Precursor reaction kinetics control compositional grading and size of CdSe1-: XSx nanocrystal heterostructures

Article abstract of DOI:10.1039/c9sc00989b

We report a method to control the composition and microstructure of CdSe_1-xS_x nanocrystals by the simultaneous injection of sulfide and selenide precursors into a solution of cadmium oleate and oleic acid at 240 °C. Pairs of substituted thio- and selenoureas were selected from a library of compounds with conversion reaction reactivity exponents (k_E) spanning 1.3 × 10^-5 s^-1 to 2.0 × 10^-1 s^-1. Depending on the relative reactivity (k_Se/k_S), core/shell and alloyed architectures were obtained. Growth of a thick outer CdS shell using a syringe pump method provides gram quantities of brightly photoluminescent quantum dots (PLQY = 67 to 90%) in a single reaction vessel. Kinetics simulations predict that relative precursor reactivity ratios of less than 10 result in alloyed compositions, while larger reactivity differences lead to abrupt interfaces. CdSe_1-xS_x alloys (k_Se/k_S = 2.4) display two longitudinal optical phonon modes with composition dependent frequencies characteristic of the alloy microstructure. When one precursor is more reactive than the other, its conversion reactivity and mole fraction control the number of nuclei, the final nanocrystal size at full conversion, and the elemental composition. The utility of controlled reactivity for adjusting alloy microstructure is discussed.

Full text of DOI:10.1039/c9sc00989b

Chemical
Science
View Article Online
View Journal
EDGE ARTICLE
Precursor reaction kinetics control compositional
grading and size of CdSe1ꢀxSx nanocrystal
heterostructures†
Cite this: DOI: 10.1039/c9sc00989b
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
Leslie S. Hamachi, a Haoran Yang,b Ilan Jen-La Plante,a Natalie Saenz,a Kevin Qian,a
Michael P. Campos,a Gregory T. Cleveland,a Iva Rreza,a Aisha Oza,c
d
b
de
Willem Walravens,
Andrew C. Crowther
Emory M. Chan,
*
Zeger Hens,
c
a
*
and Jonathan S. Owen
We report a method to control the composition and microstructure of CdSe1ꢀxSx nanocrystals by the
simultaneous injection of sulfide and selenide precursors into a solution of cadmium oleate and oleic
acid at 240 ꢁC. Pairs of substituted thio- and selenoureas were selected from a library of compounds
with conversion reaction reactivity exponents (kE) spanning 1.3 ꢂ 10ꢀ5 sꢀ1 to 2.0 ꢂ 10ꢀ1
s
ꢀ1. Depending
on the relative reactivity (kSe/kS), core/shell and alloyed architectures were obtained. Growth of a thick
outer CdS shell using a syringe pump method provides gram quantities of brightly photoluminescent
quantum dots (PLQY ¼ 67 to 90%) in a single reaction vessel. Kinetics simulations predict that relative
precursor reactivity ratios of less than 10 result in alloyed compositions, while larger reactivity differences
lead to abrupt interfaces. CdSe1ꢀxSx alloys (kSe/kS ¼ 2.4) display two longitudinal optical phonon modes
with composition dependent frequencies characteristic of the alloy microstructure. When one precursor
is more reactive than the other, its conversion reactivity and mole fraction control the number of nuclei,
the final nanocrystal size at full conversion, and the elemental composition. The utility of controlled
reactivity for adjusting alloy microstructure is discussed.
Received 26th February 2019
Accepted 14th May 2019
DOI: 10.1039/c9sc00989b
rsc.li/chemical-science
By injecting a mixture of sulde and selenide precursors with
carefully selected conversion reactivity, one can, in principle,
Introduction
adjust the solute evolution during growth and the resultant
alloy microstructure (Scheme 1). Such an approach could
reduce the number of steps required to prepare CdSe/CdS het-
erostructures, nanocrystals that are highly desirable lumines-
cent downconverters for solid-state lighting, electronic displays,
and biological imaging applications.16–20 In particular, grading
the interface between CdSe and CdS is thought to reduce the
rate of Auger recombination,21 a multi-exciton recombination
process that is sensitive to the microstructure of quantum dots
Molecular precursors with tailored reactivity can precisely
control the rate of solute supply and the extent of nucleation
during colloidal crystallizations.1–4 Recent examples illustrate
this principle using libraries of chalcogenone derivatives to
synthesize CdS, CdSe, PbS, and PbSe nanocrystals.5–8 By
choosing different substitution patterns, one can control the
reactivity, the extent of nucleation, and the size following
quantitative precursor conversion.5–7 Such a method avoids the
need to terminate the precursor reaction prematurely, modify
the surfactant mixture, or adjust the reaction temperature to
prepare a desired size, which can inadvertently change the
surface chemistry and composition of the nal product.9–15
aDepartment of Chemistry, Columbia University, New York, New York 10027, USA.
E-mail: jso2115@columbia.edu
bThe Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, CA 94720,
USA
cDepartment of Chemistry, Barnard College, New York, New York 10027, USA. E-mail:
acrowthe@barnard.edu
dPhysics and Chemistry of Nanostructures Group (PCN), Ghent University, B-9000
Ghent, Belgium
eCenter of Nano and Biophotonics, Ghent University, B-9000 Ghent, Belgium
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9sc00989b
Scheme 1
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
and of great importance to quantum dot lasers and down con- relative reactivity and the mole fraction of added precursor
verting materials for solid state lighting.22–30 However, recent systematically inuence the elemental distribution of CdSe/CdS
theoretical and experimental studies on the link between heterostructures and alloys, as well as their nal crystal size.
graded alloys and Auger kinetics do not agree on the magnitude
of this effect.21,27,30–32 Greater control over the compositional
grading between interfaces could therefore enable more
Results
N-Monosubstituted and N,N0-disubstituted imidazolidine
selones (Se-Im(R1,R2) R ¼ H, Me, Et, iPr, t-Bu, Ph) and pyrimi-
dine selones (Se-Pym(R1,R2) R ¼ H, Me, Et, iPr) were used to
synthesize CdSe nanocrystals (Scheme 2, Fig. 1 and S1†). Se-Im
and Se-Pym can be prepared in a single synthetic step by
reuxing diamines with triethyl orthoformate and elemental
selenium.56 With the exception of N,N0-di-tert-butyl derivatives
and Se-Pym(R1,R2) compounds, > 50% yields of analytically pure
material can be isolated following recrystallization from ethyl
acetate or acetonitrile.
The corresponding thiones (S-Im and S-Pym) are synthesized
from the N,N0-substituted diamines and carbon disulde, thi-
ophosgene,57–59 or thiocarbonyldiimidazole.59,60 To avoid toxicity
concerns, all thiones in this study were prepared from thio-
carbonyldiimidazole. 1H NMR studies show that the second
C–N bond forming step is inhibited by tert-butyl substituents,
which can lead to oligomerization rather than cyclization.
Similarly, syntheses of S-Pym from N,N0-disubstituted-1,3-
diaminopropanes provided lower yields and must be conduct-
ed under dilute conditions to prevent oligomerization.58
To improve the reproducibility and convenience of the
nanocrystal synthesis, 50 gram quantities of cadmium oleate
were isolated from cadmium triuoroacetate, triethylamine,
and oleic acid in acetonitrile and stored in a nitrogen lled
glove box prior to use. In this way, water and/or acetic acid
generated during the synthesis of cadmium carboxylates from
cadmium oxide or cadmium acetate can be avoided.61 Tetra-
glyme solutions of chalcogenoureas were injected into a solu-
tion of oleic acid and cadmium oleate at 240 ꢁC, producing CdS
and CdSe nanocrystals over the course of minutes to hours
depending on the precursor structure (Fig. 1).
systematic investigations of multi-exciton photophysics.
CdSe/CdS heterostructures are typically synthesized using
layer-by-layer, seeded growth, or one-pot methods where
a mixture of sulfur and selenium precursors are coinjected.
Layer-by-layer methods provide precise control over composi-
tion by slowly injecting precursors via syringe pump22,27,30 or via
successive ion layer adsorption and reaction (SILAR),26,33 but
these methods require extensive reaction times and/or precise
half monolayer additions to avoid separate nucleation of the
shell material.22,24,34,35 Heterostructured and alloyed nano-
particles of CdSe1ꢀxSx,25,36–41 CdTe1ꢀxSx,42 CdTe1ꢀxSex,43,44
45
PbSe1ꢀxSx, PbTe1ꢀxSx, and PbTe1ꢀxSex are also prepared in
a single step by coinjection of phosphine sulde and sele-
nides,25,36–38,43–45
bis(trimethylsilyl)chalcogenides,45
and
elemental chalcogen precursors.39–41 While it is possible to
adjust the luminescence and absorption properties by tuning
the ratio of sulde and selenide precursors, it is unclear how
these changes inuence the elemental distribution or size of the
nal product, nor the role of alloy microstructure on photo-
physical properties.
Finally, the microstructure of heterostructured and alloyed
nanoparticles is difficult to probe with a resolution higher than
5 nm, which is on the order of the shell and core thickness of
typical colloidal quantum dots. Scanning transmission electron
microscopy coupled with energy dispersive spectroscopy
(STEM-EDX) or high angle annular dark eld (HAADF STEM)
imaging are well suited to determining the elemental distribu-
tion in large (d $ 30 nm) particles,27,46,47 but elemental mapping
of small particles is much more challenging,48 and effects from
curvature, anisotropic shell growth, and intermixing are
convoluted. Although powder X-ray diffraction is also used to
characterize such samples,49 Scherrer line broadening and the
presence of stacking faults makes it difficult to differentiate
between compositions. Raman spectroscopy can probe the
composition of CdS1ꢀxSex alloys, although there are conicting
claims regarding the number of characteristic Raman modes,
and it is difficult to distinguish alloy and core/shell quantum
dots on the basis of Raman spectroscopy alone.39,50,51 X-ray
photoelectron spectroscopy (XPS), photoemission spectros-
copy, energy dispersive X-ray spectroscopy (EDX), Rutherford
backscattering, and solid-state 113Cd and 77Se nuclear magnetic
resonance (NMR) spectroscopy have all been used to monitor
the radial evolution of nanocrystals during growth, but these
techniques do not address effects from anisotropy.29,39,40,51–55
These limitations obscure the study of alloy microstructure and
how it is controlled by precursor reactivity.
Reaction kinetics are monitored by diluting aliquots to
a
known concentration before analysis with UV-visible
To gain more precise control over the microstructure and
photophysical performance of luminescent CdSe/CdS quantum
dots, we synthesized new chalcogenourea precursors, measured
their relative reactivities, and explored the one-pot synthesis of
alloys and heterostructures. Our results demonstrate how the
Scheme 2
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
replacing methyl with isopropyl causes an order of magnitude
reduction in reactivity. An exception is Se-Im(t-Bu2), which is
anomalously reactive and likely follows a different conversion
mechanism. In many cases, N-alkyl-substituted precursors are
less reactive than N-aryl-substituted precursors. Pyrimidine
chalcogenones are typically less reactive than the imidazolidine
chalcogenones, perhaps because their substituents are more
directed toward the chalcogen atom and have a greater steric
inuence. All of these substituent effects can be explained by
a mechanism involving Lewis acidic activation of the chalco-
genourea prior to cleavage of the C]E bond, a mechanism
analogous to the one proposed for phosphine chalcogenides
and thiocarbonates.7,63
If the C]E bond cleavage reaction is preceded by a fast
preequilibrium Lewis acid activation, one precursor could, in
principle, reduce the available cadmium and the conversion
kinetics of the other. In one example studied here the conver-
sion kinetics of a mixed precursor synthesis was monitored
using 1H NMR spectroscopy and compared to the kinetics of the
single component reactions (see ESI†). In that case, the
conversion reactivity of the chalcogenoureas appears orthog-
onal, i.e. it is not altered by the presence of a second precursor.
