Engineering Sensitized Photon Upconversion Efficiency via Nanocrystal Wavefunction and Molecular Geometry

Article abstract of DOI:10.1002/anie.202009066

Triplet energy transfer from inorganic nanocrystals to molecular acceptors has attracted strong attention for high-efficiency photon upconversion. Here we study this problem using CsPbBr₃ and CdSe nanocrystals as triplet donors and carboxylated anthracene isomers as acceptors. We find that the position of the carboxyl anchoring group on the molecule dictates the donor-acceptor coupling to be either through-bond or through-space, while the relative strength of the two coupling pathways is controlled by the wavefunction leakage of nanocrystals that can be quantitatively tuned by nanocrystal sizes or shell thicknesses. By simultaneously engineering molecular geometry and nanocrystal wavefunction, energy transfer and photon upconversion efficiencies of a nanocrystal/molecule system can be improved by orders of magnitude.

Full text of DOI:10.1002/anie.202009066

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Engineering Sensitized Photon Upconversion Efficiency via
Nanocrystal Wavefunction and Molecular Geometry
Authors: Shan He and Kaifeng Wu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202009066
Link to VoR: https://doi.org/10.1002/anie.202009066

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
Engineering Sensitized Photon Upconversion Efficiency via
Nanocrystal Wavefunction and Molecular Geometry
Shan He, Runchen Lai, Qike Jiang, Yaoyao Han, Xiao Luo, Yuyang Tian, Xue Liu and Kaifeng Wu*
[*]
Dr. S. He, Dr. R. Lai, Y. Han, Dr. X. Luo, Y. Tian, X. Liu, Prof. Dr. K. Wu
State Key Laboratory of Molecular Reaction Dynamics, Dynamics Research Center for Energy and Environmental Materials, Dalian Institute of Chemical
Physics, Chinese Academy of Sciences, Dalian, Liaoning 116023, China
E-mail: kwu@dicp.ac.cn
Q. Jiang
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, Liaoning 116023, China
Y. Han
University of the Chinese Academy of Sciences, Beijing 100049, China
Supporting information for this article is given via a link at the end of the document.
Abstract: Triplet energy transfer from inorganic nanocrystals to
molecular acceptors has attracted strong attention as a promising
approach to high-efficiency photon upconversion that is important for
energy-related applications. Because energy transfer is a donor-
acceptor interfacial process, it is essential to simultaneously
engineer the nanocrystal and molecular parameters in order to
uncover unified physical principles that are transferable among
different systems. Here we study this problem using CsPbBr3 and
CdSe nanocrystals as triplet donors and carboxylated anthracene
isomers as acceptors. We find that the position of the carboxyl
anchoring group on the molecule dictates the donor-acceptor
coupling to be either “through-bond” or “through-space”, while the
relative strengths of the two coupling pathways is controlled by the
“wavefunction-leakage” of nanocrystals that can be quantitatively
tuned by nanocrystal sizes or shell thicknesses. By simultaneously
engineering molecular geometry and nanocrystal wavefunction,
design also has an important impact on TET. The efficiencies of
TET from CdSe NCs to anthracene acceptors functionalized with
carboxylate, dithiocarbamate or diphosphate groups were found
to be very different.[8] Interestingly, the substitution position of
the functional group on the molecule is important as well. For
example, the photon upconversion efficiencies of CdSe NCs
with different carboxylated anthracene isomers differed by more
than 10-fold.[8a] These effects were often tentatively attributed to
electronic coupling strengths (orbital overlaps) between NCs and
molecular acceptors.[8a] However, the exact mechanism by
which the electronic coupling differs remains unclear.
Considering that TET occurs across the donor-acceptor
interface, it is essential to simultaneously engineer the
nanocrystal and molecular parameters in order to uncover
physical principles that are transferable among various systems.
Such unified principles, however, are still lacking in the field,
which is the motivation of the current work. Here we study TET
from CsPbBr3 NCs of varying sizes to carboxylated anthracene
isomers, using a combination of static and time-resolved
spectroscopy, and investigate their impact on photon
upconversion efficiencies. CsPbBr3 perovskite NCs were chosen
because of their unique “defect-tolerance” and distinctive
spectroscopic features well suited to TET studies.[4i, 4j, 6c, 9] We
find that the NC-molecule coupling mechanisms are “through-
bond” and “through-space” when the carboxyl and anthracene
moieties in the molecule are co-planar or orthogonal,
respectively. The relative strength of the two coupling
mechanisms is controlled by the “wavefunction-leakage” of NCs,
with small-size, strongly-confined NCs favoring the “through-
space“ mechanism. Such a wavefunction-engineering strategy is
further validated using CdSe@ZnS core@shell NCs of varying
shell thicknesses.
energy transfer and photon upconversion efficiencies of
a
nanocrystal/molecule system can be improved by orders of
magnitude.
