Quantifying Ligand Exchange on InP Using an Atomically Precise Cluster Platform

Article abstract of DOI:10.1021/acs.inorgchem.8b03524

The surface chemistry of a colloidal nanoparticle is intrinsic to both its structure and function. It is therefore necessary to characterize the surfaces of colloidal materials to rationally underpin any synthetic, catalytic, or transformative mechanisms they enable. Here we characterize the surface properties of colloidal InP clusters and quantum dots by examining the binding of traditional stabilizing ligands including carboxylates, phosphonates, and thiolates. By using the In₃₇P₂₀X₅₁ (X = carboxylate) cluster species as an ideally monodisperse and well-defined starting scaffold, we quantify surface-exchange equilibria. Using quantitative ¹H and ³¹P NMR spectroscopy, we show that 1:1 metathesis-type binding models are insufficient to fully describe the surface dynamics. In particular, for the case of the highly reversible carboxylate ligand exchange, a more detailed isotherm approach using a two-site, competitive model is necessary. This model is used to deconvolute L- A nd X-type binding modalities. We additionally quantify the reversible and irreversible ligand-exchange reactions observed in the thiolate and phosphonate systems.

Full text of DOI:10.1021/acs.inorgchem.8b03524

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Quantifying Ligand Exchange on InP Using an Atomically Precise
Cluster Platform
Andrew Ritchhart and Brandi M. Cossairt*
Department of Chemistry, University of Washington, Seattle, Washington 98195-1700, United States
S
* Supporting Information
ABSTRACT: The surface chemistry of a colloidal nanoparticle is intrinsic to both its
structure and function. It is therefore necessary to characterize the surfaces of
colloidal materials to rationally underpin any synthetic, catalytic, or transformative
mechanisms they enable. Here we characterize the surface properties of colloidal InP
clusters and quantum dots by examining the binding of traditional stabilizing ligands
including carboxylates, phosphonates, and thiolates. By using the In37P20X51 (X =
carboxylate) cluster species as an ideally monodisperse and well-defined starting
scaffold, we quantify surface-exchange equilibria. Using quantitative 1H and 31P NMR
spectroscopy, we show that 1:1 metathesis-type binding models are insufficient to
fully describe the surface dynamics. In particular, for the case of the highly reversible carboxylate ligand exchange, a more
detailed isotherm approach using a two-site, competitive model is necessary. This model is used to deconvolute L- and X-type
binding modalities. We additionally quantify the reversible and irreversible ligand-exchange reactions observed in the
thiolate and phosphonate systems.
Furthermore, quantum dots (QDs) are frequently treated with
additives such as HF and Zn2+ in order to improve photo-
luminescent quantum yields by removing defects such as
dangling bonds at the surface.26−28 Finally, postsynthetic ligand
modification is required for tuning interparticle charge and
exciton transfer in thin films of colloidal semiconductor
nanocrystals.29 The nature of these applications implicitly
requires postsynthetic surface modification because the resulting
surface chemistries would otherwise be significantly altered or
the chemistry would preclude particle nucleation and growth.30
Therefore, a robust understanding of the ligand behavior not
only directly benefits nanoparticle synthetic design but also can
enable it to be separately optimized and better compartmental-
ized from applications.
Previous work in the literature has elegantly defined the
binding affinities and characteristics of common ligands in the
cases of CdSe,31−34 PbS,35,36 and perovskites.37,38 Using NMR
spectroscopy, it has been shown that equilibrium models for the
binding of ligands on these surfaces can be accurately and easily
measured. These previous analyses, however, suffer from
fundamental limitations of measuring the ensemble properties
of inherently polydisperse samples. Because these systems are
not perfectly uniform and vary with respect to the particle size,
and, moreover, the number and type of binding sites, simplifying
assumptions must be made or the model must take these factors
into consideration. This greatly complicates the modeling,
reducing the translatability and interpretation across particle
sizes, compositions, and morphologies. Critically, many of these
systems lack experimental quantification of ligand binding
INTRODUCTION
■
Our understanding of the chemistry of colloidal semiconductor
nanoparticles has advanced to enable progress in applying these
materials in a wide range of applications including catalysis,1,2
photovoltaics,3,4 displays,5 and imaging.6,7 Recent efforts have
focused on developing nontoxic materials, such as InP, for large-
scale, highly distributed applications.8,9 Research into II−IV and
III−V nanomaterials has been tailored to the applications at
hand by manipulating the quantum confinement effect as a
function of the size,10,11 shape,12,13 and composition.14 Under-
pinning this research, and fundamental to colloidal nanoparticle
function and synthesis, is the nature of the nanoparticle surface.
The surface and ligand layer represent a highly complex
component of the nanoparticle composition that has influence
on the structural, electronic, and reactivity properties of the
nanomaterial.15,16 Accurately quantifying the binding properties
of ligands is critical for the design of size-tunable and anisotropic
nanoparticle syntheses,17,18 as well as for postsynthetic
modifications such as shelling, passivation, or cation ex-
change.19−21 A detailed, analytical understanding of the surface
structure and ligand coordination properties of a colloidal
system must underscore any rational design of the nanoscale
properties in those systems.
Postsynthetic ligand modification specifically has enabled a
host of targeted applications. Exchanging ligands of differing
hydrophobicity has long been known to facilitate changes in the
nanoparticle solubility, a critical parameter to control
applications in biosensing.22 In a similar manner, modifying
mixed-ligand shells to tune the hydrophobicity has been used to
coordinate the formation of nanoparticle superstructures.23 For
catalysis applications, ligand exchange has been used to improve
the C−C coupling rate24 as well as H2 production.25
Received: December 18, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
modes in the first place, relying on theory or analogy to distantly
related molecular structures. Here we illustrate the veracity of
using the precisely known In37P20X51 (X = carboxylate) cluster
as a model for the InP nanocrystal surface. This cluster can be
synthesized and purified on a large scale, and its surface
chemistry is precisely known from single-crystal X-ray
diffraction analysis.39 Notably, despite the rising prominence
of InP nanocrystals for emissive applications, no such
investigation into the ligation and surface binding properties
of InP has been performed to date.
