11:
“To summarize, X
-ray photoelectron spectroscopy (XPS) data and its theoretical
modeling indicated a non-
conformal growth for Pt deposition via UHV evaporation of
films with thicknesses < 1 nm. For films with theoretical thickness equivalent to 1 nm
and larger –
as determined by monitoring the rate of deposition using a quartz crystal
microbalance (QCM) in UHV –
the fitting procedure of the XPS data revealed a
decrease in the size effects (line shape asymmetry, shift of the Pt 4f binding energy,
linewidth broadening) associated with non-
conformal growth (Fig. S8b). Based on these
results, Pt films with QCM thickness equivalent to 1 nm were used for all of the
experiments discussed herein.”
(3) The authors directly conclude from the bulk
-sensitive p
hotoluminescence
measurements onto the origin of the dominating signal recorded using the surface
sensitive 2PPE method. I would suggest being more conservative with such statements
as it is well known that the surface electronic structure can differ signi
ficantly from the
bulk electronic structure as the bulk termination can results in a surface reconstruction
and also the formation of quite different types of defects.
We thank the reviewer
for this valuable comment. We agree
that the current version of
the manuscript reads as it was our intension to draw conclusions about the actual type
and density of the surface defects from photoluminescence measurements. We point
out
that
their
energetic position
in the bulk (measured by PL) and at the surface (tr
-
2PPE)
match and conclude that “
surface accumulation of the electrons occurs at levels
isoenergetic with the bulk trap band“
. We are aware of the fact that the surface
reconstruction and adsorption of
adlayers
or atoms can decisively alter the surface
properties compared to the bulk. We clarified the text regarding
this point.
Associated changes to the main
text:
Page 8:
“In order to gain additional information about the origin of the filling of these defect
levels located within the band gap photoluminescence (PL) measurements were
conducted exhibiting primarily bulk sensitivity and, therefore, represent a well-
suited
complementary method to the surface sensitive 2PPE technique. The experiments were
conducted at the same stages of sample preparation (reconstructed, Pt
-deposited)
without breaking UHV conditions and applying identical pump laser pulse conditions as
utilized in the 2PPE measurements (Supplementary section 5). Both stationary and
time
-resolved photoluminescence measurements (PL) strongly suggest that Cu
vacancies act as the dominant defect type in the investigated material, and the
photoemission energy is i
n close accord with the energetic position of the defect state in
the band gap. The different surface treatments did not alter the PL signals, suggesting a
relatively high concentration of Cu vacancies in the bulk and that the surfaces
contribution as well
as influence to the charge recombination is negligible. We point out
that the found energetic match does not allow to draw direct conclusions about the type
and density of defects at the surfaces. Therefore, additional surface-
sensitive
experiments (LEED,
XPS) were performed to obtain information about the
stoichiometry, structure, and energetics (band bending, band alignment) of the material
(see Supplementary section 6).
“
(4) In line 112 the authors state that the Cu
2
O-derived features on the Pt covered
sample primarily originate from a combination of directly emitted electrons from the
Cu
2
O substrate (I assume this refers to Cu
2
O areas not covered by Pt) in conjunction
with electrons probed through the Pt top layer. I have to admit that I do not complet
ely
understand how the authors come to this conclusion from their discussion on the
electron mean free path in the same paragraph. Particularly I am wondering how the
authors can exclude that the Cu
2
O-derived features on the Pt covered sample
exclusively a
rise from directly emitted electrons from the Cu
2
O substrate?
We agree with the reviewer that this point needs to be clarified. Given that the diameter
of the
probe
laser spot was on the order of 50μm, it implies that the
probed
area of the
Pt covered sample contained both Pt islands and uncovered Cu
2
O areas. Hence, we
obtained a
n averaged
2PPE signal containing both contributions. In addition, since the
probing depth of the 2PPE measurement exceeds the thickness of the non-
conformal
Pt
layer, the signal contains both electrons probed through the Pt and
as well as
emitted
from the uncovered Cu
2
O substrate.
We have modified the corresponding section in the
manuscript
to
extend and clarify this point.
Associated changes to the main text:
Page
7:
“For low
photoelectron kinetic energies, the inelastic mean free paths are relatively
large, exceeding tens of nanometers
29
. However, because elastic electron-
acoustic
phonon scattering overcomes the energy loss scattering events, the photoelectron
escape depth is substantially reduced for such low kinetic energies
30,31
. Hence, the
thickness region sampled by 2PPE is typically on the order of a few
nanometer
s
25,32,33
.
In addition to the above mentioned surface sensitivity of 2PPE, it is important to note
that
due
to the non-
conformal coverage of the Pt adlayer and the surface area
accessible by the probe pulse with diameter of ~50 μm both Pt covered and uncovered
Cu
2
O areas were simultaneously
sampled.