Introduction
Sensitization of molecular spin-triplet states is essential for
many applications, including but not limited to organic
photoredox catalysis, photodynamic therapy, room temperature
phosphorescence, and triplet-triplet annihilation based photon
upconversion.[1] Development of efficient triplet sensitizers has
thus been an important subject of research in photochemistry.[1-2]
It was until very recently that colloidal semiconductor
nanocrystals (NCs), which have been under study for over three
decades,[3] were recognized as excellent triplet sensitizers.[4]
Compared to traditional organic or organometallic sensitizers,
NCs have many advantages such as negligible intersystem
crossing energy loss, strong light absorption and facile spectral
tunability.[4f-h, 5]
Results and discussion
In the past few years, researchers have systematically
investigated how the NC parameters, such as sizes[4i, 6] and shell
thicknesses[7], influence the efficiency of triplet energy transfer
(TET) from NCs to molecular acceptors. For example, the NC
size effect in triplet sensitization has been reported for CdSe
NCs,[6a] PbS(Se) NCs[6b] and more recently CsPbBr3 perovskite
NCs.[4i, 6c] In general, a decrease in the NC size can enhance
both the driving force[6b] and the electronic coupling[4i, 6c] required
for TET, which increases the efficiencies of TET and photon
upconversion based on it. Besides NC parameters, molecular
We first examined the performances of triplet-triplet
annihilation photon upconversion (TTA-UC) using CsPbBr3 NCs
with surface-anchored anthracene carboxylic acid (ACA) ligands.
Fig. 1a shows the UV-vis absorption spectra of three different-
sized CsPbBr3 NCs prepared using a hot injection method;[4i, 10]
see Methods in the Supporting Information (SI) for details. The
lowest energy absorption peaks of these samples are situated at
450, 454 and 468 nm, which are used label the NCs in the
following. According to their transmission electron microscope
(TEM) images (Fig. S1), these samples correspond to NC sizes
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
(edge lengths) of ~3.2, 3.4 and 4.3 nm, respectively, which are
consistent with the previously-reported sizing curve for CsPbBr3
NCs.[11] The size distributions of these NCs are in the range of
4.6 to 9.4% (Fig. S1), resulting in well-resolved, multiple
excitonic absorption peaks in Fig. 1a.
ACA ligands further pass the triplet energy to DPA molecules
(ET ~ 1.77 eV) in the solution (TET-2) which emit upconverted
photons via TTA. Because of a weak exchange interaction on
the order of meV in CsPbBr3 NCs,[16] we can assume that the
lowest exciton energy determined from the absorption spectrum
is the triplet energy, which is 2.76, 2.73 and 2.65 eV for NC450,
NC454 and NC468, respectively. Thus, the driving forces for
TET from NCs to ACA ligands are very strong.
In order to efficiently excite the CsPbBr3 NCs but not ACA
(absorption spectra in Fig. S2) and DPA (absorption spectrum in
Fig. S3) in the mixture, we used a continuous wave (cw) laser of
443 nm; see SI for details. As shown in Fig. 1d, for the
NC/ACA/DPA mixture, the UC light can be seen through a short-
pass filter that cuts off the laser light and the residual NC
emission. Fig. S4 shows the PL spectra of NC/ACA/DPA
systems with different NC sizes under 443 nm excitation, where
both the residual emission of NCs and the UC emissions can be
observed. The UC emission intensity increases with excitation
power first quadratically and then linearly (Fig. S5), which is a
typical feature of sensitized TTA-UC.[1]
Table 1. TTA-UC and TET efficiencies of CsPbBr3 NC/ACA systems.
Figure 1. CsPbBr3 perovskite NC-ACA systems for photon upconversion.
(a) UV-vis absorption spectra of CsPbBr3 NCs with lowest exciton peaks at
450 (yellow), 454 (orange) and 468 nm (red). The peak positions are used to
label the samples. (b) Chemical structures of the 2-anthracene carboxylic acid
(ACA) and 9-ACA isomers. Gray: C; white: H; red: O. (c) Schematic and
electronic level diagram of the system for photon upconversion (UC). The
system comprises ACA-anchored CsPbBr3 NCs and diphenylanthracene (DPA)
dispersed in deaerated hexane solvent. (d) Pictures of a typical NC/ACA/DPA
mixture (left) and a free NC solution (right) under 443 nm excitation. The UC
light can be seen through a short-pass filter (left).
NC450
NC454
NC468
2-ACA
9-ACA
2-ACA
9-ACA
2-ACA
9-ACA
Φ’UC (%)
QPL (%)
13.0
98.1
97.4
5.3
12.4
96.0
95.7
3.4
6.9
0.6
6.7
6.9
35.0
35.7
21.7
22.2
84.8
85.6
The chemical structures of the two ACA isomers, 2-ACA
and 9-ACA, are shown in Fig. 1b. They were anchored onto NC
QTR-PL (%)
12]
surfaces using a ligand exchange procedure (Methods).[4b,
The numbers of 2-ACA and 9-ACA ligands per NC were
controlled to be similar for each sample; these numbers are
~198, 175 and 255 for NC450, NC454 and NC468, respectively
(Table S1). In this way, any difference between the TET or TTA-
UC efficiencies of each NC anchored with 2-ACA and 9-ACA
ligands is not caused the difference in the number of ACA
molecules. Note that, because the carboxyl group is known to
be deprotonated upon anchoring the molecules onto NC
surfaces,[13] we show the deprotonated forms of the ACA
molecules and all the following discussions on ACA refer to
deprotonated ACA. Density functional theory (DFT)
optimizations of the ground-state geometries indicate that the
dihedral angles between the carboxylate and anthracene
moieties are close to 0 and 90 degrees, respectively, for 2-ACA
and 9-ACA (Fig. 1b). This is consistent with previous reports and
can be rationalized by that a large steric hindrance effect
between the carboxylate group and the peripheral H atoms on
anthracene exists in 9-ACA but not in 2-ACA.[14] As a result,
there is a resonance effect (i.e., wavefunction delocalization)
between the carboxylate and anthracene moieties for 2-ACA but
not for 9-ACA, which explains the red-shift (lower energy) of the
absorption onset of 2-ACA compared to that of 9-ACA (Fig. S2).