RESULTS AND DISCUSSION
Carboxylate−Carboxylic Acid Exchange. Carboxylates
have historically been the ligand of choice for the synthesis of
InP nanoparticles both for general laboratory use and for their
application in commercialized display technologies.5 The
exchange between X-type carboxylate ligands (Figure 1) such
■
In order to answer outstanding questions surrounding InP
nanoparticle surface chemistry, we adopt an analytical 1H NMR
spectroscopic approach used to great success in the CdSe and
PbS literature.31,35,40 Traditional aliphatic ligands are not
diagnostically useful in the 1H NMR spectrum of nanoparticles
because of the excessive overlapping and polydispersity of their
alkyl resonances in the upfield region. By using alkene-labeled
ligands, it is possible to quantitatively measure both free and
bound ligands on the nanoparticle surface in the relatively clean
4.0−6.0 ppm region. It is convenient for our purposes then to
use oleate ligands bearing an internal alkene on the native
particle surface in conjunction with exchanging ligands
bearing terminal alkenes to enable a complete deconvolution
and analytical assessment of the ligand-exchange equilibria. The
1
Figure 1. H NMR spectra of the alkene region for titration of DDA
into a solution of the Ol-capped In37P20 cluster, from 0 equiv (red) to
139 equiv (violet). Resonances at 5.8−5.9 and 5.0−5.2 ppm correspond
to DDA, while those at 5.4−5.7 ppm correspond to Ol. Inset: 31P NMR
spectra of the starting cluster (blue) and the cluster + 139 equiv of DDA
(orange).
1
sources of H resonance shifts in nanoparticle solutions have
recently been examined in detail, including the source of the
peak shifts and broadening of ligated species.41 By using a high-
field instrument, ligated molecules can be reliably shifted and
deconvoluted from their free counterparts. Herein we will use
these 1H NMR properties to model ligand-exchange equilibria
as reversible chemical processes (Scheme 1) using quantitative
as oleate (Ol) and dodec-11-enoic acid (DDA) on nanoparticle
surfaces has often been modeled as a metathesis-type
equilibrium, as described by eq 1.
k1
[Ol]B + [DDA]F XooY [Ol]F + [DDA]B
Scheme 1. Reaction Scheme Showing the Equilibrium
Exchange of Oleate Ligands on InP Particles for Incoming
Terminal Alkene-Labeled Ligands, Including Carboxylic
Acid, Phosphonic Acid, and Thiol
k−1
(1)
[Ol]F[DDA]B
[Ol]B[DDA]F
Keq
=
(2)
This model is indifferent to binding modes of the ligand and
site differentiation on the nanocrystal. The crystal structure of
the In37P20X51 cluster with carboxylate ligands is known and
exhibits exclusively bidentate and bridging−bidentate X-type
binding across a relatively uniform surface. Attempting to model
the exchange as an X-type metathesis and to determine Keq via eq
2 shows a nonlinear trend that can be interpreted as two distinct
equilibrium regimes (Figure 2A). The difference in these
regimes is much greater than would be expected from the
difference in the chemical potential between facets of InP;42
moreover, a nonstoichiometric exchange ratio is seen in the early
regime with substoichiometric ligand displacement. Therefore,
we consider an alternative, neutral L-type binding mode in
addition to the X-type exchange to account for this increase in
coverage.
To model this type of binding, we adopt a modified multisite
Langmuir isotherm model (eq 3) that enables us to account for
multiple species as well as multiple binding modes. Additional
details and derivation are available in section SI 1. Equation 3
represents the binding of a titrated acid (DDA) using a multisite-
competitive Langmuir isotherm, which models the bound
fraction of ligand, θDDA, as a function of the free ligand
concentrations and equilibrium constants. This model allows for
fitting of the fraction of sites belonging to each mode per
nanoparticle, the L-type equilibria of each acid, and the relative
NMR integration and fitting in conjunction with mesitylene as
an internal standard and a highly purified 1.3 nm In37P20X51
cluster as a precise starting point. Per-particle measurements are
normally highly limited by not just the polydispersity but also
the inherent difficulty in determining the nanoparticle
concentrations. Using an atomically defined starting point
eliminates data convolution arising from ensemble measure-
ments and allows for highly precise correlations of per-particle
properties such as the ligand count and density.
X-type binding affinity of each carboxylate, such that KXDDA
/
B
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
we would predict the L-type binding to be monodentate and
concomitant with neighboring shifts of bound carboxylate to a
monodentate binding mode in order to maintain both the
coordination number and charge balance (Scheme 2).
KXDDA[DDA]F
θDDA = χ
x 1 + KXDDA[DDA]F + KXOl[Ol]F
KLDDA[DDA]F
+ χ
l 1 + KLDDA[DDA]F + KLOl[Ol]F
(3)
i
KXDDA[DDA]F
1 + KXDDA[DDA]F + KXOl[Ol]F
j
j
[DDA]B = [MSC]jnX
j
j
k
y
KLDDA[DDA]F
1 + KLDDA[DDA]F + KLOl[Ol]F
z
z
+ nL
z
z
z
(4)
{
Scheme 2. Schematic Representation of L-Type Carboxylic
Acid Coordination with a Concomitant Shift in a Bidentate
Carboxylate to a Monodentate Binding Mode
Figure 2. (A) Metathesis-style plot of Keq via eq 2 for the titration of
DDA into a solution of Ol-capped cluster. Instead of a single slope
corresponding to Keq, empirically two regimes of differing equilibrium
constants are seen. (B) Same data replotted using an isotherm fit via eq
The difference between the L- and X-type binding modes was
further investigated using variable-temperature 1H NMR
spectroscopy (Figure 3 and section SI 2). Because the
4. This fit gives the following values: KXeq = 0.80, KLDDA = 2.3, KLOl
=
1.9, nX = 51, and nL = 8.
KXOl = KXeq. We note that based on the range of calculated
ligand desorption energies for different sites on the In37P20
cluster and the 1:1 nature of this exchange process, we would not
expect to be able to distinguish individual X-type sites.45
Considering that a ligand bound in the X- versus L-type mode
will have a nearly indistinguishable terminal alkene resonance,
we can further rearrange eq 3 into directly measurable
1
concentrations for quantitative H NMR analysis via eq 4.