A complete analysis of these effects requires a detailed study
beyond the scope of this report, and the role of the conversion
reaction mechanism or the orthogonality of the precursor
reactions is not considered further below.
With a range of precursor reactivity in hand, we explored the
synthesis of CdSe1ꢀxSx nanocrystals using a mixture of two
precursors with appropriately matched reactivity. We sought to
evaluate whether the relative reactivity could be used to dictate
the microstructure according to Scheme 1. When a fast sele-
nium precursor and a much slower sulfur precursor (N-butyl-N0-
pyrrolidineselenourea and S-Im(Me2), kSe/kS ꢃ 1000) are coin-
jected, a CdSe nanocrystal nucleates and grows to 2.2 nm, as can
be observed in the UV-vis absorption spectra of timed aliquots.
Aer growth of the CdSe core, CdS deposition increases the
absorbance at high energy (<512 nm) and red shis the
absorption onset (Fig. 2A). At the same time, the photo-
luminescence quantum yield (PLQY) increases, signaling that
CdS is depositing on the CdSe core. No evidence for the
nucleation of pure phase CdS was obtained under these
conditions.
Fig. 1 (A) Synthesis of cadmium chalcogenide nanocrystals from (B) S-
Im(Ph2) and (C) Se-Im(Ph2). Comparison of mass balance and kinetics
as measured by UV-vis absorption and H NMR spectroscopies. Error
1
bars correspond to 10% error in NMR measurement.
absorption and NMR spectroscopy. The disappearance of the
precursors and the formation of the corresponding urea and
carboxylic anhydride coproducts4–7 match the yield of CdSe62
and CdS7 measured with absorption spectroscopy (Fig. 1). This
observation allows the precursor conversion kinetics to be
estimated by monitoring nanoparticle formation using UV-vis
spectroscopy. A similar strategy has been used in our previous
studies of CdS,7 CdSe,4 PbS,5 and PbSe.6
A single exponential t of the conversion and yield are used
to extract a reactivity exponent (kE) for each precursor (sꢀ1
)
(Table 1). These are distinct from a rst order conversion
reaction rate constant because the kinetics experiments are not
run under pseudo rst order conditions and appear to follow
higher order kinetics. Regardless, the exponents provide a way
to conveniently order the relative conversion reactivity over
several orders of magnitude under a standard set of conditions.
Five orders of magnitude in reactivity are apparent, with many
pairs of sulfur and selenium precursors within an order of
magnitude of each other. A few of the selenourea precursors
produce a partial yield of CdSe, and an initial rate was used to
estimate the relative reactivity (see ESI†). These include Se-
Im(H,Ph), Se-Im(H,Et), Se-Im(H,iPr), Se-Im(Me2), Se-Im(t-Bu2),
Se-Pym(H,Me), Se-Pym(Me2), and Se-Pym(Et2), which typically
produce between 50–75% yields. An explanation for the low
yield was not pursued, and with the exception of Se-Im(t-Bu2)
these selenoureas are not used further below.
The inverse architecture is obtained when a reactive sulfur
precursor is paired with a slower selenium precursor (N-
dodecyl-N0-hexylthiourea and Se-Im(Et2), kSe/kS ꢃ 0.001). CdS
rapidly nucleates upon which CdSe deposits over an hour,
shiing the absorption edge into the visible. The expected CdS/
CdSe core/shell heterostructure is supported by the evolution of
the spectrum during growth (Fig. 2B), and a long photo-
luminescence decay time and low PLQY (ESI†) that is consistent
with a quasi-Type II band structure. Unfortunately, in these
examples, the nanocrystals were too small to characterize using
STEM EDX.
Several reactivity trends emerge from the range of kE. Cyclic
precursors are substantially less reactive than acyclic structures
with the same substituents. For example, S-Im(Me2) is less
reactive than tetramethylthiourea by an order of magnitude.
Sterically encumbered precursors are also less reactive;
In order to synthesize larger CdSe/CdS nanocrystals suitable
for elemental mapping, Se-Im(t-Bu2) and S-Im(Me2) (kSe/kS
100) are co-injected (see ESI†). The spectral evolution of
ꢃ
aliquots, and the photoluminescence quantum yield (60%) and
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Table 1 Chalcogenourea precursors and their reactivity
Compound
E
R1
R2
kEa (sꢀ1
)
Initial rateb (mM sꢀ1
)
% Yielda [CdE]
S-Im(H,Ph)
S-Im(H,Me)
S-Im(H,Et)
S-Im(Ph2)
S-Pym(H,Me)
S-Im(H,iPr)
S-Im(Me2)
S-Pym(Me2)
S-Im(Et2)
S-Pym(Et2)
S-Pym(H,iPr)
S-Im(iPr2)
S-Pym(iPr2)
Se-Im(H,Ph)
Se-Im(H,Et)
Se-Im(H,iPr)
Se-Im(Ph2)
Se-Pym(H,Me)
Se-Im(t-Bu2)
Se-Im(Me2)
Se-Pym(Me2)
Se-Im(Et2)
S
S
S
S
S
S
S
S
S
S
S
S
H
H
H
Ph
H
Ph
Me
Et
Ph
Me
iPr
Me
Me
Et
2.5 ꢂ 10ꢀ3
1.2 ꢂ 10ꢀ3
6.6 ꢂ 10ꢀ4
2.2 ꢂ 10ꢀ4
1.6 ꢂ 10ꢀ4
1.3 ꢂ 10ꢀ4
6.5 ꢂ 10ꢀ5
5.5 ꢂ 10ꢀ5
2.2 ꢂ 10ꢀ5
1.8 ꢂ 10ꢀ5
1.6 ꢂ 10ꢀ5
1.5 ꢂ 10ꢀ5
1.3 ꢂ 10ꢀ5
2.0 ꢂ 10ꢀ1
6.3 ꢂ 10ꢀ2
4.0 ꢂ 10ꢀ2
2.0 ꢂ 10ꢀ2
1.5 ꢂ 10ꢀ2
7.4 ꢂ 10ꢀ3
3.3 ꢂ 10ꢀ3
2.2 ꢂ 10ꢀ3
1.3 ꢂ 10ꢀ3
1.0 ꢂ 10ꢀ3
9.5 ꢂ 10ꢀ4
2.1 ꢂ 10ꢀ4
2.1 ꢂ 10ꢀ2
1.3 ꢂ 10ꢀ2
5.6 ꢂ 10ꢀ3
3.7 ꢂ 10ꢀ3
2.1 ꢂ 10ꢀ3
2.0 ꢂ 10ꢀ3
3.9 ꢂ 10ꢀ4
9.9 ꢂ 10ꢀ4
2.9 ꢂ 10ꢀ4
3.4 ꢂ 10ꢀ4
6.0 ꢂ 10ꢀ4
6.6 ꢂ 10ꢀ4
1.5 ꢂ 10ꢀ4
3.7 ꢂ 10ꢀ1
3.5 ꢂ 10ꢀ1
1.8 ꢂ 10ꢀ1
2.9 ꢂ 10ꢀ1
8.0 ꢂ 10ꢀ2
4.1 ꢂ 10ꢀ2
2.6 ꢂ 10ꢀ2
3.3 ꢂ 10ꢀ2
1.1 ꢂ 10ꢀ2
1.1 ꢂ 10ꢀ2
1.2 ꢂ 10ꢀ2
4.0 ꢂ 10ꢀ3
93
100
100
98
100
100
100
100
100
100
96
100
100
44
79
64
98
75
53
82
81
77
70
98
H
Me
Me
Et
Et
H
iPr
iPr
H
H
H
Ph
H
t-Bu
Me
Me
Et
Et
iPr
iPr
Et
iPr
iPr
iPr
Ph
Et
iPr
Ph
Me
t-Bu
Me
Me
Et
S
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se-Pym(Et2)
Se-Pym(iPr2)
Se-Im(iPr2)
Et
iPr
iPr
98
a
Kinetics measured by UV-Vis absorption under standard reaction conditions: chalcogenourea (10 mM), cadmium oleate (12 mM), oleic acid (24
mM) at 240 ꢁC. Uncertainty (25%) in the rate constant is estimated from the uncertainty in the extinction coefficient (10%) described in the ESI.
b
Initial rate estimated from the slope of the rst two UV-vis absorption aliquots (see ESI for examples).
photoluminescence lifetime (23 ns) are characteristic of a CdSe/ the syringe pump addition, greatly facilitates the preparation of
CdS core/shell architecture. Scanning transmission electron multi-gram quantities of brightly luminescent materials.
microscopy and energy dispersive X-ray spectroscopy (STEM
In the examples above, the ux of solute during nucleation is
EDX) support the expected CdSe/CdS microstructure although dominated by the more reactive precursor. Provided the
the CdS shell and the CdSe core regions are anisotropically precursors are orthogonal to the crystallization, the solute
distributed (Fig. 2C). Despite the anisotropy of the shell growth, supplied by the more reactive precursor will dictate the number
the observed microstructure supports the claim that the more of nuclei and the nal size. For example, when several different
reactive precursor forms the nanocrystal core.
selenium precursors are used together with a less reactive
To access heterostructures desirable for photoluminescence thiourea coreactant (kE ¼ 10ꢀ1 to 10ꢀ3 sꢀ1: N-butyl-N0-pyrroli-
applications, a method was developed to shell the CdSe1ꢀxSx dine selenourea, Se-Im(Ph2), or Se-Im(t-Bu2); vs. kE ¼ 10ꢀ4 to
nanocrystals described above with CdS. The low reactivity of 10ꢀ5 sꢀ1: tetramethylthiourea or S-Im(Me2)) nanocrystals with
tetraalkylthioureas toward cadmium oleate at room tempera- a range of emission wavelengths are obtained (lmax ¼ 515–622
ture allows their mixture to be slowly added to a solution of nm) including blue-green emitters that are otherwise chal-
cores using a syringe pump. Dropwise addition at 240 ꢁC lenging to synthesize (Fig. 4). TEM analysis shows that the
produces nanocrystals up to 6 nm in diameter. Attempts to grow nanocrystal size (d ¼ 3.5–5.7 nm) tracks inversely with the
the nanocrystals beyond this diameter caused precipitation reactivity of the selenium precursor. Given that the mole frac-
during growth. However, upon adding tri-n-octylphosphine and tion of the precursors is xed across these experiments,
2-hexyldecanoic acid to the shelling solution and increasing the (33 : 66), we conclude that the shi in wavelength is a conse-
temperature to 290 ꢁC, even thicker shells could be grown quence of the change in size. The difference in nal volume
without precipitation of the nanocrystals. Using this method, corresponds to a ꢃ4-fold change in the number of nuclei,
CdS/CdSe/CdS spherical quantum wells24 with PLQY up to 90% behavior that is consistent with the extent of nucleation of pure
could be grown in a single pot (Scheme 1) (Fig. 3). Photo- phase CdSe cores discussed below (see below).
luminescence decay times (30–90 ns) increase as the CdS shell
Adjusting the ratio of sulfur and selenium precursors to tune
thickens. Much like the CdSe/CdS core/shell structures above, the composition and emission wavelength can also inuence
the nal shape appears to be bipyramidal with the selenium off- the rate of solute generation and the extent of nucleation. For
center.64–66 The nal nanocrystals are primarily composed of example, the emission wavelength of CdS/CdSe core/shell
CdS in the outer shell, which combined with the ease of scaling nanocrystals could be tuned from lmax ¼ 472 to 513 nm by
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
Fig. 3 (A) UV-vis and photoluminescence spectra, (B) TEM of CdS/
CdSe/CdS spherical quantum wells whose synthesis is described in the
ESI.† Scale bar is 10 nm in length.
size, which is unfortunately too small to accurately measure in
the examples shown in Fig. 5.