We therefore expect that the Cu
2
O-derived
features on the Pt covered sample primarily originate from a combination of directly
emitted electrons from the Cu
2
O substrate in conjunction with electrons probed through
the Pt top layer.
”
(
5) The conclusion that the characteristic signal ri
se in the band-
gap feature arises from
carrier diffusion to the surface in the defect band seems to me very reasonable. I was
surprised that the observed timescales were not related to the time-
resolved PL data on
the (much longer) defect depopulation times. Furthermore, the authors observe and
mention in the main text a two-
component signal rise with a pronounced fluence
dependence of the fast rise-
time. What can we learn from these observations? This
point is not discussed at all in the rest of the manusc
ript.
We agree that a direct
correlation
of the time scales
between
the
two methods seems
an attractive opportunity to learn more about the physics involved. However, we believe
that
the very different time resolutions
, recorded time windows
and information depth of
both measurements
complicate their
interpretation
and comparison
with regard to
dynamics
. Although we see a decrease in
the tr
-2PPE defect signal within the
available
2ns time window
, that might be assigned to a (bulk)
defect
depopulation,
it is not clear
how the signal further evolves in time and how fast it reaches zero level or whether it
converges into a
long-
lived offset.
On the other hand, the defect related PL signal
reaches
its maximum at about 10ns. We further point to
the fact that the 2PPE signal is
particularly
sensitive to the charge carrier dynamics in the space-
charge region that
might considerably deviate from the bulk
, especially with regard of the initial diffusion
process of charge carriers into the surface tr
ap states
. Dynamic
changes in surface
band alignment and the space charge region
induced by the optical
pump
could also
explain the fluence dependent rise of the 2PPE signal.
Due to this
number of potential
limitations
great care are must be taken to draw
clear conclusions and
we decided to
not address this topic in m
ore detail
without further investigation.
(6) What mechanism channels excited electrons at the surface into sites associated with
a high dislocation density (line 214)? Shouldn’t such type
of process result in the
formation of an in plane surface potential gradient which at some point will stop the
channeling process.
The diffusion process towards
the surface is driven by the downward band bending.
Due to the non-
uniform defect distribution
at the reconstructed surface charge carrier
accumulation will be statistically enhanced at sites associated with a higher dislocation
density. We agree that the current version of the manuscript seems to suggest a
specific effect being the origin for the charge separation.
We also agree that a non-
uniform charge distribution at the surface would ultimately lead to potential gradients
that will counteract the diffusion process. However, as pointed out answering question 4
the limited spatial resolution in t
he tr
-2PPE as well as XPS measurements do not allow
to draw any conclusion on this level of detail.
Associated changes to the main
text:
Page 12:
“The long rise time of the integrated 2PPE signal, which extended into the ns time
domain, is consistent with
these states being filled with bulk electrons that drift or diffuse
toward the surface with the electrons then primarily accumulated at sites
with a high
dislocation density.
“
(7) The author mention thermally assisted hopping within the VCu defect band.
What
energies are we talking about? Here, experiments performed at different temperatures
would be nice and would strongly support the interpretation in terms of bulk to surface
diffusion.
We thank the reviewer for this comment and have added a reference that emphasizes
low electron mobility in Cu
2
O using terahertz spectroscopy implying non-
conduction
band and activated transport. Considering the energetic position of the defect band
electrons cannot repopulate the conduction band and must hop between defect levels.
To
go into full details of the bulk transport mechanism temperature dependent
experiments would be nice but would go beyond the scope of this manuscript.
Associated changes to the main
text:
Page 12:
“This fit well to the low electron
mobility estimated by Paracchino et al.
39
between
2.7−6.3 cm
2
V
−1
s
−1
in electrodeposited Cu
2
O using time-
resolved terahertz
spectroscopy. Considering the position of the V
Cu
defect band located deep below the
conduction band minimum bulk electron transport
to the surface would consistently
occur via thermally assisted hopping within the defect band, in accord with the moderate
values for minority
-carrier mobility and diffusion length from other reports for Cu
2
O
11,40–
42
.”
(8) In the abstract the authors claim that their data suggest that the Pt adlayer reduced
the surface defect states. Later in the manuscript (line 147) they write that ‘...either the
defect states disappeared after deposition of Pt, or that the Pt provided an alternative
pathway for the photoexcited electrons that prevented filling of defect states.’ From the
following discussion in the manuscript I learned that none of these statements is finally
correct: The defects did not disappear but become hidden under
neath the Pt
-island.
These defect states become still populated (at least as implied by Fig. 4(b)). However,
in the presence of the Pt overlayer they get much faster depopulated in comparison to
the reconstructed surface due to coupling to platinum electronic states.