Using this model, we see a much-improved fit and
experimentally determine the equilibrium constants (Figure
2B). Our experimental value for KXeq of 0.80 is remarkably
similar to that for KXeq reported by Dempsey et al. for a virtually
identical pair of ligands on CdSe QDs,31 suggesting that some
relative binding properties of ligands may translate very well
between nanoparticle systems. This may be especially true for
Cd2+- and In3+-based systems given the similarities in the cation
size and Lewis acidity. While the basicity of carboxylic acids is
well characterized43,44 the L-type equilibria of carboxylic acids
on nanoparticle surfaces have, to our knowledge, never been
quantified. Here we find the equilibrium constants for the L-type
binding of typical aliphatic carboxylates to be on the order of KL
= 2.0 and that L-type binding accounts for approximately 15% of
the total ligation when saturated. Given that L-type binding
would proceed as a Lewis base interaction with a surface In ion,
this ratio would likely decrease with the decreasing concen-
tration of surface In that has been theorized and observed on
larger particles.45 Because these In sites are coordinatively
saturated at the onset through bidentate binding of carboxylate,
Figure 3. Van’t Hoff plots for the temperature-dependent equilibria of
DDA versus Ol at low-concentration (blue) versus high-concentration
(orange) regimes of added DDA.
theoretical number of L-type sites is small because of steric
crowding on the cluster surface and the binding is favorable, the
low free acid concentration regime can be considered to be L-
exchange-dominated, whereas the high free acid regime would
theoretically be X-exchange-dominated and L-type-saturated.
Constructing a Van’t Hoff plot by examining Keq as a function of
the temperature in the high-concentration regime at 44 equiv of
acid (115 mM acid) reveals a nearly iso-Gibbs energy reaction.
This is a reasonable result given that the number of molecules in
the system remains unchanged and that the acids are extremely
similar to an equilibrium constant quite close to 1. By contrast,
examining the low, L-type-dominated regime at 7 equiv (18
mM) reveals weakly positive enthalpy and entropy terms of ΔH°
= 8 kJ/mol and ΔS° = 26 J/(mol K), respectively, in line with
C
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
similar measurements in the literature. These values equate to a
ΔG value of 412 J/mol at room temperature and suggest that L-
type binding is slightly unfavorable, whereas the isotherm fitting
indicated it was slightly favorable. This may be an artifact of X-
type exchange or other rearrangements convoluting the
measurement; however, it would also be unsurprising that
carboxylic acids bind less favorably than traditional Lewis basic
L-type ligands such as amines or phosphines. Positive entropy
terms for ligand-exchange reactions have been observed
previously and assigned by others in the literature as disorder
in the ligand shell, i.e., the entropy of mixing.31,42 Entropy purely
from mixing would, however, be proportional to the molar
fractions of ligands being mixed, and thus we would expect an
order of magnitude lower entropy for the low equivalent
exchange regime, treated as ideal species of just 1.2 J/(mol K).46
We have previously shown crystallographically that carboxylates
have four distinct X-type binding modes on InP that are
predominantly bridging and bidentate. We hypothesize that L-
type carboxylate binding would be associated with local
rearrangement, with neighboring ligands shifting from bidentate
to monodentate in order to alleviate steric hindrance and reveal
additional coordination sites. This denticity rearrangement has
been seen crystallographically during the adsorption of water to
the InP cluster.47 Such a rearrangement could result in a net
increase in the entropy related to the increase in available
microstates of a monodentate species versus a more rigidly
bound bidentate ligand. Such conformational entropies have
been calculated to be on the order of 10 J/(mol K) per degree
denticity in other systems.48
The excitonic feature of the cluster as measured by UV−vis
spectroscopy was tracked as a function of the added titrant and
underwent a slight 20 meV red shift followed by a 60 meV blue
shift over the course of the titration at room temperature
(section SI 3). This change is inconsistent with what has been
seen in the Z-type etching of indium carboxylate from the cluster
by amine.47 Rather, this strongly indicates the influence of the
ligand shell over the excitonic wave function, which is magnified
because of the cluster’s small size. Given the electronic similarity
of oleic acid and DDA this shift is more attributable to changes in
the surface environment as opposed to ligand identity, with
changes in the binding modes and increased total coverage being
the operative differences. As such, this feature is not only tunable
but also highly reversible and a useful metric for conveniently
tracking cluster purification. The 31P NMR spectrum does not
show an appreciable change upon the addition of excess
carboxylic acid (Figure 1, inset). Given that exchange is fast
relative to the NMR time scale and that the net density of oxygen
bonds at the surface does not change, this is not surprising and
supports the stability of the cluster toward excess carboxylic acid.
Carboxylate−Phosphonic Acid Exchange. Phosphonic
acids are widely used in the semiconducting nanocrystal
literature as robust X-type capping ligands. As ligands, they are
characterized by their variable denticity binding modes and
strong, often irreversible, binding affinity.49,50 These properties
have been the driving arguments for synthetic observations such
as anisotropic growth51,52 and shelling inhibition, as well as
influencing the physical properties of nanoparticles through
enhanced thermal, photochemical, and oxidative stability.53−55
In the case of InP, phosphonate capping ligands have been little
explored, and no bottom-up syntheses of InP using phosphonate
ligands have been reported beyond a phosphonate-capped
cluster species.56 This is likely due to the extreme stability that
these ligands impart upon low-molecular-weight, oligomeric,
and cluster intermediates.
Dianionic binding of phosphonates and displacement of
carboxylate are widely reported in the QD literature. This is
readily predictable because phosphonic acids are roughly 3
orders of magnitude more acidic than analogous carboxylic acids
and subsequent bidentate binding is heavily favored by the
chelate effect.57 Assuming the dianionic nature of the
phosphonate ligand, the theoretical point of complete
stoichiometric exchange on the starting Ol-ligated In37P20X51
cluster by adding 10-eneundecylphosphonic acid (UDPA) is
25.5 equiv. Titrations toward this stoichiometric point at room
temperature are associated with a steady blue shift of the
excitonic feature in the UV−vis spectrum across a range of 10
meV, beyond which a decrease in the intensity and a loss of the
lowest energy features and solution color are observed,
suggesting cluster decomposition (section SI 4). The initial
hypsochromic shift is reasonably attributable to what has been
proposed in the literature as ligand effects on excitonic wave
functions as a function of the headgroup electronegativity.58−60
The highest occupied and lowest unoccupied molecular orbitals
of the cluster are calculated to reside close to the surface and can
be influenced by binding agents, as we have shown previously.39
Additionally, the exchange of carboxylate for phosphonate
ligands dictates a significant degree of surface rearrangement,
corroborated by an increase in the cluster symmetry, as seen by
31P NMR spectroscopy (section SI 5). During this room
temperature ligand exchange, however, structural rearrange-
ment is not equivalent to the alternate zincblende cluster
structure observed from bottom-up synthesis at elevated
temperatures. Rather, the In37P20 cluster core is evidently
kinetically stable across the complete range of ligand exchange.