To understand how adjusting the amount of the more reac-
tive precursor dictates the extent of nucleation and the nal
size, pure phase CdS and CdSe nanocrystals were prepared by
adjusting the amount of added chalcogenourea (0.1–1 equiv.: N-
hexyl-N0-dodecylthiourea, N-hexyl-N0,N0-dibutylthiourea) or Se-
Im(Ph2) under otherwise identical conditions (see ESI†). As the
amount of injected chalcogenourea increases, the number of
nuclei increases linearly, a change that will increase the number
of core nanocrystals on which to grow shell material. Interest-
ingly, nucleation is not observed below a threshold concentra-
tion of injected precursor (thiourea ꢃ1 mM, selenourea ꢃ0.1
mM), which may reect a solubility limit of the monomers that
is greater for CdS than CdSe. In the synthesis of CdS/CdSe core/
shell nanocrystals shown in Fig. 5, increasing the amount of N-
hexyl-N0-dodecylthiourea should therefore reduce the nal size
and enhance the blue shi for quantum connement reasons.
Fig. 2 Example UV-vis and photoluminescence spectra of (A) CdSe/
CdS core/shell and (B) CdS/CdSe core/shell nanocrystals synthesized
from (A) N-butyl-N0-pyrrolidine selenourea and S-Im(Me2) and (B) N-
dodecyl-N0-hexylthiourea and Se-Im(Et2). (C) STEM EDX elemental
maps of nanocrystals synthesized from Se-Im(t-Bu2) and S-Im(Me2)
(blue ¼ cadmium, red ¼ selenium, green ¼ sulfur.) STEM EDX line scan
collected in area denoted by dashed white box. Scale bars are 4 nm in
length.
reducing the amount of N-hexyl-N0-dodecylthiourea co-injected
with Se-Im(Et2) (kSe/kS ꢃ 0.001) from 0.5–0.2 equivalents
(Fig. 5). Consistent with the compositional variation, the lumi-
nescence color red-shis with the increased mole fraction of
selenium. However, the shi can also be affected by the nal
Fig. 4 Size and photoluminescence wavelength tunability of CdSe/
CdS core/shell nanocrystals made using N-butyl-N0-pyrrolidine sele-
nourea (A, D), Se-Im(Ph2) (B, E), and Se-Im(t-Bu2) (C, F) paired with
a slower sulfur precursor S-Im(Me2). Note that the yield of Se-Im(t-
Bu2) reaches 50%, an effect that decreases the Se : S ratio of the final
structure, and will increase the thickness of the CdS shell. These two
effects have opposite effects on the band gap. Note that the reaction
with N-butyl-N0-pyrrolidine is done at a lower concentration to avoid
mixing limited kinetics.
Scheme 3
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
elsewhere). The sizes and lattice constants of these materials
smoothly span the range of sizes produced in single component
reactions, and follow Vegard's law (e.g. CdS: d ¼ 2.1 nm; a ¼ 5.81
˚
˚
A, and CdSe: d ¼ 3.0 nm; a ¼ 6.09 A). These data are consistent
with the formation of an alloyed composition, and indicate that
the extent of nucleation from the mixed solute is similar to
behavior of a pure phase synthesis. We conclude that the extent
of nucleation is similar to the pure phase and core–shell
nanocrystals and is controlled by the rate of solute generation
during nucleation.
The closely matched precursor reactivity intermixes the CdS
and CdSe as can be seen in Raman spectra acquired at 77 K
(Fig. 6B). Two vibrational bands at ꢃ200 cmꢀ1 and ꢃ300 cmꢀ1
are assigned to CdSe and CdS phonon modes, respectively. Each
band is t to two peaks, a more intense longitudinal optical (LO)
phonon with a lower energy surface optical (SO) phonon
shoulder (see ESI† for more detail on tting methodology and
peak assignments). The LO peaks red-shi from the frequency
of pure CdSe or CdS as its mole fraction is reduced, consistent
with the literature on bulk67 and nanoscale51 CdSe1ꢀxSx alloys.
The presence of two Raman LO peaks is characteristic of
a CdSe1ꢀxSx alloy because the mass of sulfur is less than the
reduced mass of CdSe (mS < mCdSe) which meets a criterion for
two-mode behavior.67 In contrast, recent studies report single
mode Raman spectra of alloyed CdSe1ꢀxSx nanocrystals at
a Se : S ratio near 50 : 50.39,50 We, and others,51 have been unable
to reproduce this result, making it unclear whether the observed
single mode behavior is anomalous or an aspect of a different
microstructure (see ESI†). We conclude that the frequency
dependence of the Raman bands on the mole fraction of sulfur
and selenium is consistent with an alloyed material.
Fig. 5 Absorption (solid line) and photoluminescence (dashed line) of
CdS/CdSe core/shell nanocrystals synthesized by adjusting ratio of N-
hexyl-N0-dodecyl thiourea and Se-Im(Et2) (A) S : Se ¼ 0.5 : 0.5, (B)
S : Se ¼ 0.25 : 0.75, (C) S : Se ¼ 0.2 : 0.8, (D) S : Se ¼ 0.0 : 1.0. The
broad fluorescence feature shown in A–C arises from trap emission.
Alloys were prepared using a pair of precursors whose
conversion reactivity is closely matched (N-methyl-N,N0-
diphenyl thiourea and Se-Im(Ph2), kSe/kS ꢃ 2.4) (Fig. 6A). With
nearly matched reactivity, both the sulde and selenide solute
are produced at rates that reect the starting ratio of precursors
throughout the reaction and can allow substantial intermixing
of the phases. The emission wavelengths of the alloys span the
blue/green region of the spectrum with the endpoints dened
by pure phase CdS (d ¼ 2.1 nm) and CdSe (d ¼ 3.0 nm) materials
prepared from the constituent precursors. Although the nal
size of the alloys is too small to be measured accurately with
TEM, we were able to determine the size using X-ray pair
distribution function analysis (these methods will be reported
Discussion
Compositional grading
To understand the inuence of the precursor reactivity on the
nal composition, the elemental distribution was simulated
using a rate equation model based on several simplifying
Fig. 6 (A) UV-vis and photoluminescence spectra of CdS (i), CdSe0.25S0.75 (ii), CdSe0.5S0.5 (iii), CdSe0.75S0.25 (iv), and CdSe (v) nanocrystals. (B) 77 K
Raman spectra (lexcite ¼ 406 nm) of same alloys showing 2-mode behavior. (C) Peak frequencies of Raman modes vs. the mole fraction of sulfur.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
assumptions: (1) the rate at which sulde or selenide deposit on work is required to understand the inuence of solute that is
the nanocrystal surface is equal to the instantaneous solute below the nucleation threshold. These factors limit the range of
generation rate (M sꢀ1), (2) solute generation exhibits rst order compositions that can be tuned from a single injection. Thus,
kinetics in the chalcogen precursor, i.e. the rate of solute gener- a single injection synthesis of the kind presented above requires
ation is the product of the chalcogenourea concentration and the precise control over both the reactivity and the mole fraction of
reactivity exponent (kE[E]), (3) the nanocrystals are spherical and the precursors, especially if the size of the nal nanocrystal is to
grow isotropically without preference for one chalcogenide or the be controlled.
other. Many complicating scenarios could lead to deviations
Synthesizing higher delity structures (e.g. with phase pure
from this simplied behavior including: a surface reaction cores, a thick graded interface, and a phase pure shell) could, in
limited growth mechanism that selectively favors sulde or principle, be achieved with precursors that have “laddered”
selenide deposition, or causes anisotropic growth, or the nite reactivity (Fig. 7, red trace). However, as the nanocrystals grow
size of the nucleus, which could skew the composition of the to large sizes and the outer layers make up a greater fraction of
core. Regardless, using these inputs, we simulated the radial the whole, the range of reactivity that is required to achieve
prole of nanocrystals produced from a wide range of Se : S precisely composed and distinct regions leads to impractically
precursor ratios and relative conversion reactivity (Fig. 7).
long reaction times. We estimate that several days of reaction
A number of useful observations can be made from these time are required to produce pure phase core/graded interface/
simulations. (1) Solute supply kinetics must be within an order pure phase shell structures with radii >10 nm from a single
of magnitude to substantially alloy the two components, while injection of precursors (see ESI†). Although a variable temper-
precursors with >10ꢂ difference in reactivity provide abrupt ature prole can further optimize the timing, limitations on the
(>90% pure) interfaces. Thus to obtain a homogeneously range of useful concentration, reaction volume, time, temper-
distributed alloy by injecting a pair of sulde and selenide ature, and reagent stability will also limit the utility of more
partners requires precisely matched reactivity. (2) Hetero- complex precursor mixtures.
structures with thick outer shells require a large amount of shell
precursor. The large amount of shell precursor increases the
rate at which the shell material is generated during nucleation.
Size control
This can cause alloying of the core and shell unless the two Adjusting the reactivity and the mole fraction of each precursor
precursors have very different reactivity. For example, a hetero- also inuences the extent of nucleation and the nal nano-
structure with a 4 nm CdSe core and a 3 nm thick CdS shell is crystal size. This is clearly observed in Fig. 4 and 5 above, where
ꢃ90% CdS by volume and requires ꢃ10ꢂ the amount of sulde increasing the reactivity or the mole fraction of the precursor
precursor. To prepare such a heterostructure with core and shell that causes nucleation reduces the nal size and blueshis the
regions that are >90% phase pure, the reactivity of the precursor emission wavelength. Similar behavior is observed when
that nucleates the core must be at least 100ꢂ greater than the synthesizing alloys using sulde and selenide precursors with
shell precursor. A further complication is the threshold amount matched conversion reactivities. The extent of nucleation is
of precursor required to cause nucleation, which in the case of similar to a linear combination of the number of pure phase
CdSe (ꢃ0.2 mM) is much smaller than CdS (ꢃ0.9 mM). The nanocrystals, weighted by the mole fraction of each precursor.
difference suggests that pure phase CdSe nuclei may be more The result is surprising given that the nucleation barrier can be
likely to form from a mixed solute composition. More detailed expected to depend on the composition of the nucleus. While
the inuence of the solute composition on nucleation is
deserving of additional study, it is clear that the precursor
reactivity remains a powerful tool to tailor the size of hetero-
structures and alloys.
The number of pure phase CdSe and CdS nuclei produced
from the chalcogenoureas is plotted versus the initial precursor
reaction rate in Fig. 8. Faster conversion reactivity nucleates
greater numbers of nanocrystals as has been described
earlier.2,4–7,68 Interestingly, the nucleation of CdS and CdSe
exhibit different dependences on the conversion reactivity, that
may arise from differences in the growth kinetics and/or
nucleus size.1 Compared to CdSe, greater numbers of CdS
nanocrystals nucleate at a given solute supply rate, and in
several cases, nucleation of CdSe nanocrystals does not follow
a simple trend.