We will illustrate later that the resonance effect plays a key role
in the electronic coupling mechanism between NCs and ACA.
The calculated UC quantum yields (Φ’UC) are summarized in
Table 1; calculation details are provided in the SI. We can
identify three important trends when comparing these yields.
First, the UC yields generally decrease with increasing NC sizes.
Considering that the exciton energies of all three NCs (2.65-2.76
eV) are much higher than the triplet energy of ACA (1.83 eV),
the size-dependent electronic coupling for TET from NCs to ACA
is believed to be responsible for the observed trend.[4i] Secondly,
for each NC sample, NC/2-ACA has a higher UC yield than
NC/9-ACA, which is exactly opposite to previous observation for
CdSe NCs.[8a] Last but not the least, the UC yields of NC/9-ACA
samples display a stronger size-dependence than those of
NC/2-ACA. As a result, the yield difference increases from ~2.4-
fold for NC450 to 11.5-fold for NC468. Note that we also
measured TTA-UC using 1-ACA ligand, another common ACA
isomer, and the UC yields always fall in between the results of 2-
ACA and 9-ACA (Table S2). In order to understand these trends,
we performed detailed static and time-resolved spectroscopic
experiments.
The static PL spectra in Fig. S4 already imply the key
process determining the TTA-UC yields. The UC emission is
strong when the NC emission is efficiently quenched. The NC
emission quenching efficiencies by ACA (QPL) are tabulated in
Table 1, which reveals a clear correlation between QPL and Φ’UC
.
For TTA-UC measurements, we combined ACA-anchored
Thus, Φ’UC is mainly determined by the efficiency of TET from
CsPbBr3 NCs to ACA ligands (TET-1 in Fig. 1c), and the change
in NC sizes or ACA isomers mainly affects this TET efficiency.
Time-resolved PL measurements were also used to calculate
the quenching rates and efficiencies (see Fig. S4 and SI for
CsPbBr3 NCs and diphenylanthracene (DPA) in hexane solution,
4k, 6a, 15]
with the later acting as effective annihilators.[4d,
As
schematically shown in Fig. 1c, photoexcited NCs sensitize ACA
triplet (3ACA*; ET ~ 1.83 eV) through a TET process (TET-1); the
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
details). The quenching efficiencies calculated in this way (QTR-PL
agree well with the static PL quenching efficiencies (Table 1).
The quenching rates, i.e., TET rates (kTET), are in the range of
0.05 to 6.45 ns-1. We note that, because 9-ACA quenches the
PL of NC468 by only 6.7%, the PL quenching caused by trap
states introduced in the ligand exchange process is believed to
be negligible, consistent with previous studies.[4b, 9a]
)
thus a higher TTA-UC efficiency. TA spectra of NC450 and
NC468 and their NC/ACA complexes are shown in Figs. S6 and
S7 and the overall observations are similar to those of NC454.
The large difference between the rates of TET from
CsPbBr3 NCs to 2-ACA and 9-ACA cannot be attributed to
energetics (driving forces), as the excited state energies and
redox potentials of the two ACA isomers are very similar (Fig.
S8). This is consistent with the conclusion from a previous study
of TET from CdSe NCs to ACA isomers.[8a] Interestingly,
however, as we mentioned above, for CdSe NCs TET to 9-ACA
Because time-resolved PL only measures the excited state
decay of NCs, we also performed transient absorption (TA; see
Methods in the SI) experiments to directly observe both the
excited state decay of NCs and formation of ACA triplets.[4b] Figs. is more efficient than to 2-ACA.[8a] Rationalizing the opposite
2a and 2b show the TA spectra of NC454 in the absence and
presence of 2-ACA ligands, respectively, measured under the
same conditions (460 nm pump). The TA features of free NCs
decay away within ~10 ns due to radiative and/or nonradiative
recombinations of band edge electrons and holes. In the
presence of 2-ACA, the TA features of NCs show rapid decay
within ~1 ns, and meanwhile, a new absorptive feature at ~430
nm emerges which can be assigned to the T1→Tn transition of
ACA triplets (3ACA*).[4b] This triplet spectrum becomes especially
obvious on the μs timescale when the NC features completely
disappear (Fig. 2b). In line with TR-PL measurements, 9-ACA
quenches the NC TA features much more slowly than 2-ACA, as
clearly shown by the comparison of the XB kinetics probed for
NC454, NC454/2-ACA and NC454/9-ACA in Fig. 2c.
trends observed for CsPbBr3 and CdSe NCs might be a key to
understanding TET from NCs to surface-anchored acceptors.