By 1H NMR spectroscopy, we measure the stoichiometry of this
exchange between carboxylate and phosphonate to be 2.10
0.06 carboxylates per phosphonic acid (Figure 4), in strong
agreement with the initial hypothesis of bidentate, irreversible
binding as the phosphonate dianion.
Beyond the stoichiometric point of complete exchange, there
is no appearance of a free phosphonic acid resonance in ether the
1H or 31P{1H} NMR spectra in these experiments. Coupled with
bleaching of the UV−vis spectroscopic features, this suggests
cluster decomposition in the presence of excess acid. The
stoichiometric point of etching to form In2PA3 and PH3 would
be 55.5 equiv of added phosphonic acid; however, we do not
observe a distinct crossover at this point. Etched PH3 accounts
for less than 3% of the phosphorus measured via 31P NMR
spectroscopy, and the continued absence of free phosphonic
acid suggests that InP decomposition by phosphonic acid does
not proceed through primarily PH3 displacement. Instead,
broadening of the bound phosphonate resonances and eventual
solidification of the sample imply the formation of complex
InP(PO3)(PO3H) species. Despite this transformation being as
completely forward driven as ligand exchange, it is only observed
after ligand exchange is complete, prior to which at least 20
unique bound phosphonate resonances are observable in the 31P
NMR spectrum. Oligomeric phosphonate products such as this
are likely what inhibits bottom-up growth of phosphonate-
capped InP nanoparticles.
Carboxylate−Thiol Exchange. While not a general-use
ligand choice for InP synthesis compared to carboxylates, thiols
hold great interest in InP syntheses insofar as they relate to
shelling strategies to access InP@ZnS and related hetero-
structures.61,62 A preliminary treatment of the exchange
D
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Polyexponential decay curves mean that a sum of diffusing
products is being measured, supporting the hypothesis of
particle etching. Using biexponential fitting, three distinct
species are found and can be independently correlated with
the chemical shifts of bound or free regions (Figure 5). These
Figure 4. (A) 1H NMR spectra of the alkene region for the titration of
UDPA into a solution of Ol-capped In37P20 cluster, from 0 equiv (red)
to 51 equiv (violet). Resonances at 5.8−5.9 and 5.0−5.2 ppm
correspond to UDPA, while those at 5.4−5.7 ppm correspond to Ol.
(B) Quantification of the displacement of carboxylate (blue) by
phosphonic acid (orange). The solid black lines correspond to
theoretical 1:1 (UDPA/UDPA, increasing) and 2:1 (DDA/UDPA,
decreasing) stoichiometries, respectively.
1
Figure 5. H DOSY NMR spectra from low gradient (red) to high
gradient (violet), overlaid with corresponding pairs of diffusion
constants acquired from biexponential fitting. The mean diffusion
constants in benzene and proposed assignments are as follows (cm2/s):
1 × 10−5 (red) free acid, 5 × 10−7 (blue) etched species, and 4 × 10−6
(green) cluster.
equilibrium between 10-undecene-1-thiol (UDTh) and Ol
ligands on the cluster using eq 2 results in a two-regime
progression similar to that seen in the carboxylate-exchange
case. Quantifying the exchange by total ligands bound per
particle, however, reveals a trend inconsistent with a mixed L-
and X-type binding model. The low-concentration exchange
species were found to have diffusion constants D = 1 × 10−5, 5 ×
10−7, and 4 × 10−6 cm2/s, which we assign as the free acid/thiol,
cluster, and etched product, respectively. Accounting for solvent
viscosity differences, these values are consistent with the
literature.41,64
The etched product appears to contain both thiolate and
carboxylate. Using the Stokes−Einstein equation, this material
has a hydrodynamic radius of 1.0 nm, approximately 3 times that
of the free acid. Specifically assigning this species is very
challenging because many small oligomers can be drawn with
mixed numbers of carboxylates and thiolates, especially
considering the many bridging modes of carboxylate. No such
decomposition was observed by 1H and 31P{1H} NMR
spectroscopy after exposing a sample of cluster to 16 equiv of
thiol for 12 h, while decomposition was effectively immediate
upon exposure to an excess of 35 equiv of thiol. This observation
is unlike the Z-type displacement by amines on both InP65 and
CdSe,63 where the initial displacement is very rapid and then
subsequently slows, likely speaking to a different displacement
mechanism. Whereas the carboxylate- and phosphonate-
exchange systems were found to be relatively analogous to
those found for CdSe nanoparticles, the case for thiol differs
quite dramatically. There is no evidence to suggest that thiol
shows L-type behavior or that thiolate irreversibly binds on InP.
Additionally, no evidence for the formation of disulfides was
regime exhibits a highly stoichiometric 50
1 total X-type-
bound ligand (section SI 6). Following a rigid one-to-one
exchange rate ratio strongly implies thiol does not L-type-bind
to InP. Additionally, the exchange causes a significant loss of 31P
environment symmetry, as observed by NMR spectroscopy,
which likely signifies surface reconstruction (section SI 7). This
surface rearrangement appears to preclude carboxylate L-type
binding, as is also observed in the case of phosphonate
coordination. Modeled as pure X-type exchange, this regime
follows a thiolate for carboxylate exchange with KXeq = 3.9
through 36 equiv of added thiol, meaning that a favorable
exchange for thiol is seen through approximately 50% starting
ligand substitution. This favorable binding is consistent with
thiolate being a much softer Lewis base than carboxylate and
In3+ being a relatively soft Lewis acid. By contrast, in the high-
concentration regime, a roughly linear increase in the total
number of bound ligands is observed at a ratio of approximately
1 per 6 equiv of thiol added, or equivalently per increase in 0.6
mM thiol.