Changing the S : Se ratio, as is oen used to change the
Fig. 7 Simulated radial elemental distributions from the first order rate
constants shown. [Se]0 and [S]0 denote the initial concentrations of the
Se and S precursors, respectively. [Se1], [Se2], [Se3], and [Se4] indicate
composition of nanocrystals in an alloy synthesis,39,43,52 can
have a complex inuence over the emission wavelength of the
nal nanocrystal. While increasing the S : Se ratio generally
blue shis the optical spectrum, the change will also adjust the
Se precursors with different reactivities. The final nanocrystal size is
chosen to be 5 nm in all cases.
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
appear to be a meaningful consequence of the growth mecha-
nism rather than a signature of irreproducible reactivity.
Moreover, we were able to optimize one pot syntheses of several
desirable architectures that are described in the ESI,† including
narrow-band cyan emitters that are challenging to produce by
conventional methods and thick shell spherical quantum wells
with PLQY near 90%.25,39,52
Conclusion
We report the synthesis of new cyclic thione and selone
precursors whose reactivity with cadmium oleate at 240 ꢁC can
be used to control the composition of CdSe1ꢀxSx hetero-
structures and alloys. These precursors are the least reactive
among a library of chalcogenoureas that can be used to
synthesize metal chalcogenide nanocrystals, extending the
range of reactivity by several orders of magnitude. By coinject-
ing a pair of thione and selone precursors, the extent of
nucleation and the intermixing of selenide and sulde regions
can be controlled to form heterostructures or alloys. The
number of nanocrystals that nucleate depends on the relative
reactivity and the molar ratio of the co-injected precursors, in
a manner that enables precise control of the nal architecture.
The dependence of the size and composition on the mole ratio
and reactivity ratio underscores the importance of precursor
design and controlled and reproducible reactivity. Advances
along these lines can lead to economical methods to manu-
facture complex heterostructures in fewer synthetic steps.
Fig. 8 Number of nanocrystals vs. initial rate of reaction; CdS (yellow)
and CdSe (red).
number of particles nucleated, particularly if the precursors
have very disparate reactivity. Whether the number of nano-
crystals increases or decreases depends on which precursor is
more reactive. When kS ꢄ kSe, increasing the S : Se reduces the
amount of selenium precursor that causes nucleation, increases
the nanocrystal size and reduces the connement; these effects
counteract the blue shi. On the other hand, when kS [ kSe, an
increase in the S : Se ratio decreases the size and further blue
shis the emission; a cooperative effect on the emission wave-
length. Moreover, Fig. 8 suggests these effects will be greater
when nucleating CdS, which is more sensitive to the conversion
reactivity.
Experimental
General considerations
The appearance of multiple nanocrystal populations
occurred in a signicant fraction of reactions studied over the
course of an extensive synthetic investigation (see ESI†).
Heterogeneity during the nucleation of pure phase CdSe and
the homogeneous nucleation of shell material, especially
during the growth of larger heterostructures, were both
observed. In several cases, the homogeneity could be improved
by adding oleic acid or adjusting the total reaction concentra-
tion, while in others these changes were not effective. Although
the origins of this behavior are unclear, the incomplete yield
produced by several of the selenoureas, the bimodal distribu-
tions observed during the synthesis of pure phase CdSe, and the
variations in the number of CdSe nanocrystals across the range
of reactivity in Fig. 8, suggest that the reactivity of the sele-
nourea precursors may not be orthogonal to the crystallization
in all cases.
On the other hand, reactions that rst nucleate CdS reliably
produce single populations of emitters with a narrow FWHM
across a range of thiourea and oleic acid concentrations (0.1–1
equiv. of thiourea and 0–10 equiv. of oleic acid).7 Thus, CdS/
CdSe/CdS spherical quantum well microstructures were more
reliably prepared with narrow photoluminescence than the
more traditional CdSe/CdS architecture. Although the origins of
the multiple populations proved too complex to rationalize in
all cases, the results described herein are reproducible and
All manipulations were performed in air unless otherwise
indicated. Toluene (99.5%), methyl acetate (99%), ethyl acetate
($99.8%), benzene (99.8%), hexanes (98.5%), methanol
(99.8%), ethanol ($99.8%), dichloromethane ($99.5%), chlo-
roform ($99.8%), acetone ($99.8%), acetonitrile (99.5%),
cadmium nitrate tetrahydrate (98%), sodium hydroxide
($98%), sodium bicarbonate ($99.7%), hydrochloric acid
(37%), sodium chloride ($99%), sodium sulfate ($99%), tet-
ramethylthiourea (98%), phenyl isothiocyanate (98%), pyrroli-
dine (99%), dimethyl terephthalate ($99.0%), triethyl
orthoformate (98%), triuoroacetic anhydride ($99%), tri-
uoroacetic acid (99%), selenium powder ꢃ100 mesh (99.99%),
and trioctylphosphine (97%) were obtained from Sigma Aldrich
and used without further purication. Oleic acid (99%) and 1,10-
thiocarbonyldiimidazole ($95.0% or 90%) were obtained from
either Sigma Aldrich or Alfa Aesar and used without further
purication. Diphenyl ether (99%), 1-octadecene (90%), hex-
adecane (99%), and tetraethylene glycol dimethyl ether (“tetra-
glyme” $99%) were obtained from Sigma Aldrich, stirred with
calcium hydride overnight, and distilled prior to use. 2-Hex-
yldecanoic acid (96%) was obtained from Sigma Aldrich, stirred
with sodium sulfate overnight, and distilled prior to use.
Chloroform-d (99.8%), benzene-d6 (99.5%), and methylene
chloride-d2 (99.8%) were obtained from Cambridge Isotopes
and used without further purication. Cadmium oxide (99.99%)
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
was obtained from Strem and used without further purication. le under vacuum overnight before cooling the sample to 77 K
N-Methylethylenediamine (95%), N-isopropylethylenediamine over approximately an hour using a LakeShore 325 temperature
(98%), N,N0-dimethylethylenediamine (99%), N,N0-diphenyl controller.
ethylenediamine
(98%),
N,N0-diisopropylethylenediamine
The Raman spectra are measured on a home-built micro-
(99%), N-methyl-1,3-propanediamine (98%), N-isopropyl-1,3- Raman spectrometer. These measurements use 0.5–1 mW of
propanediamine (95%), N,N0-dimethyl-1,3-propanediamine light from an Ondax diode laser with l ¼ 406 nm. This laser
(97%), N,N0-diethyl-1,3-propanediamine (97%), and N,N0- light diffracts off an Ondax ASE lter to remove spontaneous
diisopropyl-1,3-propanediamine (96%) were obtained from emission from the laser, enters a Nikon Eclipse Ti/U inverted
Sigma Aldrich and used without further purication. N-Ethyl- microscope, and is focused by a 40ꢂ /0.6 N.A. objective to a ꢃ1
ethylenediamine (98%), and N,N0-diethylethylenediamine mm diameter spot on the sample in the cryostat chamber. The
(96%) were obtained from Alfa Aesar and used without further backscattered light is collected by the same objective and passes
purication. N-Phenylethylenediamine (98% or 99%) was ob- through a 50 mm confocal pinhole. The light passes through two
tained from either Sigma Aldrich or Acros Organics and used polarizers and an angle-tuned Semrock 407 nm RazorEdge long
without further purication. Column chromatography was pass lter. The polarizers transmit P-polarized light with respect
performed with 40–63 mm silica gel (230–400 mesh).
to the long pass lter angle, which permits measurement of low
Tetrasubstituted, trisubstituted, disubstituted thioureas, sele- frequency vibrational modes (<300 cmꢀ1). The light enters
noureas, and thiocarbonates were prepared as described previ- a 300 mm Acton SP2300 spectrometer with an 1800 g mmꢀ1
ously. N-Hexyl-N0-dodecylthiourea,5 N-hexyl-N0N0-dibutylthiourea,5 grating before reaching a Pixis 400 CCD imaging detector. We
N-methyl-N,N0-diphenylthiourea,7 dipyrrolidinethiourea,7 and N- typically measure multiple consecutive spectra using 300 to 450
butyl-N0-pyrrolidineselenourea.6 All thiocarbonates described in second exposure times. These exposure times prevent detector
the particle number vs. rate plot were synthesized as previously saturation from uorescence or Rayleigh scattering, which
described.7
slightly transmits through the angle-tuned long pass lter.
Measurements are performed on multiple spots on a sample,
and consecutive spectra are measured on each spot to ensure
that no sample degradation occurred. Total exposure time for
a given sample is 20 to 90 minutes. We calibrate the spectra
using an argon lamp at the beginning and end of a set of
measurements for a given sample. Typical spectral resolution is
5 cmꢀ1 based on the FWHM of the l ¼ 415.86 nm peak in the
argon spectrum.
Instrumentation
UV-visible absorbance spectra were obtained using a Perki-
nElmer Lambda 950 spectrophotometer equipped with deute-
rium and halogen lamps. Photoluminescence measurements
were performed using a Fluoromax 4 from Horiba Scientic,
and photoluminescence quantum yields were determined using
a quanta-phi integrating sphere accessory according to a previ-
ously described procedure.69 Powder X-ray diffraction (XRD) was
measured on a PANalytical X'Pert Powder X-ray diffractometer.
Transmission electron microscopy (TEM) was performed on
a FEI T12 BioTWIN and a FEI Talos F200X. Fourier transform
infrared (FT-IR) spectra were obtained using a PerkinElmer
Spectrum One FT-IR Spectrometer operating with an attenuated
total reectance (ATR) accessory.
Synthesis of cadmium oleate
Cadmium oxide (99.99%) (9.5 g, 74 mmol, 1 eq.) and acetoni-
trile (95 mL) are stirred at room temperature. Triuoroacetic
acid (1.6 mL, 20.9 mmol, 0.28 eq.) and triuoroacetic anhydride
(11 mL, 79 mmol, 1.1 eq.) are added slowly and stirred for one
hour. The cadmium oxide fully dissolves, yielding a clear
colorless solution. To a 4 L Erlenmeyer ask, oleic acid (46.7
mL, 148 mmol, 2 eq.), dichloromethane (740 mL), and trie-
thylamine (26.3 mL, 188.7 mmol, 2.55 eq.) are added. The
cadmium triuoroacetate solution is then added dropwise to
the oleic acid solution with stirring. An additional 600 mL of
acetonitrile are added resulting in the formation of a white
STEM-EDX
Elemental maps were recorded on a FEI Talos F200X operated at
80 kV and equipped with a SuperX detector. Samples were drop
cast and heated to 50 C in a vacuum oven for 1 week prior to
ꢁ
imaging in order to assist in removal of volatile organics. The X-
ray signals of a 15.8 ꢂ 15.7 nm map consisting of 233 ꢂ 231
pixels were recorded over a period of 105 minutes, during which
the sample remained stable. At each pixel, a 0–20 keV spectrum
was recorded over 2048 channels. From this data cube, the
following slices were taken to obtain individual elemental
maps: SK : 2.23–2.48 keV; SeL : 1.16–1.60 keV; SeK : 11.00–11.40
ꢁ
precipitate. The mixture is heated to 60 C in order to dissolve
the precipitate, and the ask is slowly cooled to room temper-
ature and then put in a ꢀ22 ꢁC freezer. The resulting white
powder is isolated by vacuum ltration and washing with 1 L
acetonitrile, being careful to thoroughly stir the slurry and
break up large chunks. The product is dried under vacuum to
yield a ne, uffy white powder. Typical yields are 49 g (98%). 1H
NMR (CDCl3, 400 MHz) d ¼ 0.90 (t, 6H, CH3), 1.23–1.40 (m, 40H,
(CH2)6 and (CH2)4), 1.59 (m, 4H, COCH2CH2), 2.03 (m, 8H, ]
a
a,b
a
keV; SeK : 12.36–12.68 keV; CdL : 3.00–3.65 keV.