We approach this problem by analyzing the fundamental
photophysics of semiconductor NCs. The CdSe NCs used in
previous studies often have the lowest energy absorption peak
at ~500-520 nm, corresponding to a NC diameter of ~2.3-2.5
nm.[4b, 8a] The sizes of CsPbBr3 NCs used here are in the range
of 3.2-4.3 nm. Because the Bohr exciton diameters of bulk CdSe
and CsPbBr3 are ~11.2 (ref[17]) and 7 nm (ref[18]), respectively,
CdSe NCs are more strongly quantum-confined than CsPbBr3
NCs. Quantum confinement strongly affects the electronic
coupling between NCs and surface-anchored acceptors
because more strongly confined NCs tend to have higher carrier
wavefunction amplitudes on NC surfaces.[3a, 19] The difference in
surface wavefunction amplitudes and associated electronic
coupling strengths with surface-anchored acceptors might be
responsible for the opposite trends observed for CdSe and
CsPbBr3 NCs.
In order to verify the wavefunction argument, we
systematically tune wavefunction amplitudes on CdSe NC
surfaces by coating them with ZnS shells of varying
thicknesses[20]. Because of a type-I band alignment between
CdSe and ZnS, the ZnS shell functions as a tunneling layer for
charge/triplet energy transfer from the CdSe core to ACA ligands.
Details for the synthesis of CdSe@ZnS core@shell NCs are
provided in Methods in the SI. The shell thicknesses were
controlled to correspond to 0.5, 1 and 2 monolayers (MLs) of
ZnS; see TEM images in Fig. S9. UV-vis absorption spectra of
the core and core@shell NCs are shown in Fig. S10. The lowest
energy absorption peak only slightly shifts from 505 to 515 nm
(48 meV shift) upon coating the core with 2 MLs of ZnS. Thus,
these core@shell NCs can be used to control wavefunction
amplitudes on NC surfaces with the energetics remaining
virtually unchanged.
Figure 3. Transient absorption measurements. TA spectra of (a) CsPbBr3
NC454 and (b) NC454/2-ACA at indicated time delays following the excitation
by 460 nm pulses. The triplet absorption feature (3ACA*) of 2-ACA is amplified
by a factor of 3 for clarity. (c) TA kinetics probed at the XB features of free
NCs (red), NC/2-ACA (orange) and NC/9-ACA (yellow). (d) Triplet decay
kinetics of 2-ACA (orange) and 9-ACA (yellow) sensitized by NC454.
Table 2. TET efficiencies of CdSe@ZnS NC/ACA systems.
Interestingly, the lifetimes of sensitized 2-ACA and 9-ACA
triplets are remarkably different. As shown in Fig. 2d, the 1/e
decay times are ~50 μs and 1 ms for 2-ACA and 9-ACA,
respectively, with the latter value consistent with the literature
result.[4b] The reason for the difference is not clear yet, but is
likely associated with the resonance effect between the
carboxylate and anthracene moieties existing in 2-ACA but not
in 9-ACA. The resonance effect renders the excited state of 2-
ACA susceptible to more nuclear degrees of freedom. The 20-
fold longer triplet lifetime of 9-ACA than 2-ACA, combined with
the higher TTA-UC efficiencies of NC/2-ACA systems, further
supports that the efficiency of TET from NCs to ACA ligands
(TET-1) rather than from ACA ligands to DPA (TET-2)
determines the TTA-UC efficiency. Otherwise, a longer triplet
lifetime of 9-ACA should result in a higher TET-2 efficiency and
CdSe
CdSe@0.5ZnS
2-ACA
9-ACA
2-ACA
9-ACA
QPL (%)
72.9
70.9
89.2
87
33.6
34.1
41.2
41
QTR-PL (%)
CdSe@1ZnS
CdSe@2ZnS
2-ACA
9-ACA
2-ACA
9-ACA
QPL (%)
20.6
19.4
13
15.8
15.4
3.8
4.3
QTR-PL (%)
12.1
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
a
ACA ligands were functionalized onto CdSe@ZnS
core@shell NC surfaces in a similar way as CsPbBr3 NCs. The
average numbers of 2-ACA and 9-ACA ligands per NC are
calculated to be ~13, 11, 10 and 13 for CdSe, CdSe@0.5ZnS,
CdSe@1ZnS and CdSe@2ZnS NCs, respectively (Table S1).