1
Because there is no observed L-type binding, superstoichio-
metric binding in this second regime can only be explained by
particle etching, exposing more metal atoms either on the
surface or in solution. Similar observations have been made in
the Z-type etching of CdSe nanoparticles by concentrated
found in the H NMR data, which would have a diagnostic
triplet at 2.5 ppm. The etched product of high-concentration
thiol exposure to the InP cluster most closely resembles Z-type
exchange observed in the presence of alcohols in metal
chalcogenide systems.
Carboxylate−Carboxylic Acid Exchange on QDs. The
In37P20X51 cluster has proven to be a robust analogue for InP
QDs in many regards. It possesses an In-rich stoichiometry com-
primary alcohols.63 Using H DOSY NMR spectroscopy, we
1
measure the signal versus gradient decay rates for the bound
species to be clearly nonmonoexponential (section SI 8).
E
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
parable to that established for InP QDs, and is passivated with
the same commonly used ligands. Given the cluster’s greater
curvature and more confined electronic structure, the similarity
of the ligand affinities between cluster and nanoparticle is not
immediately obvious.66 To establish the veracity of using the InP
cluster as a model system for larger InP nanostructures, we have
replicated the carboxylate-exchange experiment on larger InP
QDs. These particles were purified in the same manner as the
cluster and determined to be 3 nm by transmission electron
microscopy analysis (section SI 9). Solution concentrations
were determined via the literature extinction coefficient values,66
and the exchange was carried out over a range similar to that of
the cluster on a percent ligand basis using the same Ol starting
and DDA titrant systems.
colloidal InP, the ligative properties can vary greatly depending
on the ligand identity but that their binding can be reliably and
quantitatively modeled. By using an InP cluster as a molecularly
precise starting point and alkene-labeled ligands, we are able to
quantitatively model the ligand binding dynamics in an
atomically precise manner using 1H NMR spectroscopy. Using
DDA, we demonstrate the broad similarity of the cluster
coordination chemistry to that of larger QDs of InP and CdSe.
By modeling the equilibrium between DDA and the starting Ol
ligands using an isotherm-based approach, we are able to for the
first time deconvolute and quantify carboxylic acid L-type
binding to a nanoparticle surface as accounting for 10−20% of
total ligand binding at saturation. Given the net increase in the
entropy associated with this type of binding and the change in
31P symmetry seen upon the binding of phosphonate and thiol,
we suggest that both X- and L-type exchanges are concomitant
with surface carboxylates shifting from bidentate to mono-
dentate and more significant structural perturbations in the case
of strongly binding ligands. UDTh was found to bind more
strongly than Ol with no tendency toward L-type binding or
disulfide formation, but was observed to cause decomposition
prior to complete exchange. Finally, UDPA was found to bind
irreversibly and at a strict 2:1 stoichiometry versus Ol, as has
been observed in several material systems. This binding was
largely free from decomposition below the complete exchange
limit, beyond which InP phosphonate oligomers began to
develop. Ultimately, we believe that observations and analytical
techniques such as these will underpin future development of
nanoparticle synthesis and technological translation via an
improvement in the rational surface design. Methodological
development in postsynthetic surface modification, including
shelling for improved photophysical properties and ligand
exchange for improved charge transport, will require a detailed
understanding of the relative binding strengths of ligands as well
as ligand decomposition pathways to achieve maximal utility.
The results of this exchange were indeed similar to those seen
in the cluster case. Attempting to model the exchange
mechanism as pure X-type metathesis yielded a nonlinear fit
similar to that of Keq that was much-improved upon the use of
isotherm fitting via eq 3 (Figure 6). The experimental value for
Figure 6. Isotherm fitting via eq 4 of DDA titration into a solution of
Ol-capped InP QDs from 0 to 110 equiv. Fitting gives the following
values: KXeq = 0.88, KLDDA = 2.3, KLOl = 3.5, nX = 88, and nL = 10.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.8b03524.
the X-type equilibrium was measured to be KXeq = 0.88. This
value is in strong agreement with the cluster system. The ratio of
L-to-X-type sites was calculated to be approximately 10% of the
total sites, which is significantly less than the 15% seen in the
cluster case. A decrease in the available surface-In concentration
with increasing particle size has recently been computationally
predicted to be as much as 30% lower for a 3.28 nm nanoparticle
compared to the In37P20X51 cluster.45 This change would
naturally lead to a decrease in the L-to-X-type ligand ratio
because there are fewer sites in general to bind and X-type
coordination takes priority for charge-balance requirements.
The absolute number of bound ligands on a per-particle basis
was also found to be lower than expected. At 98 bound ligands
per particle, there are approximately 25% fewer bound ligands
than literature and geometric considerations would otherwise
predict on a per-particle basis.45,66 This discrepancy could
reasonably be attributed to ligand stripping from over-
purification, as has been observed previously,67,68 or simply an
error arising from the inherent difficulty in quantifying the
nanoparticle concentrations. Overall, the nanoparticle system is
demonstrably similar to the cluster system with nearly
equivalent relative ligand affinities in conjunction with a
predictable decrease in the L-site availability.
■
S
Full synthetic procedures and additional analytical details
(PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: cossairt@uw.edu.
ORCID
Brandi M. Cossairt: 0000-0002-9891-3259
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
We gratefully acknowledge support from the University of
Washington and the National Science Foundation under Grant
CHE-1552164.
REFERENCES
■
(1) Sambur, J. B.; Chen, P. Approaches to Single-Nanoparticle
Catalysis. Annu. Rev. Phys. Chem. 2014, 65, 395−422.
(2) Xia, Y.; Yang, H.; Campbell, C. T. Nanoparticles for Catalysis. Acc.
Chem. Res. 2013, 46, 1671−1672.
CONCLUSIONS
■
Ligand shells are a complex and relatively poorly understood
aspect of colloidal nanostructures. We have shown that, on
F
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(3) Semonin, O. E.; Luther, J. M.; Beard, M. C. Quantum Dots for
Next-Generation Photovoltaics. Mater. Today 2012, 15, 508−515.
(4) Lewis, N. S. Toward Cost-Effective Solar Energy Use. Science
2007, 315, 798−801.
the Composition of the Dot’s Ligand Shell. J. Am. Chem. Soc. 2017, 139,
4246−4249.