b
b
Raman spectroscopy
Nanocrystal sample thin lms for Raman spectroscopy CHCH2), 2.33 (t, 4H, COCH2), 5.36 (m, 4H, ]CH–); 13C{1H}
measurements are prepared by drop-casting concentrated NMR (CDCl3, 125 MHz) d ¼ 14.26 (CH3), 22.85 (CH2CH3), 26.15
nanocrystal solutions in hexane onto 5 ꢂ 5 mm silicon wafers. (COCH2CH2), 27.44 (]CHCH2–), 27.46 (]CHCH2–), 29.50
The samples are loaded into a Cryo Industries cryostat which is (CH2), 29.53 (CH2), 29.57 (CH2), 29.74 (CH2), 29.97 (CH2), 30.07
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
(CH2), 32.09 (CH2), 35.73 (COCH2), 129.80 (]CH–),130.10 (] (m, 4H, 4-CH2 & 5-CH2), 4.78 (m, 1H, –CH), 5.91 (br, 1H, NH);
CH–),184.15 (OOC); IR (liquid cell in tetrachloroethylene): 1317, 13C{1H} (100 MHz, CDCl3): d ¼ 19.36 (–CH3), 41.52 (4-CH2),
1405, 1437, 1468, 1530, 2850, 2873, 2918, 2955, 3007 cmꢀ1; Anal. 42.88 (5-CH2), 46.92 (–CH) 182.54 (C(S)); Anal. calcd for
calcd For CdO4C36H66: C, 64.03; H, 9.85. Found: C, 63.72; H, C6H12N2S: C, 49.96; H, 8.39; N, 19.42. Found: C, 50.25; H,
9.62.
8.18; N, 19.36. MS (ASAP) m/z Calcd for [C6H12N2S + H+]: 145.08.
Found: 145.08.
1,3-Diphenylimidazolidine-2-thione (S-Im(Ph2)). 1,10-Thio-
carbonyldiimidazole (1.09 g, 1.1 eq., 6.14 mmol) was dissolved
in 10 mL of ethyl acetate. N,N0-Diphenylethylenediamine
Synthesis of cyclic thioureas S-Im(H,R), S-Im(R2), S-Pym(H,R),
and S-Pym(R2)
1-Phenylimidazolidine-2-thione (S-Im(H,Ph)). 1,10-Thio- (1.19 g, 1 eq., 5.58 mmol) was added dropwise to the solution.
ꢁ
carbonyldiimidazole (1.09 g, 1.1 eq., 6.14 mmol) was dissolved The solution was heated to 70 C and stirred for 1 hour. Aer
in 10 mL of ethyl acetate. N-Phenylethylenediamine (0.760 g, 1 cooling, the resulting precipitate is collected via ltration and
eq., 5.58 mmol) was added dropwise to the solution. The solu- washed with hexanes to yield 1.20 g (85%) of S-Im(Ph2). The
tion was heated to 70 ꢁC and stirred for 1 hour. Aer cooling, the product was further puried by recrystallization from ethyl
resulting precipitate is collected via ltration and washed with acetate and hexanes to produce pale off-white crystals. Yield:
hexanes to yield 0.905 g (91%) of S-Im(H,Ph). The product was 1.10 g (78%). 1H NMR (400 MHz, C6D6): d ¼ 3.07 (s, 4H, –CH2),
further puried by recrystallization from ethyl acetate and 6.97–7.03 (m, 2H, p-CH), 7.14–7.23 (m, 4H, m-CH), 7.56–7.62 (m,
hexanes to produce off-white crystals. Yield: 0.835 g (84%). H 4H, o-CH); 13C{1H} (125 MHz, CDCl3): d ¼ 49.41 (–CH2), 125.47
1
NMR (400 MHz, CD2Cl2): d ¼ 3.69–3.78 (m, 2H, –CH2), 4.14–4.24 (m-CH), 126.67 (p-CH), 128.89 (o-CH), 140.85 (i-CH), 181.27
(m, 2H, –CH2), 6.35 (br, 1H, NH), 7.24–7.32 (m, 1H, p-CH), 7.39– (C(S)); Anal. calcd for C15H14N2S: C, 70.83; H, 5.55; N, 11.01.
7.48 (m, 2H, m-CH), 7.58–7.65 (m, 2H, o-C); 13C{1H} (125 MHz, Found: C, 70.77; H, 5.57; N, 10.26. MS (ASAP) m/z calcd for
CDCl3): d ¼ 41.71 (5-CH2), 52.16 (4-CH2), 124.56 (p-CH), 126.44 [C15H14N2S + H+]: 255.10. Found: 255.10.
(m-C), 128.86 (o-C), 140.02 (i-C), 182.75 (C(S)); Anal. calcd for
1,3-Dimethylimidazolidine-2-thione (S-Im(Me2)). 1,10-Thio-
C9H10N2S: C, 60.64; H, 5.65; N, 15.72. Found: C, 60.41; H, carbonyldiimidazole (1.09 g, 1.1 eq., 6.14 mmol) was dissolved
5.41; N, 15.67. MS (ASAP) m/z calcd for [C9H10N2S + H+]: 179.06. in 10 mL of ethyl acetate. N,N0-Dimethylethylenediamine
Found: 179.06.
(0.492 g, 1 eq., 5.58 mmol) was added dropwise to the solution,
1-Methylimidazolidine-2-thione (S-Im(H,Me)). N-Methyl- producing heat. The reaction mixture was extracted with 10%
ethylenediamine (1.22 g, 1 eq., 16.4 mmol) was dissolved in HCl (10 mL), washed with saturated NaHCO3 (10 mL), and then
100 mL of ethyl acetate.
A
suspension of 1,10-thio- dried over Na2SO4. The resulting product was isolated and
carbonyldiimidazole (3.23 g, 1.1 eq., 18.1 mmol) in 100 mL of recrystallized from ethyl acetate and hexanes to produce color-
ethyl acetate was added to the diamine solution slowly over the less crystals of S-Im(Me2). Yield 0.436 g (60%). 1H NMR (400
course of 10 minutes. The mixture was allowed to stir at room MHz, C6D6): d ¼ 2.38 (s, 4H, –CH2), 2.81 (s, 6H, –CH3); 13C{1H}
temperature for 1 hour. The reaction mixture was gravity ltered (100 MHz, C6D6): d ¼ 24.51 (–CH3), 47.27 (–CH2), 184.25 (C(S));
to remove a red precipitate, and the ltrate was evaporated to Anal. calcd for C5H10N2S: C, 46.12; H, 7.74; N, 21.51. Found: C,
produce an orange solid. This solid was dissolved in a minimal 46.25; H, 7.73; N, 21.45. MS (ASAP) m/z calcd for [C5H10N2S +
amount of dichloromethane, and was loaded onto a silica gel H+]: 131.06. Found: 131.06.
column, and eluted with ethyl acetate. The fractions were
1,3-Diethylimidazolidine-2-thione (S-Im(Et2)). Synthesized
consolidated and evaporated to yield S-Im(H,Me) as white following the same procedure as S-Im(Me2), instead using N,N0-
1
crystals. Yield: 1.346 g (71%). H NMR (400 MHz, CDCl3): d ¼ diethylethylenediamine. Yield 0.517 g (59%). 1H NMR (500
3.13 (s, 3H, –CH3), 3.56 (t, 2H, –CH2), 3.71 (t, 2H, –CH2) 5.88 (br, MHz, C6D6): d ¼ 0.87 (t, 6H, –CH3), 2.54 (s, 4H, –CH2), 3.52 (q,
1H, NH); 13C{1H} (100 MHz, CDCl3): d ¼ 34.30 (–CH3), 41.17 (4- 4H, –CH2CH3); 13C{1H} (125 MHz, CDCl3): d ¼ 12.06 (–CH3),
CH2), 51.19 (5-CH2), 184.07 (C(S)); Anal. calcd for C4H8N2S: C, 42.20 (–CH2CH3), 45.35 (–CH2), 181.82 (C(S)); Anal. calcd for
41.35; H, 6.94; N, 24.11. Found: C, 41.88; H, 6.88; N, 24.05. MS C7H14N2S: C, 53.12; H, 8.92; N, 17.70. Found: C, 53.25; H,
(ASAP) m/z calcd for [C4H8N2S + H+]: 117.05. Found: 117.05.
8.75; N, 17.56. MS (ASAP) m/z calcd for [C7H14N2S + H+]: 159.10.
1-Ethylimidazolidine-2-thione (S-Im(H,Et)). Synthesized Found: 159.10.
following the same procedure as S-Im(H,Me), instead using N-
1,3-Diisopropylimidazolidine-2-thione (S-Im(iPr2)). Synthe-
ethylethylenediamine. White crystals. Yield: 1.56 g (73%). 1H sized following the same procedure as S-Im(Me2), instead using
NMR (400 MHz, CDCl3): d ¼ 1.19 (t, 3H, –CH3), 3.66 (m, 6H, N,N0-diisopropylethylenediamine. Yield 0.690 g (66%). 1H NMR
–CH2 & –CH2CH3), 5.83 (br, 1H, NH); 13C{1H} (100 MHz, CDCl3): (400 MHz, C6D6): d ¼ 0.89 (d, 12H, –CH3), 2.65 (s, 4H, –CH2),
d ¼ 12.13 (–CH3), 41.31 (4-CH2), 41.57 (CH2CH3), 47.90 (5-CH2), 5.19 (m, 2H, –CH); 13C{1H} (100 MHz, C6D6): d ¼ 18.64 (–CH3),
182.90 (C(S)); Anal. Calcd for C5H10N2S: C, 46.12; H, 7.74; N, 39.88 (–CH2), 46.35 (–CH), 182.06 (C(S)); Anal. calcd for
21.51. Found: C, 45.86; H, 7.57; N, 21.29. MS (ASAP) m/z calcd C9H18N2S: C, 58.02; H, 9.74; N, 15.04. Found: C, 58.12; H,
for [C5H10N2S + H+]: 131.06. Found: 131.06.
9.52; N, 15.02. MS (ASAP) m/z calcd for [C9H18N2S + H+]: 187.13.