These numbers are in general one order of magnitude lower
than those for CsPbBr3 NCs, likely because the surface of
CsPbBr3 NCs is highly dynamic,[21] allowing more ACA
molecules to bind to their surface through ligand exchange.
strongly-confined NCs
b
weakly-confined or shell-coated NCs
9-ACA
9-ACA
2-ACA
2-ACA
through-bond (TB) coupling
through-space (TS) coupling
c
d
100
CdSe
TS > TB
TS > TB
100
CdSe
CdSe@0.5ZnS
CdSe@1ZnS
10-1
10-2
10-3
CdSe@0.5ZnS
CdSe@1ZnS
CPB468
CdSe@2ZnS
Because TET from CdSe NCs to ACA ligands has already
been well established in the literature using TA,[4b, 8a] here we
use PL (Fig. S11) and TR-PL (Fig. S12) to simply quantify the
TET efficiencies. The calculated static and TR-PL quenching
efficiencies (QPL and QTR-PL) are summarized in Table 2. Indeed,
9-ACA quenches the PL of CdSe NCs more efficiently than 2-
ACA.[8a] With increasing ZnS shell thicknesses, the quenching
efficiencies of both 2-ACA and 9-ACA decrease. Notably,
however, there is an interesting crossover between the relative
quenching efficiencies of 2-ACA and 9-ACA. While 9-ACA still
quenches the PL of CdSe@0.5ZnS NCs slightly more efficiently
than 2-ACA, 2-ACA becomes a more efficient quencher for
CdSe@1ZnS and CdSe@2ZnS NCs.
CPB450
CPB450
CPB454
CPB454
10-1
CPB468
CdSe@2ZnS
TB > TS
TB > TS
Figure 3. Through-space and through-bond TET. (a) Wavefunction leakage
from strongly-confined NCs enables efficient through-space (TS) coupling with
9-ACA. (b) For weakly-confined or shell-coated NCs with weak wavefunction
leakage, through-bond (TB) coupling to 2-ACA dominates over through-space
coupling to 9-ACA. (c) Calculated products of electron and hole wavefunctions
on NC surfaces (|ΨeΨh|2) for CsPbBr3 NCs of varying sizes (blue triangles)
and CdSe@ZnS NCs of varying shell thicknesses (red squares). The solid
lines are guides to the eye. The green and blue shaded areas correspond to
through-space and through-bond dominated regimes, respectively, separated
by a gray dashed line. (d) The same plot as (c) but for calculated hole
wavefunctions on NC surfaces (|Ψh|2).
The crossover behavior observed for CdSe@ZnS NCs
suggests that carrier probability densities on NC surfaces, or
wavefunction squared (|ΨS|2), indeed paly an essential role in
the TET mechanism. For strongly-confined CdSe NCs |ΨS|2 is
large, whereas for relatively weakly-confined CsPbBr3 NCs or
shell-coated CdSe@ZnS NCs |ΨS|2 is strongly reduced. Thus,
an empirical rule is that a large |ΨS|2 favors TET to 9-ACA over
2-ACA whereas a relatively small |ΨS|2 does the opposite.
Taking into consideration the different molecular geometries of
9-ACA and 2-ACA mentioned above, we propose the following
explanations to this empirical rule.
In order to verify the proposed mechanisms, we quantify the
wavefunction squared on NC surfaces (|ΨS|2) for CsPbBr3 NCs
of varying sizes and CdSe@ZnS NCs of varying shell
thicknesses using an effective mass approximation (EMA; see
SI and Table S3 for calculation details). For CsPbBr3 NCs, the
calculated wavefunction squared amplitudes on NC surfaces for
both the electron (|Ψe|2) and the hole (|Ψh|2) decay with
increasing NC sizes (Fig. S13). For CdSe@ZnS NCs, both |Ψe|2
and |Ψh|2 decay with increasing shell thicknesses, with the latter
decaying more rapidly due to a larger effective mass for the hole
than the electron and thus a lower hole tunneling capability (Fig.
S13). In Fig. 3c, we plot the product of |Ψe|2 and |Ψh|2, |ΨeΨh|2,
for CsPbBr3 and CdSe@ZnS NCs, by scaling the value of core-
only CdSe NCs to 1. From the plot, CdSe core and
CdSe@0.5ZnS NCs have larger |ΨeΨh|2 than CdSe@1ZnS,
CdSe@2ZnS and all the CsPbBr3 NCs. This correlates well with
the observations that CdSe core and CdSe@0.5ZnS NCs
transfer triplet energy more efficiently to 9-ACA than to 2-ACA
whereas all other NCs behave exactly oppositely. Thus, we can
draw a horizontal line (gray dashed) in Fig. 3c to represent the
approximate strength for the through-bond coupling pathway;
above and below the line through-space and through-bond
mechanisms dominate, respectively.
When |ΨS|2 is large, through-space electronic coupling
between NCs and surface-anchored acceptors dominates (Fig.
3a). In this case, coupling with 9-ACA is more efficient than 2-
ACA because the center of the anthracene moiety of 9-ACA is
closer to NCs than that of 2-ACA (Fig. 3a). For weakly-confined
or shell-coated NCs with small |ΨS|2, through-space coupling is
not as efficient. In this case, the chemical resonance effect
(wavefunction delocalization) in 2-ACA, enabled by the co-
planar configuration between the carboxylate and anthracene
moieties, opens up an additional pathway for the coupling
between NCs and 2-ACA (Fig. 3b). We call this a through-bond
mechanism, with the bond referring the coordination bond
between the carboxylate group and the surface atoms of NCs.