(25) Chang, C. M.; Orchard, K. L.; Martindale, B. C. M.; Reisner, E.
Ligand Removal from CdS Quantum Dots for Enhanced Photocatalytic
H 2 Generation in PH Neutral Water. J. Mater. Chem. A 2016, 4, 2856−
2862.
́
(5) Shirasaki, Y.; Supran, G. J.; Bawendi, M. G.; Bulovic, V. Emergence
of Colloidal Quantum-Dot Light-Emitting Technologies. Nat.
Photonics 2013, 7, 13−23.
(6) Gao, X.; Yang, L.; Petros, J. A.; Marshall, F. F.; Simons, J. W.; Nie,
S. In Vivo Molecular and Cellular Imaging with Quantum Dots. Curr.
Opin. Biotechnol. 2005, 16, 63−72.
́
(26) Samadpour, M.; Boix, P. P.; Gimenez, S.; Iraji Zad, A.;
́
Taghavinia, N.; Mora-Sero, I.; Bisquert, J. Fluorine Treatment of
TiO2 for Enhancing Quantum Dot Sensitized Solar Cell Performance. J.
Phys. Chem. C 2011, 115, 14400−14407.
(7) Kairdolf, B. A.; Smith, A. M.; Stokes, T. H.; Wang, M. D.; Young,
A. N.; Nie, S. Semiconductor Quantum Dots for Bioimaging and
Biodiagnostic Applications. Annu. Rev. Anal. Chem. 2013, 6, 143−162.
(8) Hu, M. Z.; Zhu, T. Semiconductor Nanocrystal Quantum Dot
Synthesis Approaches Towards Large-Scale Industrial Production for
Energy Applications. Nanoscale Res. Lett. 2015, 10, 469.
(9) Pu, Y.; Cai, F.; Wang, D.; Wang, J.-X.; Chen, J.-F. Colloidal
Synthesis of Semiconductor Quantum Dots toward Large-Scale
Production: A Review. Ind. Eng. Chem. Res. 2018, 57, 1790−1802.
(10) Bukowski, T. J.; Simmons, J. H. Quantum Dot Research: Current
State and Future Prospects. Crit. Rev. Solid State Mater. Sci. 2002, 27,
119−142.
(11) Johansen, J.; Stobbe, S.; Nikolaev, I. S.; Lund-Hansen, T.;
Kristensen, P. T.; Hvam, J. M.; Vos, W. L.; Lodahl, P. Size Dependence
of the Wavefunction of Self-Assembled InAs Quantum Dots from
Time-Resolved Optical Measurements. Phys. Rev. B: Condens. Matter
Mater. Phys. 2008, 77, 073303.
(27) Shahid, R.; Gorlov, M.; El-Sayed, R.; Toprak, M. S.; Sugunan, A.;
Kloo, L.; Muhammed, M. Microwave Assisted Synthesis of ZnS
Quantum Dots Using Ionic Liquids. Mater. Lett. 2012, 89, 316−319.
(28) Stein, J. L.; Mader, E. A.; Cossairt, B. M. Luminescent InP
Quantum Dots with Tunable Emission by Post-Synthetic Modification
with Lewis Acids. J. Phys. Chem. Lett. 2016, 7, 1315−1320.
(29) Talapin, D. V.; Murray, C. B. PbSe Nanocrystal Solids for N- and
p-Channel Thin Film Field-Effect Transistors. Science 2005, 310, 86−
89.
(30) Wall, M. A.; Cossairt, B. M.; Liu, J. T. C. Reaction-Driven
Nucleation Theory. J. Phys. Chem. C 2018, 122, 9671−9679.
(31) Knauf, R. R.; Lennox, J. C.; Dempsey, J. L. Quantifying Ligand
Exchange Reactions at CdSe Nanocrystal Surfaces. Chem. Mater. 2016,
28, 4762−4770.
(32) Anderson, N. C.; Hendricks, M. P.; Choi, J. J.; Owen, J. S. Ligand
Exchange and the Stoichiometry of Metal Chalcogenide Nanocrystals:
Spectroscopic Observation of Facile Metal-Carboxylate Displacement
and Binding. J. Am. Chem. Soc. 2013, 135, 18536−18548.
(33) Weiss, E. A. Organic Molecules as Tools to Control the Growth,
Surface Structure, and Redox Activity of Colloidal Quantum Dots. Acc.
Chem. Res. 2013, 46, 2607−2615.
(34) Zhou, Y.; Buhro, W. E. Reversible Exchange of L-Type and
Bound-Ion-Pair X-Type Ligation on Cadmium Selenide Quantum
Belts. J. Am. Chem. Soc. 2017, 139, 12887−12890.
(35) Kessler, M. L.; Starr, H. E.; Knauf, R. R.; Rountree, K. J.;
Dempsey, J. L. Exchange Equilibria of Carboxylate-Terminated Ligands
at PbS Nanocrystal Surfaces. Phys. Chem. Chem. Phys. 2018, 20, 23649−
23655.
(36) Moreels, I.; Justo, Y.; De Geyter, B.; Haustraete, K.; Martins, J.
C.; Hens, Z. Size-Tunable, Bright, and Stable PbS Quantum Dots: A
Surface Chemistry Study. ACS Nano 2011, 5, 2004−2012.
(37) Wheeler, L. M.; Sanehira, E. M.; Marshall, A. R.; Schulz, P.; Suri,
M.; Anderson, N. C.; Christians, J. A.; Nordlund, D.; Sokaras, D.; Kroll,
T.; et al. Targeted Ligand-Exchange Chemistry on Cesium Lead Halide
Perovskite Quantum Dots for High-Efficiency Photovoltaics. J. Am.
Chem. Soc. 2018, 140, 10504−10513.
(38) Zhou, L.; Yu, K.; Yang, F.; Cong, H.; Wang, N.; Zheng, J.; Zuo,
Y.; Li, C.; Cheng, B.; Wang, Q. Insight into the Effect of Ligand-
Exchange on Colloidal CsPbBr3 Perovskite Quantum Dot/Mesopo-
rous-TiO2 Composite-Based Photodetectors: Much Faster Electron
Injection. J. Mater. Chem. C 2017, 5 (25), 6224−6233.
