Rapid phosphine-free synthesis of CdSe quantum dots: Promoting the generation of Se precursors using a radical initiator

Article abstract of DOI:10.1039/c4ta00285g

The replacement of phosphine containing compounds in the synthesis of II-VI quantum dots (QDs) via the 'hot-injection' method has received considerable attention in recent years, in particular toward scaling-up production. A key bottleneck in current appr

Full text of DOI:10.1039/c4ta00285g

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Rapid phosphine-free synthesis of CdSe quantum
dots: promoting the generation of Se precursors
using a radical initiator†
Cite this: J. Mater. Chem. A, 2014, 2,
6879
Bo Hou,a David Benito-Alifonso,a Richard Webster,b David Cherns,b
a
a
*
M. Carmen Galan and David J. Fermın
´
The replacement of phosphine containing compounds in the synthesis of II–VI quantum dots (QDs) via the
‘hot-injection’ method has received considerable attention in recent years, in particular toward scaling-up
production. A key bottleneck in current approaches is the poor solubility of elemental Se in solvents such as
1-octadecene (1-ODE) or oleylamine requiring a heating temperature of 200 ꢀC for several hours, or the
introduction of additives such as alkylthiols or NaBH4, or replacement by SeO2. In the present work, we
elucidate the fundamental steps in the dissolution of elemental Se in 1-ODE and oleylamine with the
view of facilitating the large-scale synthesis of CdSe QDs. The main organoselenium species generated
during the solubilisation of elemental Se in 1-ODE and oleylamine were identified by 1D and 2D NMR
spectroscopy (1H, 13C and 77Se). Experimental evidence suggests that the rate determining step is the
formation of Se radicals, via homolytic cleavage of the Se–Se bond, that attack the allylic proton in 1-
ODE and oleylamine. Plausible reaction pathways in both systems are proposed. Finally, we demonstrate
that the radical-mediated solubilisation of Se can be significantly accelerated by the addition of
azobisisobutyronitrile (AIBN), a common radical initiator used in the polymer industry. In this way, a
highly concentrated Se precursor was prepared. The “hot-injection” of the Se precursor into CdO
containing 1-ODE solution leads to the formation of highly luminescent CdSe QDs with a well-defined
cubic structure.
Received 17th January 2014
Accepted 18th February 2014
DOI: 10.1039/c4ta00285g
www.rsc.org/MaterialsA
have been extensively used.6 However, these phosphines are
highly ammable and corrosive,7 prompting the search for
1. Introduction
alternative approaches.
The optical properties of CdSe quantum dots (QDs) have
generated a tremendous amount of interest due to their
potential exploitation in a variety of elds including light-
emitting devices,1 imaging probes2 and absorber layers in
photovoltaic devices.3 Colloidal synthesis is widely recog-
nized as key for large-scale fabrication of highly mono-
dispersed and stable QDs. The seminal work by Murray et al.4
introduced the methodology of injecting phosphine–chalco-
genide precursors into an organometallic solution at
temperatures above 300 ꢀC.4b Another important contribution
in the eld was made by Peng and co-workers, replacing the
highly unstable organometallic precursors by metal oxides
(e.g. CdO) in the presence of stearic acid.5 With regard to the
solubilization of the chalcogenide component, Lewis bases
such as tributylphosphine (TBP) or trioctylphosphine (TOP)
Mulvaney and co-workers reported the rst phosphine-free
synthesis of CdSe, based on the dissolution of Se in 1-octade-
cene (1-ODE) aer prolonged periods of heating at high
temperatures.8 Subsequently, other research groups have
employed similar approaches to synthesize phosphine-free
nanocrystals.9 Solvents such as olive oil,10 paraffin,11 and oleyl-
amine12 have also been implemented. In addition to long time
heating protocols aimed at facilitating the dissolution of Se,
additives including NaBH4 and alkylthiols14 have been intro-
13
duced. The disadvantage of the latter approaches is the intro-
duction of additional elements (e.g. Na, B, and S) which could
appear as impurities in the nal product. Recent studies have
demonstrated that the synthesis of QDs can be carried out by
employing insolubilized Se15 or SeO216 powder with good yields.