We agree that the abstract needs clarification regarding
the final conclusion made in the
manuscript.
The
preliminary interpretation
in the result section
(Page 9)
considering
solely the tr
-2PPE
signals
state
s two different possibilities, that are either a reduction of
available surface trap state or an alternative relaxation channel that prevent
s the
accumulation of charge in the
defect state. After incorporating of all other available
information,
we concluded
that the lat
ter is the most probable explanation for the
missing signal rise.
We adapted the manuscript at both places in the text.
Associated changes to the abstract:
“The data also suggest that the Pt adlayer suppresses the slow accumulation of
electrons into the surface defect states and that the Pt is capable of
mediat
ing
the
charge transfer at the semiconductor/metal interface.
”
Associated changes to the main text:
Page 9:
“This behavior suggests that either the defect states disappeared after deposition of Pt,
or that the Pt provided an alternative pathway for the photoexcited electrons that
prevented accumulation of charge carriers in the defect states.
”
(9) The authors state that the reconstructed surface has a non-
uniform defect
distribution (line 143) (and conclude from this that the ‘long-
lived signal can consistently
be ascribed to sites associated with a high dislocation density.’) I missed somehow,
where this information is coming from. Maybe this point is somewhere hidden in the
supplementary information. In any case at least a reference must be given.
Results
obtained by
AFM and SEM
indicate a non-
uniform defect distribution (both
presented in the supplementary information)
. We agree that there is no reference in the
main text to this important information. We therefore adapt
ed
the main text accordingly.
Associated changes to the main
text:
Page
9:
"Surface analysis of
the reconstructed and Pt covered Cu
2
O sample by means
of atomic
-force and scanning electron
microscopy (AFM, SEM; see Supplementary
section
8) revealed a non-
uniform defect distribution, hence the long-
lived
signal
can
consistently be ascribed to sites associated with a high dislocation density."
(10) It seems to be very important that a reconstructed Cu2O surface instead of a non-
reconstructed surface is used in the experiment as this point is mentioned several times
throughout the manuscript. I am not familiar with the details of the water
-splitting
capabilities of Cu2O and I think it would be helpful for the non-
expert to spend a few
words in the main text or the supplementary information on this point.
We agree with the reviewer that this point needs clarification: a
ll tr
-2PPE, PL and tr
-PL
measurements were also conducted
directly
after crystal cleavage without further
surface treatment (as
-received)
resulting into qualitatively similar signals compared to
the reconstructed samples. However, signal intensities and background contributions
with tr
-2PPE
were different
and varied depending on respective sample and chosen
spot on the sample surface.
Such variations are not surprising due to the high surface
sensitivity,
so that we decided
to include only the results from the
well
-defined
and
highly reproducible surface
that were not hampered by these
difficulties.
The crystal
orientation and reconstruction procedure were
not specifically chosen with regard to
water
-splitting aspects of Cu
2
O and we will add this information
to the manuscript
.
Associated changes to the main
text:
Page 4:
“Artificially synthesized Cu
2
O crystals were chosen and an extensive
surface
reconstruction procedure was employed that ensures high reproducibility accounting
for
the surface sensitivity of
the tr
-2PPE technique.
Investigations conducted at crystals
without any surface treatment
(as-received)
yielded
qualitatively similar results
and are
not shown.
”
Reviewer #3 (Remarks to the Author):
This work measures the relaxation dynamics of photoexcited electrons in Cu2O pure
and with a Pt cover. The performed 2-
photo-
photoelectron measurements are well
designed and executed. The manuscript describes the undertaken experiments and
analyses clearly. I also summarizes the results clearly. The difference between pure
and Pt
-covered material is strikingly different and I think that the authors are right with
their analysis. The defect sites on the surface become slowly filled with bulk electrons. it
is quite fascinating to see this in real time.
I only have one very minor comment: in the abstract the authors write about
energetically "unfavorable" states. It is not upfront clear what that should mean. The
authors may want to chose a more clarifying description.
We thank the reviewer for studying our manuscript and for providing positive feedback
and comments. We
have rewritten the abstract and replaced the term “unfavorable”
accordingly.
REVIEWERS' COMMENTS:
Reviewer #1 (Remarks to the Author):
The authors have adequately addressed this reviewers comments. I can recommend publication.
Reviewer #2 (Remarks to the Author):
I carefully read the reply of the authors to the comments of all three reviewers. TIn my opinion, the
authors addressed all points raised by the reviewers in a satisfactory manner. I particularly
acknowledge that the revisions made by the author clarified
the actual focus of the study. Overall I
can agree with a publication of the manuscript. However, as already mentioned in my first review, I
am not an expert in the field of solar
-driven water splitting and have therefore difficulties to judge
whether the
novelty or relevance of the outcome of the study justifies publication in Nature
Communications.