(12) Wang, F.; Wang, Y.; Liu, Y.-H.; Morrison, P. J.; Loomis, R. A.;
Buhro, W. E. Two-Dimensional Semiconductor Nanocrystals: Proper-
ties, Templated Formation, and Magic-Size Nanocluster Intermediates.
Acc. Chem. Res. 2015, 48, 13−21.
(13) Jia, G.; Xu, S.; Wang, A. Emerging Strategies for the Synthesis of
Monodisperse Colloidal Semiconductor Quantum Rods. J. Mater.
Chem. C 2015, 3, 8284−8293.
̀
(14) Reiss, P.; Protiere, M.; Li, L. Core/Shell Semiconductor
Nanocrystals. Small 2009, 5, 154−168.
(15) Hines, D. A.; Kamat, P. V. Recent Advances in Quantum Dot
Surface Chemistry. ACS Appl. Mater. Interfaces 2014, 6, 3041−3057.
(16) Owen, J. The Coordination Chemistry of Nanocrystal Surfaces.
Science 2015, 347, 615−616.
(17) Talapin, D. V.; Nelson, J. H.; Shevchenko, E. V.; Aloni, S.;
Sadtler, B.; Alivisatos, A. P. Seeded Growth of Highly Luminescent
CdSe/CdS Nanoheterostructures with Rod and Tetrapod Morpholo-
gies. Nano Lett. 2007, 7, 2951−2959.
(18) Kim, K.; Yoo, D.; Choi, H.; Tamang, S.; Ko, J.-H.; Kim, S.; Kim,
Y.-H.; Jeong, S. Halide−Amine Co-Passivated Indium Phosphide
Colloidal Quantum Dots in Tetrahedral Shape. Angew. Chem., Int. Ed.
2016, 55, 3714−3718.
(19) Park, J.; Kim, S.-W. CuInS2/ZnS Core/Shell Quantum Dots by
Cation Exchange and Their Blue-Shifted Photoluminescence. J. Mater.
Chem. 2011, 21, 3745−3750.
(20) Wang, X.-S.; Dykstra, T. E.; Salvador, M. R.; Manners, I.; Scholes,
G. D.; Winnik, M. A. Surface Passivation of Luminescent Colloidal
Quantum Dots with Poly(Dimethylaminoethyl Methacrylate) through
a Ligand Exchange Process. J. Am. Chem. Soc. 2004, 126, 7784−7785.
(21) Groeneveld, E.; Witteman, L.; Lefferts, M.; Ke, X.; Bals, S.; Van
Tendeloo, G.; de Mello Donega, C. Tailoring ZnSe−CdSe Colloidal
Quantum Dots via Cation Exchange: From Core/Shell to Alloy
Nanocrystals. ACS Nano 2013, 7, 7913−7930.
(22) Petryayeva, E.; Algar, W. R.; Medintz, I. L. Quantum Dots in
Bioanalysis: A Review of Applications across Various Platforms for
Fluorescence Spectroscopy and Imaging. Appl. Spectrosc. 2013, 67,
215−252.
(23) Merg, A. D.; Zhou, Y.; Smith, A. M.; Millstone, J. E.; Rosi, N. L.
Ligand Exchange for Controlling the Surface Chemistry and Properties
of Nanoparticle Superstructures. ChemNanoMat 2017, 3, 745−749.
(24) Zhang, Z.; Edme, K.; Lian, S.; Weiss, E. A. Enhancing the Rate of
Quantum-Dot-Photocatalyzed Carbon−Carbon Coupling by Tuning
(39) Gary, D. C.; Flowers, S. E.; Kaminsky, W.; Petrone, A.; Li, X.;
Cossairt, B. M. Single-Crystal and Electronic Structure of a 1.3 nm
Indium Phosphide Nanocluster. J. Am. Chem. Soc. 2016, 138, 1510−
1513.
(40) De Nolf, K.; Cosseddu, S. M.; Jasieniak, J. J.; Drijvers, E.; Martins,
J. C.; Infante, I.; Hens, Z. Binding and Packing in Two-Component
Colloidal Quantum Dot Ligand Shells: Linear versus Branched
Carboxylates. J. Am. Chem. Soc. 2017, 139, 3456−3464.
(41) De Roo, J.; Yazdani, N.; Drijvers, E.; Lauria, A.; Maes, J.; Owen, J.
S.; Van Driessche, I.; Niederberger, M.; Wood, V.; Martins, J. C.; et al.
Probing Solvent−Ligand Interactions in Colloidal Nanocrystals by the
NMR Line Broadening. Chem. Mater. 2018, 30, 5485−5492.
(42) Cho, E.; Jang, H.; Lee, J.; Jang, E. Modeling on the Size
Dependent Properties of InP Quantum Dots: A Hybrid Functional
Study. Nanotechnology 2013, 24, 215201.
(43) Lee, D. G.; Sadar, M. H. The Basicity of Aliphatic Carboxylic
Acids. Can. J. Chem. 1976, 54, 3464−3469.
G
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
̈
(44) Bohm, S.; Exner, O. Basicity of Carboxylic Acids: Resonance in
(64) White, J. R. Diffusion Coefficients of Fatty Acids and Monobasic
Phosphoric Acids in n -Decane. J. Chem. Phys. 1955, 23, 2247−2251.
(65) Gary, D. C.; Petrone, A.; Li, X.; Cossairt, B. M. Investigating the
Role of Amine in InP Nanocrystal Synthesis: Destabilizing Cluster
Intermediates by Z-Type Ligand Displacement. Chem. Commun. 2017,
53, 161−164.
the Cation and Substituent Effects. New J. Chem. 2005, 29, 336−342.
(45) Zhao, Q.; Kulik, H. J. Electronic Structure Origins of Surface-
Dependent Growth in III−V Quantum Dots. Chem. Mater. 2018, 30,
7154−7165.
(46) Denbigh, K. The Principles of Chemical Equilibrium, 3rd ed.;
Cambridge University Press, 1971.
́
(66) Xie, L.; Shen, Y.; Franke, D.; Sebastian, V.; Bawendi, M. G.;
Jensen, K. F. Characterization of Indium Phosphide Quantum Dot
Growth Intermediates Using MALDI-TOF Mass Spectrometry. J. Am.