In these approaches, the reported Se concentration in precursor
solution is not higher than 0.2 mol dmꢁ3, which represents a
bottleneck towards scale-up. In order to make the reaction
suitable to the industrial scope, a good understanding of the Se
dissolution process is crucial. Raston and co-workers proposed
that Se–Se and Se–C bonding compounds are the main species
aer the dissolution of Se into 1-ODE.17 However, the
aSchool of Chemistry, University of Bristol, Cantocks Close, Bristol BS8 1TS, UK; Web:
http://www.chm.bris.ac.uk/pt/electrochemistry. E-mail: David.Fermin@bristol.ac.uk;
Tel: +44 (0)117 9288981
bSchool of Physics, University of Bristol, Tyndall Avenue, Bristol BS8 1TL, UK
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c4ta00285g
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6879–6886 | 6879

View Article Online
Journal of Materials Chemistry A
Paper
mechanism of formation and nature of the Se precursors in under Ar, leading to the dissolution of Se (clear yellow solution).
these solvents remain to be fully elucidated.
CdO (0.0254 g), stearic acid (0.228 g) and 1-ODE (20 cm3) were
In this work, we investigate the process involved in the introduced into a 250 ml three-neck round bottom ask and
ꢀ
dissolution of Se in 1-ODE and oleylamine by NMR, FTIR and dried in a vacuum at 100 C for 30 min. The metal-precursor
mass spectroscopy, with the aim of optimizing this key step solution was subsequently heated to 320 ꢀC under an Ar
which is considered the bottleneck in the phosphine-free atmosphere. Once the solution became colourless, the Se
synthesis of CdSe QDs.12b,14 Our results conrm that the process precursor was quickly injected into the cadmium solution. The
is initiated by the attack of thermally generated Se radicals to particles were allowed to grow for different times at 320 ꢀC, aer
the allylic position of 1-ODE upon heating at temperatures which the heating mantle was removed. The QDs were puried
above 200 C, in agreement with previous studies.17 We show by dissolving in a methanol–hexane mixture (v/v ¼ 3 : 1) and
ꢀ
that this reaction manifests itself by the migration of the centrifuged for 10 min to remove 1-ODE, followed by addition of
terminal olen group in 1-ODE and cis/trans isomerisation in acetone and further centrifugation. The precipitate was dis-
oleylamine, generating a range of organoselenium structures. solved in chloroform for further characterization.
Furthermore, we show for the rst time that Se dissolution in
1-ODE can be signicantly accelerated by addition of trace
amounts of the radical initiator azobisisobutyronitrile (AIBN, I).
This approach allows not only the preparation of Se precursor
solutions with concentrations as high as 1.0 mol dmꢁ3, but also
the facile synthesis of highly luminescent CdSe QDs under
phosphine-free conditions yielding phase-pure monodispersed
cubic nanocrystals.
2.4. NMR, FTIR and electron microscopy methodologies and
instrumentation
Reactions requiring anhydrous conditions were performed
under an atmosphere of either anhydrous nitrogen or argon. All
glassware was ame-dried prior to use and glass syringes and
needles were placed in an oven (150 ꢀC) for at least 2 h and
allowed to cool in desiccators under an atmosphere of anhy-
drous nitrogen. Aliquots from the reaction mixture were
collected at different stages of the heating program under an Ar
atmosphere and samples were allowed to cool to room
temperature before analysis. 1D, 2D 1H NMR and 13C NMR
spectra from 2D proton correlation spectroscopy (COSY),
carbon–proton heteronuclear single quantum correlation
(HSQC) and carbon–proton heteronuclear multiple-bond
correlation spectroscopy (HMBC) were recorded in a 400 MHz
Varian INOVA 400 instrument. Chemical shis are quoted in
parts per million (ppm) and referenced to SiMe4 (1H NMR
0 ppm) and CDCl3 (13C NMR 77.16 ppm). The 13C NMR
sequence parameters were set up as follows: a 45ꢀ pulse of 7.70
ms, s2 ¼ 1.0 s, and an acquisition time of 1.28 s. The 1D and 2D
77Se NMR spectra including 1D proton decoupled 77Se NMR and
2D proton–selenium HMBC were recorded in a 500 MHz Varian
INOVA 500 instrument. The 77Se NMR sequence parameters
were set up as follows: 45ꢀ pulse, s2 ¼ 1.0 s, and an acquisition
time of 0.860 s. A saturated solution of diphenyldiselenide in
CDCl3 was used as an external standard.19 FTIR spectra were
obtained using a Perkin-Elmer Spectrum 100 FTIR Spectrom-
eter. The structure of the CdSe dots was investigated by powder
2. Experimental
2.1. Se dissolution in 1-ODE and oleylamine
The dissolution of elemental Se (3 mmol) into 1-ODE (9 mmol,
2.88 cm3) or oleylamine (9 mmol, 4.23 cm3) was investigated by
taking aliquots of the reaction mixture during ꢀthe following
heating program: room temperature (step I), 60 C for 30 min
ꢀ
under vacuum (step II), 130 C for 30 min under vacuum (step
III), 200 ꢀC for 30 min (step IV) and 20 h (step V) under Ar. Each
aliquot was analysed by 1D 1H, 13C and 2D NMR, as well as FTIR
spectroscopy. The concentration of the nal Se solution in
1-ODE or oleylamine is 1.04 and 0.71 mol dmꢁ3, respectively.