The orthogonal geometry in 9-ACA disrupts the resonance and
hence disables the through-bond mechanism. As a result,
electronic coupling between NCs and 2-ACA is stronger than 9-
ACA when |ΨS|2 is small.
The use of |ΨeΨh|2 assumes a concerted mechanism for TET,
i.e., simultaneous transfer of the electron and the hole from NCs
to ACA. However, previous studies suggest that even the one-
step, direct TET process could be mediated by a virtual charge-
transfer state.[9a, 22] Because of the very high reduction potential
energy of ACA molecules (Fig. S8), the virtual charge-transfer
state here is likely associated with hole transfer from NCs to
ACA.[9a] For this reason, we present another set of analysis
using |Ψh|2 alone. As shown in Fig. 3d, in this case, |Ψh|2 values
of CdSe core and CdSe@0.5ZnS NCs still lie above those of all
other NCs. Thus, regardless of the detailed TET mechanisms,
we can reach the same conclusion that through-space coupling
dominates in CdSe core and CdSe@0.5ZnS NCs whereas all
other NCs fall in the through-bond regime.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
Through-space and through-bond charge and/or energy
transfer have been extensively studied in organic donor-bridge-
acceptor supramolecules[23] and more recently also in
inorganic/organic hybrid systems.[24] Through-bond interaction is
essentially a super-exchange mechanism, which involves mixing
Keywords: Triplet energy transfer • through-bond and through-
space • perovskite nanocrystals • photon upconversion
of the donor and acceptor orbitals with the orbitals of the
25]
bridge.[23a,
Because both through-space and through-bond
interactions should decay exponentially with the donor-acceptor
distance, in many cases it is a challenge to disentangle these
two mechanisms.[23a] Herein, by systematically tuning the
quantum confinement strengths of inorganic NCs and by using
carboxylated acceptors with or without a resonance effect, we
are able to differentiate the through-space and through-bond
mechanisms at the inorganic/organic interfaces.
[1]
[2]
T. F. Schulze, T. W. Schmidt, Energy & Environ. Sci.
2015, 8, 103-125.
a) T. N. Singh-Rachford, F. N. Castellano, Coord. Chem.
Rev. 2010, 254, 2560-2573; b) J. Zhao, W. Wu, J. Sun, S.
Guo, Chem. Soc. Rev. 2013, 42, 5323-5351.
[3]
[4]
a) L. E. Brus, J. Chem. Phys. 1983, 79, 5566-5571; b) A.
P. Alivisatos, Science 1996, 271, 933-937.
a) M. Wu, D. N. Congreve, M. W. B. Wilson, J. Jean, N.
Geva, M. Welborn, T. Van Voorhis, V. Bulović, M. G.
Bawendi, M. A. Baldo, Nat. Photon. 2015, 10, 31-34; b) C.
Mongin, S. Garakyaraghi, N. Razgoniaeva, M. Zamkov, F.
N. Castellano, Science 2016, 351, 369-372; c) Z. Huang,
M. L. Tang, J. Am. Chem. Soc. 2017, 139, 9412-9418; d)
Z. Huang, X. Li, M. Mahboub, K. M. Hanson, V. M.
Nichols, H. Le, M. L. Tang, C. J. Bardeen, Nano Lett.
2015, 15, 5552-5557; e) K. Okumura, K. Mase, N. Yanai,
N. Kimizuka, Chem. Eur. J. 2016, 22, 7721-7726; f) N.
Yanai, N. Kimizuka, Acc. Chem. Res. 2017, 50, 2487-
2495; g) Z. Huang, M. L. Tang, J. Phys. Chem. Lett. 2018,
9, 6198-6206; h) L. Nienhaus, M. Wu, V. Bulović, M. A.
Baldo, M. G. Bawendi, Dalton Transactions 2018, 47,
8509-8516; i) X. Luo, R. Lai, Y. Li, Y. Han, G. Liang, X.
Liu, T. Ding, J. Wang, K. Wu, J. Am. Chem. Soc. 2019,
141, 4186-4190; j) S. He, X. Luo, X. Liu, Y. Li, K. Wu, J.
Phys. Chem. Lett. 2019, 10, 5036-5040; k) Y. Han, S. He,
X. Luo, Y. Li, Z. Chen, W. Kang, X. Wang, K. Wu, J. Am.
Chem. Soc. 2019, 141, 13033-13037; l) P. Xia, E. K.
Raulerson, D. Coleman, C. S. Gerke, L. Mangolini, M. L.
Tang, S. T. Roberts, Nat. Chem. 2020, 12, 137-144; m) H.
Lu, Z. Huang, M. S. Martinez, J. C. Johnson, J. M. Luther,
M. C. Beard, Energy & Environ. Sci. 2020, 13, 1347-1376.
S. Garakyaraghi, F. N. Castellano, Inorg. Chem. 2018, 57,
2351-2359.
a) Z. Huang, X. Li, B. D. Yip, J. M. Rubalcava, C. J.