Chem. Soc. 2016, 138, 13469−13472.
(67) Roberge, A.; Stein, J. L.; Shen, Y.; Cossairt, B. M.; Greytak, A. B.
Purification and In Situ Ligand Exchange of Metal-Carboxylate-Treated
Fluorescent InP Quantum Dots via Gel Permeation Chromatography.
J. Phys. Chem. Lett. 2017, 8, 4055−4060.
(68) Doris, S. E.; Lynch, J. J.; Li, C.; Wills, A. W.; Urban, J. J.; Helms, B.
A. Mechanistic Insight into the Formation of Cationic Naked
Nanocrystals Generated under Equilibrium Control. J. Am. Chem.
Soc. 2014, 136, 15702−15710.
(47) Gary, D. C.; Petrone, A.; Li, X.; Cossairt, B. M. Investigating the
Role of Amine in InP Nanocrystal Synthesis: Destabilizing Cluster
Intermediates by Z-Type Ligand Displacement. Chem. Commun. 2017,
53, 161−164.
(48) Gaberle, J.; Gao, D. Z.; Watkins, M. B.; Shluger, A. L. Calculating
the Entropy Loss on Adsorption of Organic Molecules at Insulating
Surfaces. J. Phys. Chem. C 2016, 120, 3913−3921.
(49) Wagner, G. W.; Fry, R. A. Observation of Distinct Surface Al IV
Sites and Phosphonate Binding Modes in γ-Alumina and Concrete by
High-Field 27Al and 31P MAS NMR. J. Phys. Chem. C 2009, 113,
13352−13357.
(50) Owen, J. S.; Park, J.; Trudeau, P.-E.; Alivisatos, A. P. Reaction
Chemistry and Ligand Exchange at Cadmium−Selenide Nanocrystal
Surfaces. J. Am. Chem. Soc. 2008, 130, 12279−12281.
(51) Peng, Z. A.; Peng, X. Mechanisms of the Shape Evolution of
CdSe Nanocrystals. J. Am. Chem. Soc. 2001, 123, 1389−1395.
(52) Wang, F.; Tang, R.; Kao, J. L.-F.; Dingman, S. D.; Buhro, W. E.
Spectroscopic Identification of Tri- n -Octylphosphine Oxide (TOPO)
Impurities and Elucidation of Their Roles in Cadmium Selenide
Quantum-Wire Growth. J. Am. Chem. Soc. 2009, 131, 4983−4994.
(53) Prodanov, M. F.; Pogorelova, N. V.; Kryshtal, A. P.; Klymchenko,
A. S.; Mely, Y.; Semynozhenko, V. P.; Krivoshey, A. I.; Reznikov, Y. A.;
Yarmolenko, S. N.; Goodby, J. W.; Vashchenko, V. V. Thermodynami-
cally Stable Dispersions of Quantum Dots in a Nematic Liquid Crystal.
Langmuir 2013, 29, 9301−9309.
(54) Tang, J.; Brzozowski, L.; Barkhouse, D. A. R.; Wang, X.; Debnath,
R.; Wolowiec, R.; Palmiano, E.; Levina, L.; Pattantyus-Abraham, A. G.;
Jamakosmanovic, D.; Sargent, E. H. Quantum Dot Photovoltaics in the
Extreme Quantum Confinement Regime: The Surface-Chemical
Origins of Exceptional Air- and Light-Stability. ACS Nano 2010, 4,
869−878.
(55) Xuan, T.; Yang, X.; Lou, S.; Huang, J.; Liu, Y.; Yu, J.; Li, H.;
Wong, K.-L.; Wang, C.; Wang, J. Highly Stable CsPbBr3 Quantum
Dots Coated with Alkyl Phosphate for White Light-Emitting Diodes.
Nanoscale 2017, 9, 15286−15290.
(56) Gary, D. C.; Terban, M. W.; Billinge, S. J. L.; Cossairt, B. M. Two-
Step Nucleation and Growth of InP Quantum Dots via Magic-Sized
Cluster Intermediates. Chem. Mater. 2015, 27, 1432−1441.
(57) Franz, R. G. Comparisons of PKa and Log P Values of Some
Carboxylic and Phosphonic Acids: Synthesis and Measurement. AAPS
PharmSci 2001, 3, 1−13.
̈
̈
(58) Kroupa, D. M.; Voros, M.; Brawand, N. P.; McNichols, B. W.;
Miller, E. M.; Gu, J.; Nozik, A. J.; Sellinger, A.; Galli, G.; Beard, M. C.
Tuning Colloidal Quantum Dot Band Edge Positions through
Solution-Phase Surface Chemistry Modification. Nat. Commun. 2017,
8, 15257.
(59) Frederick, M. T.; Weiss, E. A. Relaxation of Exciton Confinement
in CdSe Quantum Dots by Modification with a Conjugated
Dithiocarbamate Ligand. ACS Nano 2010, 4, 3195−3200.
(60) Giansante, C. Surface Chemistry Control of Colloidal Quantum
Dot Band Gap. J. Phys. Chem. C 2018, 122, 18110−18116.
(61) Bang, E.; Choi, Y.; Cho, J.; Suh, Y.-H.; Ban, H. W.; Son, J. S.; Park,
J. Large-Scale Synthesis of Highly Luminescent InP@ZnS Quantum
Dots Using Elemental Phosphorus Precursor. Chem. Mater. 2017, 29,
4236−4243.
(62) Ryu, E.; Kim, S.; Jang, E.; Jun, S.; Jang, H.; Kim, B.; Kim, S.-W.
Step-Wise Synthesis of InP/ZnS Core−Shell Quantum Dots and the
Role of Zinc Acetate. Chem. Mater. 2009, 21, 573−575.
(63) Anderson, N. C.; Hendricks, M. P.; Choi, J. J.; Owen, J. S. Ligand
Exchange and the Stoichiometry of Metal Chalcogenide Nanocrystals:
Spectroscopic Observation of Facile Metal-Carboxylate Displacement
and Binding. J. Am. Chem. Soc. 2013, 135, 18536−18548.
H
DOI: 10.1021/acs.inorgchem.8b03524
Inorg. Chem. XXXX, XXX, XXX−XXX