2.2. Facile Se dissolution promoted by AIBN
˚
XRD (BRUKER D8, Cu Ka radiation l ¼ 1.54 A) and JEOL 2011
Following a similar procedure as in the previous experiment, a
mixture of Se and 1-ODE (1.04 mol dmꢁ3) was prepared with the
addition of 0.6 mg of AIBN at room temperature. The compo-
sition of the solution was investigated at room temperature
under Ar (step I) and aer heating at 200 ꢀC for 1 h (step II), 2 h
(step III) and 3 h (step IV).
high resolution TEM operating at 200 kV tted with an EDX
Oxford Instruments ISIS 300 system. Selected area electron
diffraction (SAED) analysis was performed using a Philips
EM430 instrument with a camera length set at 270 mm and the
˚
electron wavelength at 0.251 A.
2.3. Phosphine-free synthesis of CdSe QDs
2.5. Electrospray ionization (ESI) and negative ion matrix
assisted laser desorption ionization time-of-ight (MALDI-
TOF) mass spectrometry (MS) analysis of Se-containing
product
The QDs were prepared based on the method reported by Peng18
with a commonly employed Se precursor concentration (0.1 mol
dmꢁ3). Typically, Se (0.0395 g) and 1-ODE (5 cm3) are loaded
into a 50 cm3 two-neck ask and heated at 60 ꢀC under vacuum In the 1-ODE system, the Se containing products were isolated
for 1 h, followed by addition of 0.001 g of AIBN (1% mol molꢁ1). by silica gel ash column chromatography using hexane as the
The temperature is subsequently increased to 200 ꢀC for 1 h eluent. In the oleylamine system, the Se containing products are
6880 | J. Mater. Chem. A, 2014, 2, 6879–6886
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
isolated through vacuum distillation using a Kugerlrohr appa- of the IR bands at n ¼ 991 and 908 cmꢁ1 and the emergence of a
ratus. The isolated products were dried under vacuum and band at n ¼ 964 cmꢁ1 (Fig. 1e) are also consistent with the
re-dissolved into dichloromethane for ESI+ analysis. MALDI formation of 2-ODE.
analyses were recorded using a HP MALDI instrument (Applied
Biosystems 4700 Proteomics Analyzer) using gentisic acid as the V aer purication by chromatography. The NMR signals with
Fig. 2 shows the 77Se NMR analysis of the Se precursor at step
matrix.
chemical shis of 416.4 (singlet), 373.1 (singlet), 373.3 (singlet),
307.4 (singlet) and 259.2 ppm (singlet) are consistent with dia-
lkyl polyselenides species.20 1H, 13C, 77Se 1D and 2D NMR
analyses (Fig. 3 and ESI [S4]†) conrmed the formation of a
3. Results and discussion
3.1. Analysis of the Se precursor in 1-ODE
Fig. 1 displays NMR and FTIR spectra of the Se–1-ODE mixture
at various heating steps. 13C NMR spectra show the emergence
of new signals at d ¼ 124.4 and 131.7 ppm (Fig. 1a) aer heating
at 200 ꢀC for 20 h under an Ar atmosphere (step V). 1H NMR data
also revealed the appearance of a new multiplet signal at d ¼
5.42 ppm at step V (Fig. 1b). The appearance of these signals
occurs only aer the complete dissolution of Se in 1-ODE. A
comparison of the 2D HSQC NMR spectra of the solutions at
steps I (Fig. 1c) and V (Fig. 1d) reveals the appearance of cis and
trans carbon–proton correlations in the latter step, suggesting
the migration of the terminal olen moiety to form 2-octade-
cene (2-ODE). This analysis is supported by the 2D COSY NMR
spectrum of sample V showing that the olenic protons (d ¼
5.41 and 5.43 ppm) are correlated with protons found at d ¼ 1.63
and 1.94 ppm (see ESI [S3]†), while the HSQC spectrum in
Fig. S4† indicates that the protons at d ¼ 1.63 and 1.94 ppm are
bonded with carbon atoms at d ¼ 18.0 and 32.7 ppm, which are
attributed to CH2 and CH3 moieties. Furthermore, the decrease
Fig. 2 Proton decoupled 77Se NMR spectrum of the purified Se
intermediate generated after dissolution of 1-ODE after 20 h at 200 ꢀC
under Ar. The inset shows two closed peaks with chemical shifts of
373.3 and 373.1 ppm.