1-Isopropylimidazolidine-2-thione (S-Im(H,iPr)). Synthe- Found: 187.13.
sized following the same procedure as S-Im(H,Me), instead
1-Methyltetrahydropyrimidine-2(1H)-thione (S-Pym(H,Me)).
using N-isopropylethylenediamine. White crystals. Yield: 1.05 g Synthesized following the same procedure as S-Im(H,Me),
(74%). 1H NMR (400 MHz, CDCl3): d ¼ 1.17 (d, 6H, –CH3), 3.59 instead using N-methyl-1,3-propanediamine. Yield: 1.16 g
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
(54%). 1H NMR (400 MHz, CDCl3): d ¼ 2.01 (m, 2H, –CH2), 3.28 a small amount of liquid condensed in the receiving ask. The
(t, 2H, –CH2), 3.35 (t, 2H, –CH2), 3.39 (s, 3H, –CH3), 6.35 (br, 1H, reaction mixture was then allowed to cool to room temperature
NH); 13C{1H} (100 MHz, CDCl3): d ¼ 21.12 (–CH2), 40.75 (–CH2), and the triethyl orthoformate was removed by distillation under
42.12 (–CH2), 48.35 (–CH3), 178.04 (C(S)); Anal. calcd for reduced pressure. The ask was opened to air and the
C5H10N2S: C, 46.12; H, 7.74; N, 21.51. Found: C, 47.37; H, remaining solid residue was dissolved in dichloromethane,
7.62; N, 20.96. MS (ASAP) m/z calcd for [C5H10N2S + H+]: 131.06. ltered through Celite, and recrystallized once outside of the
Found: 131.06.
glovebox. Caution: The contained should be handled in an
1-Isopropyltetrahydropyrimidine-2(1H)-thione
(S-Pym(- efficient fumehood to avoid exposure to a strong odor. The
H,iPr)). Synthesized following the same procedure as S- resulting solid residue was brought into a nitrogen-lled glove
Im(H,Me), instead using N-isopropyl-1,3-propanediamine and box, where it was dissolved in acetonitrile, syringe ltered
0.36 mmol scale. Yield: 0.0263 g (46%). 1H NMR (400 MHz, (PTFE, 0.2 mm), and then puried by recrystallization from
CDCl3): d ¼ 1.16 (d, 6H, CH3), 1.96 (m, 2H, CH2), 3.24 (m, 4H, acetonitrile in a ꢀ40 ꢁC freezer for >2 hours. The resulting solid
CH2), 5.64 (m, 1H, CH), 6.25 (br, 1H, NH); 13C{1H} (100 MHz, was isolated by suction ltration using a fritted glass funnel,
CDCl3): d ¼ 19.07 (–CH3), 21.11 (–CH2), 38.99 (–CH2), 40.85 and washed thoroughly with pentane, and then dried under
(–CH2), 51.87 (–CH), 177.05 (C(S)); Anal. calcd for C7H14N2S: C, vacuum for >6 hours.
53.12; H, 8.92; N, 17.70. Found: C, 53.32; H, 8.76; N, 17.72. MS
(ASAP) m/z calcd for [C7H14N2S + H+]: 159.10. Found: 159.10.
1,3-Dimethyltetrahydropyrimidine-2(1H)-thione
1,3-Diethylimidazolidine-2-selenone
Diethylimidazolidine-2-selenone was prepared according to the
(S- general procedure above using N,N0-diethylethylenediamine
(Se-Im(Et2)).
1,3-
Pym(Me2)). Synthesized following the same procedure as S- (11.330 g, 13.97 mL, 97.5 mmol), selenium (7.699 g, 97.5 mmol),
Im(H,Me), instead using N,N0-dimethyl-1,3-propanediamine and triethyl orthoformate (28.899 g, 32.43 mL, 32.43 mmol). It is
and a 3.6 mmol scale. Yield: 0.373 g (72%). 1H NMR (400 recrystallized by addition of pentane to a saturated toluene
MHz, CDCl3): d ¼ 2.04 (m, 2H, –CH2), 3.38 (t, 4H, –CH2), 3.43 (s, solution of the crude product. Yield: 15.3 g (76.5%) 1H NMR
6H, –CH3); 13C{1H} (100 MHz, CDCl3): d ¼ 21.15 (–CH2), 43.46 (400 MHz, CDCl3): d ¼ 1.17 (t, 6H, –CH3), 3.55 (s, 4H, –CH2), 3.75
(–CH2), 48.87 (–CH3) 179.38 (C(S)); Anal. calcd for C6H12N2S: C, (q, 4H, –CH2CH3); 13C{1H} (101 MHz, CDCl3): d ¼ 12.28 (–CH3),
49.96; H, 8.39; N, 19.42. Found: C, 50.73; H, 8.27; N, 19.32. MS 44.16 (–CH2CH3), 46.37 (–CH2), 179.86 (C(Se)); 77Se {1H} (76
(ASAP) m/z calcd for [C6H12N2S + H+]: 145.08. Found: 145.08.
MHz, CDCl3): d ¼ 62.14; Anal. calcd for C7H14N2Se: C, 40.98; H,
1,3-Diethyltetrahydropyrimidine-2(1H)-thione (S-Pym(Et2)). 6.88; N, 13.65. Found: C, 41.19; H, 6.57; N, 13.70. MS (ASAP) m/z
Synthesized following the same procedure as S-Im(H,Me), Calcd for [C7H14N2Se + H+]: 207.04. Found: 207.04.
instead using N,N0-diethyl-1,3-propanediamine. Yield: 2.09 g
1-Phenylimidazolidine-2-selenone
(Se-Im(H,Ph)).
This
1
(74%). H NMR (400 MHz, CDCl3): d ¼ 1.23 (t, 6H, –CH3), 1.99 synthesis was performed on a 120 mmol scale. Yield: 6.85 g
(m, 2H, –CH2), 3.33 (t, 4H, –CH2), 3.97 (q, 4H, –CH2CH3); 13C (25%). 1H NMR (400 MHz, CD2Cl2): d ¼ 3.76 (t, 2H, –CH2), 4.18
{1H} (100 MHz, CDCl3): d ¼ 12.13 (–CH3), 21.38 (–CH2), 46.08 (t, 2H, –CH2), 6.66 (br, 1H, NH), 7.34 (t, 1H, p-CH), 7.46 (t, 2H, m-
(–CH2), 49.86 (–CH2CH3), 177.59 (C(S)); Anal. calcd for CH), 7.61 (d, 2H, o-CH); 13C{1H} (101 MHz, CD2Cl2): d ¼ 43.16
C8H16N2S: C, 55.77; H, 9.36; N, 16.26. Found: C, 56.03; H, (–CH2), 52.91 (–CH2), 125.36 (m-CH), 126.83 (p-CH), 128.67 (o-
9.22; N, 16.29. MS (ASAP) m/z calcd for [C8H16N2S + H+]: 173.11. CH), 140.64 (i-C), 180.32 (C(Se)); 77Se {1H} (76 MHz, CD2Cl2): d ¼
Found: 173.11.
133.52; Anal. calcd for C9H10N2Se: C, 48.01; H, 4.48; N, 12.44.
1,3-Diisopropyltetrahydropyrimidine-2(1H)-thione
(S- Found: C, 47.90; H, 4.49; N, 12.37. MS (ASAP) m/z calcd for
Pym(iPr2)). Synthesized following the same procedure as S- [C9H10N2Se + H+]: 227.01. Found: 227.01.
Im(H,Me), instead using N,N0-diethyl-1,3-propanediamine and
1-Ethylimidazolidine-2-selenone
(Se-Im(H,Et)).
This
7.64 mmol scale. Yield: 1.16 g (76%). 1H NMR (400 MHz, CDCl3): synthesis was performed on a 30 mmol scale. Yield: 2.01 g
1
d ¼ 1.16 (d, 12H, –CH3), 1.91 (m, 2H, –CH2), 3.18 (t, 4H, –CH2), (38%). H NMR (400 MHz, CDCl3): d ¼ 1.17 (t, 3H, –CH3), 3.58
5.91 (m, 2H, –CH); 13C{1H} (100 MHz, CDCl3): d ¼ 19.23 (–CH3), (m, 2H, –CH2), 3.67 (m, 4H, –CH2), 6.83 (br, 1H, NH); 13C{1H}
21.76 (–CH2), 39.59 (–CH2), 52.19 (–CH), 177.88 (C(S)); Anal. (101 MHz, CDCl3): d ¼ 12.24 (–CH3), 42.66 (–CH2), 43.51 (–CH2),
calcd for C10H20N2S: C, 59.95; H, 10.06; N, 13.98. Found: C, 48.08 (–CH2), 179.09 (C(Se)); 77Se {1H} (76 MHz, CD2Cl2): d ¼
61.14; H, 10.11; N, 12.79. MS (ASAP) m/z calcd for [C10H20N2S + 67.70; Anal. calcd for C5H10N2Se: C, 33.91; H, 5.69; N, 15.82.
H+]: 201.14. Found: 201.14.
Found: C, 33.98; H, 5.49; N, 15.75. MS (ASAP) m/z calcd for
[C5H10N2Se + H+]: 179.01. Found: 179.01.
1-Isopropylimidazolidine-2-selenone (Se-Im(H,iPr)). This
synthesis was performed on a 21 mmol scale. Yield: 0.84 g
(21%). 1H NMR (400 MHz, CDCl3): d ¼ 1.16 (d, 6H, –CH3), 3.56
Synthesis of cyclic selenoureas Se-Im(H,R), Se-Im(R2), Se-
Pym(H,R), and Se-Pym(R2)
Adapting a procedure from Zhou and Denk,56 selenium (97.5 (m, 4H, –CH2), 4.85 (m, 1H, –CH), 6.68 (br, 1H, NH); 13C{1H}
mmol), triethyl orthoformate (195 mmol), and the appropriate (101 MHz, CDCl3): d ¼ 19.53 (–CH3), 42.79 (–CH2), 43.07 (–CH2),
diamine (97.5 mmol) were added to a PTFE-sealable Schlenk 49.08 (–CH), 178.60 (C(Se)); 77Se {1H} (76 MHz, CD2Cl2): d ¼
ask equipped with a distillation apparatus. The reaction 70.24; Anal. calcd for C6H12N2Se: C, 37.70; H, 6.33; N, 14.66.
mixture was degassed by the freeze–pump–thaw method, placed Found: C, 37.74; H, 6.12; N, 14.64. MS (ASAP) m/z calcd for
under an argon atmosphere, and heated to 130 ꢁC with stirring [C6H12N2Se + H+]: 193.02. Found: 193.02.
for 8 hours. Over this period, the selenium dissolved and
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
13C{1H} (101 MHz, CDCl3): d ¼ 12.31 (–CH3), 21.06 (–CH2), 45.68
(–CH2), 53.26 (–CH2), 175.65 (C(Se)); 77Se {1H} (76 MHz, C6D6):
d ¼ 175.18; Anal. calcd for C8H16N2S: C, 43.84; H, 7.36; N, 12.78.
Found: C, 43.88; H, 7.31; N, 12.74. MS (ASAP) m/z calcd for
[C8H16N2S + H+]: 221.06. Found: 221.06.
1,3-Diphenylimidazolidine-2-selenone (Se-Im(Ph2)). This
synthesis was performed on a 148 mmol scale. Yield: 19.67 g
(44%). 1H NMR (400 MHz, CD2Cl2): d ¼ 4.15 (s, 4H, –CH2), 7.30–
7.36 (m, 2H, p-CH), 7.41–7.49 (m, 4H, m-CH), 7.53–7.59 (m, 4H,
o-CH); 13C{1H} (101 MHz, CD2Cl2): d ¼ 51.44 (–CH2), 126.96 (m-
CH), 127.50 (p-CH), 129.20 (o-CH), 142.08 (i-C), 181.38 (C(Se));
77Se {1H} (76 MHz, CD2Cl2): d ¼ 174.28; Anal. calcd for
1,3-Diisopropyltetrahydropyrimidine-2(1H)-selenone
(Se-
Pym(iPr2)). This synthesis was performed on a 20 mmol scale.