Bardeen, M. L. Tang, Chem. Mat. 2015, 27, 7503-7507; b)
M. Mahboub, H. Maghsoudiganjeh, A. M. Pham, Z.
Huang, M. L. Tang, Adv. Funct. Mater. 2016, 26, 6091-
6097; c) Y. Han, X. Luo, R. Lai, Y. Li, G. Liang, K. Wu, J.
Phys. Chem. Lett. 2019, 10, 1457-1463.
a) M. Mahboub, Z. Huang, M. L. Tang, Nano Lett. 2016,
16, 7169-7175; b) V. Gray, P. Xia, Z. Huang, E. Moses, A.
Fast, D. A. Fishman, V. I. Vullev, M. Abrahamsson, K.
Moth-Poulsen, M. Lee Tang, Chem. Sci. 2017, 8, 5488-
5496; c) Z. Huang, Z. Xu, M. Mahboub, X. Li, J. W. Taylor,
W. H. Harman, T. Lian, M. L. Tang, Angew. Chem. Int. Ed.
2017, 56, 16583-16587; d) Z. Huang, P. Xia, N.
Megerdich, D. A. Fishman, V. I. Vullev, M. L. Tang, ACS
Photon. 2018, 5, 3089-3096; e) M. Mahboub, P. Xia, J.
Van Baren, X. Li, C. H. Lui, M. L. Tang, ACS Energy Lett.
2018, 767-772.
This study provides important guidelines for the molecular
design of acceptors. For strongly-confined NCs with strong
wavefunction leakage onto the surfaces, the anchoring group
should be functionalized at a position that allows for a maximal
spatial overlap between the NC wavefunction and the molecular
orbitals. In other circumstances, however, the size tuning of NCs
might not be able to attain strong quantum confinement.
Accordingly, anchoring groups like carboxylate should be placed
at a position allowing for a co-planar molecular geometry and
thus a resonance effect between the group and the molecule, in
order to enable the through-bond electronic coupling mechanism.
Conclusions
In summary, we studied triplet energy transfer from CsPbBr3
perovskite NCs of varying sizes and CdSe@ZnS NCs of varying
shell thicknesses to surface-anchored, carboxylated anthracene
isomers and the associated photon upconversion performances.
We find that the NC-molecule coupling mechanisms are
“through-bond” and “through-space” when the carboxyl and
anthracene moieties in the molecule are co-planar or orthogonal,
respectively. The relative strength of the two coupling
mechanisms can be controlled by the “wavefunction-leakage” of
NCs, through either NC size or shell thickness. By
simultaneously engineering nanocrystal wavefunction leakage
and molecular geometry, triplet energy transfer and photon
upconversion efficiencies of the CsPbBr3 NC/anthracene system,
for example, can be improved by orders of magnitude. This
study represents important progress towards a molecular-level
understanding of energy migration across the inorganic/organic
interface, and provides transferable guidelines for NC
engineering and molecular design for efficient energy transfer
useful in many applications.
[5]
[6]
[7]
[8]
[9]
a) P. Xia, Z. Huang, X. Li, J. J. Romero, V. I. Vullev, G. S.
H. Pau, M. L. Tang, Chem. Commun. 2017, 53, 1241-
1244; b) J. De Roo, Z. Huang, N. J. Schuster, L. S.
Hamachi, D. N. Congreve, Z. Xu, P. Xia, D. A. Fishman, T.
Lian, J. S. Owen, M. L. Tang, Chem. Mat. 2020, 32, 1461-
1466.
a) X. Luo, Y. Han, Z. Chen, Y. Li, G. Liang, X. Liu, T. Ding,
C. Nie, M. Wang, F. N. Castellano, K. Wu, Nat. Commun.
2020, 11, 28; b) X. Luo, G. Liang, Y. Han, Y. Li, T. Ding, S.
He, X. Liu, K. Wu, J. Am. Chem. Soc. 2020, 142, 11270-
11278.
Notes
The authors declare no competing financial interest.
Acknowledgments
K.W. acknowledges financial support from the National Natural
Science Foundation of China (21975253, 51961165109), the
Strategic Pilot Science and Technology Project of Chinese
Academy of Sciences (XDA21010206), the Dalian City
Foundation for
(2018J12GX051) and Dalian Institute of Chemical Physics
(DICP I201914).
[10]
[11]
Y. Dong, T. Qiao, D. Kim, D. Parobek, D. Rossi, D. H.
Son, Nano Lett. 2018, 18, 3716-3722.
a) Y. Li, T. Ding, X. Luo, Z. Chen, X. Liu, X. Lu, K. Wu,
Nano Res. 2019, 12, 619–623; b) Y. Li, R. Lai, X. Luo, X.
Liu, T. Ding, X. Lu, K. Wu, Chem. Sci. 2019, 10, 5983-
5989; c) M. C. Brennan, J. E. Herr, T. S. Nguyen-Beck, J.
Zinna, S. Draguta, S. Rouvimov, J. Parkhill, M. Kuno, J.
Science and Technology Innovation
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
Am. Chem. Soc. 2017, 139, 12201-12208; d) J. Maes, L.