Fig. 1 13C-NMR analysis of Se in 1-ODE at room temperature (step I),
60 ꢀC for 30 min under vacuum (step II), 130 ꢀC for 30 min under
vacuum (step III), 200 ꢀC for 30 min (step IV) and 20 h (step V) under Ar
(a). 1H-NMR spectra of the solution precursor at steps I and V, as well as
of 1-ODE after heating at 200 ꢀC for 20 h under Ar in the absence of Se
(b). The highlighted areas in (a) and (b) correspond to the signal
associated with the olefin functional group. HSQC 2D NMR spectra of
Se containing solutions at step I (c) and step V (d). The arrows indicate
the correlation between 13C–1H from cis and trans olefin moieties.
FTIR analysis of the precursor solutions at steps I to V (e). The arrows
indicate a vibrational mode associated with the olefin moiety in trans Fig. 3 1H and 77Se HMBC NMR analysis of Se containing species. As
2-octadecene, while the dotted line indicates the position of the indicated in the image, 77Se NMR signals at d ¼ 373.3 and 373.1 ppm
equivalent vibrational model in 1-ODE.
correlate with the 1H multiplet at d ¼ 1.73 and 1.54 ppm.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6879–6886 | 6881

View Article Online
Journal of Materials Chemistry A
Paper
carbon–selenium bond during the solubilisation process. The
species formed are a mixture of structural isomers with similar
properties, complicating their isolation. A thorough study of the
mixture shows, as a common feature, the presence of an olenic
group alpha to the C–Se bond.
1
Fig. 3 shows the 2D H and 77Se HMBC NMR spectra of Se
containing species, conrming the Se bonding to a secondary
carbon. The 2D 1H and 13C HMBC NMR spectra in Fig. S5 and S6
(see ESI [S4]†) suggest that the secondary carbon bound to Se is
Scheme 1 Proposed mechanism for the thermally induced Se radical
attack of the allylic proton in 1-ODE, leading to the dissolution of Se
and the migration of the terminal olefin. Highlighted in the blue frame
is the olefinic group alpha to the C–Se bond which is a common
feature present in the organoselenium species.
1
next to an olenic carbon and a saturated carbon. The 2D H
and 13C HSQC NMR spectra in Fig. S7 (see ESI [S4]†) further
conrm that the olenic moiety is vicinal to the C–Se bond and
(II) is proposed to be the general core structure.
3.2. Dissolution of Se into oleylamine
Mass values collected from ESI and MALDI MS analyses
suggested the species (III) as plausible structures being formed.
Detail mass spectra can be found in the ESI (S1 and S2).† A
correlation of NMR and mass spectrometry analysis results is
listed in Table S1.†
Based on the analysis of the NMR, FTIR, and MS data shown
in Fig. 1–3 and Table S1,† the dissolution of Se in 1-ODE can be
rationalized in terms of a radical process thermally initiated by
the homolytic cleavage of the Se–Se bond as highlighted in
Scheme 1. The key step is the attack of the allylic proton in 1-ODE
by Se radicals leading to the migration of the double bond and
subsequent H transfer from the Se intermediate species.17,21 This
is further supported by subsequent experiments where no
isomerization is observed upon extensive heating of 1-ODE in the
absence of Se (shown in Fig. 1b and also ESI [S5]†). Moreover, no
dissolution of Se takes place under similar conditions if 1-ODE is
replaced by octadecane, which lacks the alkene functionality.
In addition to 1-ODE, oleylamine is another solvent oen used
for the preparation of Se precursors.12a In our recent work,
organoselenium species were detected by employing mass
spectroscopy and 1H and 13C NMR spectroscopy, during the Se
dissolution in oleylamine.22 Although isomerization of the
double bond was identied in this instance, the depletion of the
olen moiety was also observed, which did not occur in the case
of 1-ODE.22 Fig. 4 shows the HSQC NMR analysis of Se con-
taining species at steps I and IV. A signicant decrease in the
signals at 13C d ¼ 130 ppm and 1H d ¼ 5.35 ppm upon heating at
ꢀ
200 C over prolonged periods of time (step IV, highlighted in
Fig. 4b) is consistent with our previous reported results.22 The
disappearance of this signal occurs concomitantly with the
dissolution of Se.