Yield: 0.15 g (3%) 1H NMR (400 MHz, CDCl3): d ¼ 1.21 (d, 12H,
–CH3), 1.92 (m, 2H, –CH2), 3.18 (t, 4H, –CH2), 6.08 (m, 2H, –CH);
13C{1H} (101 MHz, CDCl3): d ¼ 19.25 (–CH3), 21.41 (–CH2), 39.77
(–CH2), 56.20 (–CH), 176.12 (C(Se)); 77Se {1H} (76 MHz, C6D6):
d ¼ 175.96; Anal. calcd for C10H20N2Se: C, 48.58; H, 8.15; N,
11.33. Found: C, 48.98; H, 7.93; N, 11.40. MS (ASAP) m/z calcd
for [C10H20N2Se + H+]: 249.09. Found: 249.09.
C
15H14N2Se: C, 59.81; H, 4.68; N, 9.30. Found: C, 59.68; H,
4.58; N, 9.24. MS (ASAP) m/z calcd for [C15H14N2Se + H+]: 303.04.
Found: 303.04.
1,3-Ditertbutylimidazolidine-2-selenone (Se-Im(t-Bu2)). This
synthesis was performed on a 103 mmol scale. Yield: 0.81 g
1
(3%). H NMR (400 MHz, CDCl3): d ¼ 1.69 (s, 18H, –CH3), 3.47
(s, 4H, –CH2); 13C{1H} (101 MHz, CDCl3): d ¼ 28.92 (–CH3), 45.71
(–CH2), 58.07 (–C(CH3)3), 179.09 (C(Se)); 77Se {1H} (76 MHz,
CDCl3): d ¼ 274.39; Anal. calcd for C11H22N2Se: C, 50.57; H,
8.49; N, 10.72. Found: C, 50.66; H, 8.10; N, 10.79. MS (ASAP) m/z
calcd for [C11H22N2Se + H+]: 263.10. Found: 263.10.
Synthesis of CdS, CdSe, and CdSe1ꢀxSx nanocrystals
In a nitrogen-lled glove box, a three-neck round bottom ask
is loaded with cadmium oleate (0.18 mmol, 0.122 g), 1-octa-
decene (14.25 mL, 11.2 g, 44.4 mmol), and oleic acid (0.102 g,
0.114 mL, 0.36 mmol). A 4 mL vial was lled with the desired
sulfur and/or selenium precursor (0.15 mmol) and diphenyl
ether (0.78 mL, 0.80 g) or tetraglyme (0.75 mL, 0.75 g) and
sealed with a rubber septum. Most compounds are more
soluble in tetraglyme, however diphenyl ether is necessary to
monitor reactions by 1H NMR spectroscopy in the 3–4 ppm
range. The three-neck round bottom ask is transferred to
a Schlenk line and heated to 240 ꢁC under Ar. Occasionally,
precursor solutions were heated or sonicated to ensure
a homogenous solution prior to injection. The sulfur and/or
selenium precursor solution is then injected into the
cadmium oleate solution, and the reaction mixture is stirred
for the appropriate time. The resulting nanocrystals were
isolated from the reaction mixture by precipitation with
acetone and centrifugation. The colored residue is redispersed
in hexane (10 mL). Acetone (5–10 mL) is added in 0.5 mL
portions to precipitate cadmium oleate as a white solid,
without precipitating the nanocrystals. The suspension was
centrifuged, the supernatant collected, and the nanocrystals
precipitated by adding 25 mL of acetone. The nanocrystals
were washed three additional times by redispersion in toluene
and precipitation with methyl acetate. For specic syntheses
of core/shell and alloyed nanocrystals, please refer to the ESI.†
1,3-Dimethylimidazolidine-2-selenone (Se-Im(Me2)). This
synthesis was performed on a 168 mmol scale. Yield: 15.01 g
(50%). 1H NMR (400 MHz, C6D6): d ¼ 2.38 (s, 4H, –CH2), 2.89 (s,
6H, –CH3); 13C{1H} (101 MHz, C6D6): d ¼ 36.62 (–CH3), 48.50
(–CH2), 183.46 (C(Se)); 77Se {1H} (76 MHz, C6D6): d ¼ 110.90;
Anal. calcd for C5H10N2Se: C, 33.91; H, 5.69; N, 15.82. Found: C,
33.93; H, 5.47; N, 15.77. MS (ASAP) m/z calcd for [C5H10N2Se +
H+]: 179.01. Found: 179.01.
1,3-Diisopropylimidazolidine-2-selenone (Se-Im(iPr2)). This
synthesis was performed on a 86 mmol scale. Yield: 11.17 g
(55.9%). 1H NMR (400 MHz, C6D6): d ¼ 0.88 (d, 12H, –CH3), 2.63
(s, 4H, –CH2), 5.33 (m, 2H, –CH); 13C{1H} (101 MHz, C6D6): d ¼
19.11 (–CH3), 41.05 (–CH2), 48.70 (–CH), 181.30 (C(Se)); 77Se {1H}
(76 MHz, C6D6): d ¼ 107.23; Anal. calcd for C9H18N2Se: C, 46.35;
H, 7.78; N, 12.01. Found: C, 46.41; H, 7.51; N, 12.14. MS (ASAP)
m/z calcd for [C9H18N2Se + H+]: 235.07. Found: 235.07.
1-Methyltetrahydropyrimidine-2(1H)-selenone
(Se-
Pym(H,Me)). This synthesis was performed on a 32 mmol scale.
Yield: 0.62 g (11%). 1H NMR (400 MHz, CDCl3): d ¼ 2.07 (m, 2H,
–CH2), 3.23 (m, 2H, –CH2), 3.36 (t, 2H, –CH2), 3.50 (s, 3H, –CH3),
6.70 (br, 1H, NH); 13C{1H} (101 MHz, CDCl3): d ¼ 20.71 (–CH2),
40.57 (–CH2), 45.08 (–CH2), 48.04 (–CH3), 173.35 (C(Se)); 77Se
{1H} (76 MHz, C6D6): d ¼ 205.43; Anal. calcd for C5H10N2Se: C,
33.91; H, 5.69; N, 15.82. Found: C, 3.99; H, 5.48; N, 15.80. MS
(ASAP) m/z calcd for [C5H10N2Se + H+]: 179.01. Found: 179.01.
1,3-Dimethyltetrahydropyrimidine-2(1H)-selenone
(Se-
Pym(Me2)). This synthesis was performed on a 45 mmol scale.
Yield: 0.69 g (8%). 1H NMR (400 MHz, CDCl3): d ¼ 2.05 (m, 2H,
–CH2), 3.36 (t, 4H, –CH2), 3.55 (s, 6H, –CH3); 13C{1H} (101 MHz,
CDCl3): d ¼ 20.92 (–CH2), 46.94 (–CH2), 48.46 (–CH3), 177.76
(C(Se)); 77Se {1H} (76 MHz, C6D6): d ¼ 221.22; Anal. calcd for
C6H12N2Se: C, 37.70; H, 6.33; N, 14.66. Found: C, 37.72; H,
6.16; N, 14.63. MS (ASAP) m/z calcd for [C6H12N2Se + H+]: 193.02.
Found: 193.02.
Nanocrystal formation kinetics via absorption spectra
Aliquots of approximately 0.1 mL were taken from a CdS or CdSe
nanocrystal reaction and deposited into a previously weighed
vial. A mass of toluene equal to 2.5ꢂ the weight of the aliquot
was added to the vial to standardize aliquot concentration. UV-
vis absorption spectra were taken of each aliquot and the
concentration of cadmium chalcogenide in the aliquot was
calculated from the size-dependent extinction coefficient at the
rst excitonic absorption maximum for CdS7 or the size-
independent absorption coefficient of CdSe at 350 nm.62 The
kinetics collected from each reaction were t to a single expo-
nential whose value is reported in Table 1.
1,3-Diethyltetrahydropyrimidine-2(1H)-selenone
(Se-
Pym(Et2)). This synthesis was performed on a 52 mmol scale.
1
Yield: 0.23 g (2%). H NMR (400 MHz, CDCl3): d ¼ 1.24 (t, 6H,
–CH3), 2.00 (m, 2H, –CH2), 3.31 (t, 4H, –CH2), 4.07 (q, 4H, –CH2);
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
1
and J.S.O. wrote the manuscript. All authors revised the
manuscript.
Precursor conversion kinetics via H NMR
Quantitative aliquots of 200 mL were taken from a CdS, CdSe, or
mixed precursor reaction and diluted with 300 mL of CD2Cl2 and
100 mL of a 22.4 mM solution of dimethyl terephthalate dis-
1
Notes and references
solved in CD2Cl2. Quantitative H NMR spectra were collected
with a relaxation delay time of 30 s. Diphenyl ether must be
used as the injection solvent instead of tetraglyme in order to
monitor precursor disappearance kinetics between 3.0 and
4.5 ppm. Precursor disappearance was measured versus the
dimethyl terephthalate internal standard and compared with
the appearance of nanocrystals as measured using UV-vis
absorbance spectroscopy.
1 D. B. Chu, J. S. Owen and B. Peters, J. Phys. Chem. A, 2017,
121, 7511–7517.
2 S. Abe, R. K. Cabek, B. De Geyter and Z. Hens, ACS Nano,
2012, 6, 42–53.
3 T. Sugimoto and F. Shiba, Colloids Surf., A, 2000, 164, 205–
215.
4 J. S. Owen, E. M. Chan, H. Liu and A. P. Alivisatos, J. Am.
Chem. Soc., 2010, 132, 18206–18213.
5 M. P. Hendricks, M. P. Campos, G. T. Cleveland, I. Jen-La
Plante and J. S. Owen, Science, 2015, 348, 1226–1230.
6 M. P. Campos, M. P. Hendricks, A. N. Beecher, W. Walravens,
R. A. Swain, G. T. Cleveland, Z. Hens, M. Y. Sfeir and
J. S. Owen, J. Am. Chem. Soc., 2017, 139, 2296–2305.
7 L. S. Hamachi, I. Jen-La Plante, A. C. Coryell, J. De Roo and
J. S. Owen, Chem. Mater., 2017, 29, 8711–8719.
8 Y. Guo, S. R. Alvarado, J. D. Barclay and J. Vela, ACS Nano,
2013, 7, 3616–3625.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
The authors would like to acknowledge Mr Soham Banerjee and
Prof. Simon Billinge for X-ray PDF analysis of alloy nanocrystals.
Synthesis of core–shell nanocrystals with emission in the green
spectral region was supported by the US Army Research Labo-
ratory and the US Army Research Office under contract number
9 W. W. Yu and X. Peng, Angew. Chem., Int. Ed., 2002, 41, 2368–
2371.
W911NF-12-1-0594 (MURI). Precursor reaction kinetics, syringe 10 C. R. Bullen and P. Mulvaney, Nano Lett., 2004, 4, 2303–2307.
pump shelling methods, kinetics simulations, and synthesis of 11 J. van Embden and P. Mulvaney, Langmuir, 2005, 21, 10226–
core–shell nanocrystals with emission in the red spectral region
were supported by the U.S. Department of Energy's Office of 12 S. Abe, R. K. Capek, B. D. Geyter and Z. Hens, ACS Nano,
Energy Efficiency and Renewable Energy (EERE) under the 2013, 7, 943–949.
Building Technologies Office (BTO) under contract number DE- 13 K. De Nolf, R. K. Capek, S. Abe, M. Sluydts, Y. Jang,
10233.
EE0007628. Work at the Molecular Foundry was supported by
the Office of Science, Office of Basic Energy Sciences, of the U.S.
J. C. Martins, S. Cottenier, E. Lifshitz and Z. Hens, J. Am.
Chem. Soc., 2015, 137, 2495–2505.