Balcaen, E. Drijvers, Q. Zhao, J. De Roo, A. Vantomme,
F. Vanhaecke, P. Geiregat, Z. Hens, J. Phys. Chem. Lett.
2018, 9, 3093-3097; e) Y. Li, K. Wu, Angew. Chem. Int.
Ed. 2020, 10.1002/anie.202004668.
[19]
a) W. A. Tisdale, X.-Y. Zhu, Proc. Natl. Acad. Sci. 2011,
108, 965-970; b) H. Zhu, N. H. Song, T. Q. Lian, J. Am.
Chem. Soc. 2010, 132, 15038-15045; c) H. Zhu, Y. Yang,
K. Hyeon-Deuk, M. Califano, N. Song, Y. Wang, W.
Zhang, O. V. Prezhdo, T. Lian, Nano Lett. 2014, 14, 1263-
1269.
[12]
a) C. Mongin, P. Moroz, M. Zamkov, F. N. Castellano, Nat.
Chem. 2018, 10, 225-230; b) H. Lu, X. Chen, J. E.
Anthony, J. Johnson, M. C. Beard, J. Am. Chem. Soc.
2019, 141, 4919-4927.
[20]
[21]
C. Pu, J. Zhou, R. Lai, Y. Niu, W. Nan, X. Peng, Nano
Res, 2013, 6, 652-670.
J. De Roo, M. Ibáñez, P. Geiregat, G. Nedelcu, W.
Walravens, J. Maes, J. C. Martins, I. Van Driessche, M. V.
Kovalenko, Z. Hens, ACS Nano 2016, 10, 2071-2081.
R. D. Harcourt, G. D. Scholes, K. P. Ghiggino, J. Chem.
Phys. 1994, 101, 10521-10525.
a) D. Gosztola, B. Wang, M. R. Wasielewski, J.
Photochem. Photobiol., A 1996, 102, 71-80; b) M. D.
Ward, Chem. Soc. Rev. 1997, 26, 365-375; c) E. A.
Weiss, L. E. Sinks, A. S. Lukas, E. T. Chernick, M. A.
Ratner, M. R. Wasielewski, J. Phys. Chem. B 2004, 108,
10309-10316; d) D. Bellows, T. Goudreault, S. M. Aly, D.
Fortin, C. P. Gros, J.-M. Barbe, P. D. Harvey,
Organometallics 2010, 29, 317-325.
a) A. J. Morris-Cohen, M. D. Peterson, M. T. Frederick, J.
M. Kamm, E. A. Weiss, J. Phys. Chem. Lett. 2012, 3,
2840-2844; b) X. Li, Z. Huang, R. Zavala, M. L. Tang, J.
Phys. Chem. Lett. 2016, 7, 1955-1959.
G. D. Scholes, K. P. Ghiggino, A. M. Oliver, M. N. Paddon-
Row, J. Am. Chem. Soc. 1993, 115, 4345-4349.
[13]
[14]
D. M. Kroupa, M. Vörös, N. P. Brawand, B. W. McNichols,
E. M. Miller, J. Gu, A. J. Nozik, A. Sellinger, G. Galli, M. C.
Beard, Nat. Commun. 2017, 8, 15257.
a) W. Rodríguez-Córdoba, R. Noria-Moreno, P. Navarro,
J. Peon, J. Lumin. 2014, 145, 697-707; b) T. C. Werner, T.
Matthews, B. Soller, J. Phys. Chem. 1976, 80, 533-541; c)
N. Ghoneim, D. Scherrer, P. Suppan, J. Lumin. 1993, 55,
271-275.
[22]
[23]
[15]
[16]
X. Li, A. Fast, Z. Huang, D. A. Fishman, M. L. Tang,
Angew. Chem. Int. Ed. 2017, 56, 5598-5602.
M. A. Becker, R. Vaxenburg, G. Nedelcu, P. C. Sercel, A.
Shabaev, M. J. Mehl, J. G. Michopoulos, S. G.
Lambrakos, N. Bernstein, J. L. Lyons, T. Stöferle, R. F.
Mahrt, M. V. Kovalenko, D. J. Norris, G. Rainò, A. L. Efros,
Nature 2018, 553, 189-193.
[24]
[25]
[17]
[18]
D. J. Norris, M. G. Bawendi, Phys. Rev. B 1996, 53,
16338-16346.
L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R.
Caputo, C. H. Hendon, R. X. Yang, A. Walsh, M. V.
Kovalenko, Nano Lett. 2015, 15, 3692-3696.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202009066
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Text for Table of Contents
Shan He, Runchen Lai, Qike Jiang,
photon
2-ACA
YaoYao Han, Xiao Luo, Yuyang Tian, Xue
Liu and Kaifeng Wu
upconversion
By simultaneously engineering
9-ACA
molecular
geometry
and
nanocrystal
wavefunction,
Page 1-Page 6
energy transfer and photon
upconversion efficiencies of a
nanocrystal/molecule system
can be improved by orders of
magnitude.
through-bond
Engineering Sensitized Photon
Upconversion Efficiency via Nanocrystal
Wavefunction and Molecular Geometry
nanocrystal
through-space
7
This article is protected by copyright. All rights reserved.