Department of Energy under Contract No. DE-AC02- 14 K. Nose, H. Fujita, T. Omata, S. Otsuka-Yao-Matsuo,
05CH11231. Raman spectroscopy measurements were sup- H. Nakamura and H. Maeda, J. Lumin., 2007, 126, 21–26.
ported by a Barnard College Presidential Research Award and 15 B. Mahler, N. Lequeux and B. Dubertret, J. Am. Chem. Soc.,
the National Science Foundation Division of Chemistry under 2010, 132, 953–959.
award number 1709287. Some of this work was performed at the 16 B. D. Mangum, T. S. Landes, B. R. Theobald and J. N. Kurtin,
Simons Electron Microscopy Center and National Resource for Photonics Res., 2017, 5, A13–A22.
Automated Molecular Microscopy located at the New York 17 Y. Kang, Z. Song, X. Jiang, X. Yin, L. Fang, J. Gao, Y. Su and
Structural Biology Center, supported by grants from the Simons F. Zhao, Nanoscale Res. Lett., 2017, 12, 154.
Foundation (349247), NYSTAR, and the NIH National Institute 18 X. Dai, Z. Zhang, Y. Jin, Y. Niu, H. Cao, X. Liang, L. Chen,
of General Medical Sciences (GM103310). We would like to J. Wang and X. Peng, Nature, 2014, 515, 96.
acknowledge Columbia University's Nanoscience Initiative for 19 K. J. McHugh, L. Jing, A. M. Behrens, S. Jayawardena,
the use of its shared facilities, and the Fu Foundation School of
Engineering and Applied Science and the Faculty of Arts &
W. Tang, M. Gao, R. Langer and A. Jaklenec, Adv. Mater.,
2018, 30, 1706356.
Sciences for their support of the CNI Facilities. We acknowledge 20 S. Mazumder, R. Dey, M. Mitra, S. Mukherjee and G. Das, J.
the European Union's Framework Programme for Research and Nanomater., 2009, 2009, 38.
Innovation Horizon 2020 (2014-2020) under the Marie 21 G. E. Cragg and A. L. Efros, Nano Lett., 2010, 10, 313–317.
Skłodowska-Curie Grant Agreement COMPASS No. 691185. 22 O. Chen, J. Zhao, V. P. Chauhan, J. Cui, C. Wong,
Author contributions L.S.H., I.J.L., K.Q., M.P.C., and I.R. per-
formed precursor synthesis, nanocrystal synthesis, and inter-
D. K. Harris, H. Wei, H. Han, D. Fukumura, R. K. Jain and
M. G. Bawendi, Nat. Mater., 2013, 12, 445–451.
preted data. G.T.C. performed precursor synthesis. H.Y. 23 J. van Embden, J. Jasieniak and P. Mulvaney, J. Am. Chem.
performed nanocrystal synthesis, kinetics simulations, and Soc., 2009, 131, 14299–14309.
interpreted data. N.S., A.O., and A.C.C. performed Raman 24 B. G. Jeong, Y. Park, J. H. Chang, I. Cho, J. K. Kim, H. Kim,
experiments and interpreted Raman data. W.W., L.S.H., and
I.J.L. performed TEM experiments. L.S.H., H.Y., N.S., A.C.C.,
K. Char, J. Cho, V. I. Klimov, P. Park, D. C. Lee and
W. K. Bae, ACS Nano, 2016, 10, 9297–9305.
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
25 J. Zhang, Q. Yang, H. Cao, C. I. Ratcliffe, D. Kingston, 47 T. P. A. Ruberu, H. R. Albright, B. Callis, B. Ward, J. Cisneros,
Q. Y. Chen, J. Ouyang, X. Wu, D. M. Leek, F. S. Riehle and
K. Yu, Chem. Mater., 2016, 28, 618–625.
26 Y. Chen, J. Vela, H. Htoon, J. L. Casson, D. J. Werder,
D. A. Bussian, V. I. Klimov and J. A. Hollingsworth, J. Am.
Chem. Soc., 2008, 130, 5026–5027.
H. Fan and J. Vela, ACS Nano, 2012, 6, 5348–5359.
48 J. T. Held, K. I. Hunter, N. Dahod, B. Greenberg,
D. Reifsnyder Hickey, W. A. Tisdale, U. Kortshagen and
K. A. Mkhoyan, ACS Appl. Nano Mater., 2018, 1, 989–996.
´
49 R. Xie, U. Kolb, J. Li, T. Basche and A. Mews, J. Am. Chem.
27 M. Nasilowski, P. Spinicelli, G. Patriarche and B. Dubertret,
Nano Lett., 2015, 15, 3953–3958.
Soc., 2005, 127, 7480–7488.
50 N. Tschirner, H. Lange, A. Schliwa, A. Biermann,
C. Thomsen, K. Lambert, R. Gomes and Z. Hens, Chem.
Mater., 2012, 24, 311–318.
51 P. Mukherjee, S. J. Lim, T. P. Wrobel, R. Bhargava and
A. M. Smith, J. Am. Chem. Soc., 2016, 138, 10887–10896.
`
28 P. Reiss, M. Protiere and L. Li, Small, 2009, 5, 154–168.
29 D. D. Sarma, A. Nag, P. K. Santra, A. Kumar, S. Sapra and
P. Mahadevan, J. Phys. Chem. Lett., 2010, 1, 2149–2153.
30 W. K. Bae, L. A. Padilha, Y. Park, H. McDaniel, I. Robel,
J. M. Pietryga and V. I. Klimov, ACS Nano, 2013, 7, 3411–3419. 52 L. A. Swafford, L. A. Weigand, M. J. Bowers, J. R. McBride,
´
31 Y. S. Park, A. V. Malko, J. Vela, Y. Chen, Y. Ghosh, F. Garcıa-
J. L. Rapaport, T. L. Watt, S. K. Dixit, L. C. Feldman and
S. J. Rosenthal, J. Am. Chem. Soc., 2006, 128, 12299–12306.
53 H. Qian, X. Qiu, L. Li and J. Ren, J. Phys. Chem. B, 2006, 110,
9034–9040.
´
Santamarıa, J. A. Hollingsworth, V. I. Klimov and H. Htoon,
Phys. Rev. Lett., 2011, 106, 187401.
32 D. Morgan, K. Gong, A. M. Kelley and D. F. Kelley, J. Phys.
Chem. C, 2017, 121, 18307–18316.
54 E. Jang, S. Jun and L. Pu, Chem. Commun., 2003, 2694–2695.
33 J. J. Li, Y. A. Wang, W. Guo, J. C. Keay, T. D. Mishima, 55 L. Piveteau, T.-C. Ong, B. J. Walder, D. N. Dirin, D. Moscheni,
¨
M. B. Johnson and X. Peng, J. Am. Chem. Soc., 2003, 125,
12567–12575.
B. Schneider, J. Bar, L. Protesescu, N. Masciocchi and
A. Guagliardi, ACS Cent. Sci., 2018, 4, 1113–1125.
34 K. Boldt, N. Kirkwood, G. A. Beane and P. Mulvaney, Chem. 56 Y. Zhou and M. K. Denk, Tetrahedron Lett., 2003, 44, 1295–
Mater., 2013, 25, 4731–4738. 1299.
35 Y. Ghosh, B. D. Mangum, J. L. Casson, D. J. Williams, 57 M. K. Denk, S. Gupta, J. Brownie, S. Tajammul and
H. Htoon and J. A. Hollingsworth, J. Am. Chem. Soc., 2012,
134, 9634–9643.
36 T. P. A. Ruberu and J. Vela, ACS Nano, 2011, 5, 5775–5784.
A. J. Lough, Chem.–Eur. J., 2001, 7, 4477–4486.
58 A. V. Bogatskii, N. G. Luk'yanenko and T. I. Kirichenko,
Khim. Geterotsikl. Soedin., 1982, 6, 723–737.
37 N. Al-Salim, A. G. Young, R. D. Tilley, A. J. McQuillan and 59 H. A. Staab, Angew. Chem., Int. Ed., 1962, 1, 351–367.
J. Xia, Chem. Mater., 2007, 19, 5185–5193.
38 C. Hsia, A. Wuttig and H. Yang, ACS Nano, 2011, 5, 9511–
9522.
60 M. Horikawa, J. Busch-Petersen and E. J. Corey, Tetrahedron
Lett., 1999, 40, 3843–3846.
´
´
61 R. l. Garcıa-Rodrıguez, M. P. Hendricks, B. M. Cossairt,
39 T. Aubert, M. Cirillo, S. Flamee, R. Van Deun, H. Lange,
H. Liu and J. S. Owen, Chem. Mater., 2013, 25, 1233–1249.
C. Thomsen and Z. Hens, Chem. Mater., 2013, 25, 2388–2390. 62 C. A. Leatherdale, W. K. Woo, F. V. Mikulec and
40 J. Ouyang, M. Vincent, D. Kingston, P. Descours, M. G. Bawendi, J. Phys. Chem. B, 2002, 106, 7619–7622.
T. Boivineau, M. B. Zaman, X. Wu and K. Yu, J. Phys. 63 H. Liu, J. S. Owen and A. P. Alivisatos, J. Am. Chem. Soc., 2007,
Chem. C, 2009, 113, 5193–5200. 129, 305–312.
41 H. Cao, J. Ma, L. Huang, H. Qin, R. Meng, Y. Li and X. Peng, J. 64 R. Tan, Y. Yuan, Y. Nagaoka, D. Eggert, X. Wang, S. Thota,
Am. Chem. Soc., 2016, 138, 15727–15735.
42 N. P. Gurusinghe, N. N. Hewa-Kasakarage and M. Zamkov, J.
Phys. Chem. C, 2008, 112, 12795–12800.
43 R. Wang, O. Calvignanello, C. I. Ratcliffe, X. Wu, D. M. Leek,
M. B. Zaman, D. Kingston, J. A. Ripmeester and K. Yu, J. Phys.
Chem. C, 2009, 113, 3402–3408.
P. Guo, H. Yang, J. Zhao and O. Chen, Chem. Mater., 2017,
29, 4097–4108.
65 A. C. Berends, W. Van Der Stam, J. P. Hofmann, E. Bladt,
J. D. Meeldijk, S. Bals and C. de Mello Donega, Chem.
Mater., 2018, 30, 2400–2413.
66 K. A. Abel, J.-C. Boyer, C. M. Andrei and F. C. van Veggel, J.
Phys. Chem. Lett., 2011, 2, 185–189.
67 I. F. Chang and S. S. Mitra, Phys. Rev., 1968, 172, 924.
44 R. E. Bailey and S. Nie, J. Am. Chem. Soc., 2003, 125, 7100–
7106.
45 D. K. Smith, J. M. Luther, O. E. Semonin, A. J. Nozik and 68 E. M. Chan, C. Xu, A. W. Mao, G. Han, J. S. Owen, B. E. Cohen
M. C. Beard, ACS Nano, 2010, 5, 183–190. and D. J. Milliron, Nano Lett., 2010, 10, 1874–1885.
46 E. Bladt, R. J. A. van Dijk-Moes, J. Peters, F. Montanarella, 69 N. C. Anderson, M. P. Hendricks, J. J. Choi and J. S. Owen, J.
C. de Mello Donega, D. Vanmaekelbergh and S. Bals, J. Am.
Chem. Soc., 2016, 138, 14288–14293.
Am. Chem. Soc., 2013, 135, 18536–18548.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019