Fig. 5 displays FTIR spectra of Se containing species at steps I
to IV. The cis and trans isomerization is detected in the case of
Se dissolution. As shown in Fig. 5, cis hydrogen out of plane
deformation (as highlighted, ‘band F’, step I) almost dis-
appeared and an increase of trans (band ‘E’) is observed aer
the total dissolution of Se (step IV). This observation is consis-
tent with the HSQC NMR results shown in Fig. 4 and our
previous report which further conrmed the disappearance of
the olenic moiety during the Se dissolution in oleylamine.22
Fig. 6 shows the 77Se NMR spectrum of the Se precursor
generated at step V. Two major signals at d ¼ 336.2 and 525.6
ppm show the characteristic peaks of polyalkyl polyselenides
compounds.20 This result is similar to our previous ndings on
Se dissolution into 1-ODE. Moreover, the 77Se NMR spectrum
shows the presence of a multiplet (highlighted in the inset
image, d ¼ 336.2 ppm) indicating the formation of the polyalkyl
polyselenides structure.
1
Fig. 7 displays the 2D H and 77Se HMBC analysis of the Se
precursor in oleylamine. The main 77Se signal at d ¼ 336.2 ppm
correlates with protons at chemical shis of d ¼ 2.81, 1.55 and
1.32 ppm. It should be noted that these results are comparable
6882 | J. Mater. Chem. A, 2014, 2, 6879–6886
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 4 2D HSQC NMR analysis of (a) Se and oleylamine mixture before
heating; (b) Se and oleylamine after heating at 200 ꢀC for 20 h.
to those in Fig. 3, except for the new signal at 2.81 ppm.
1
Moreover, the H and 13C HMBC NMR spectra in Fig. S9a (see
ESI [S6]†) suggest that the secondary carbon bound to Se is next
to carbon atoms with chemical shis at d ¼ 26.79 and 47.24
ppm. The HSQC spectrum in Fig. S9b† shows these carbon
atoms are CH and CH2 moieties. The analysis also shows that
CH is attached to nitrogen (1H at d ¼ 2.76 and 13C d ¼ 26.78
ppm). Therefore, besides structure II and polyalkyl polyselenide
structures proposed previously,22 structure IV is proposed to be
one of the other core structures formed aer consumption of
the olen moiety.
The disappearance of the olen group during Se dissolution
in oleylamine can be explained by nucleophilic attack of oleyl-
amine. Besides the double bond, oleylamine also possesses a
nucleophilic amine group capable of reacting with Se23 forming
Fig. 6 Proton decoupled 77Se NMR spectrum of the purified Se
intermediate generated after dissolution in oleylamine after 20 h at
200 ꢀC under Ar. The inset shows the enlarge image of the multiplet
peak at d ¼ 336.2 ppm.
N–Se intermediates able to form N-centred radical species. occurs with subsequent addition to the alkyl chain as shown in
Therefore, the organoselenium structures generated in the Scheme 2.
oleylamine approach are expected to be more complicated.
Based on the NMR and FTIR analyses and previous results, a
plausible reaction mechanism for the depletion of the olen
3.3. Se dissolution into 1-ODE promoted by AIBN
moiety during Se dissolution in oleylamine is proposed in
Scheme 2. A radical process is thermally initiated by the
homolytic cleavage of the Se–Se bond as highlighted in Scheme
2. The key step is the attack on the cis allylic proton in oleyl-
amine by Se radicals, leading to the isomerization of the double
bond and subsequent H transfer from the Se intermediate
species.21 Concomitantly, nucleophilic addition from the amine
On the basis that Se solubilisation in 1-ODE occurs via radical
species, we investigated whether trace amounts of AIBN could
accelerate this process. For instance, recent studies by Yu et al.
have shown that radical cycloaddition reactions involving
elemental chalcogens can be accelerated in the presence of
AIBN.24 Fig. 8 shows clear evidence of migration of the olen
group, and the associated Se dissolution, just aer 1 h heating
Fig. 5 FTIR analysis of the reaction of oleylamine with Se (I) oleylamine mixed with Se before heating; (II) oleylamine and Se mixture after heating
at 60 ꢀC under vacuum for 30 min; (III) oleylamine and Se after heating at 130 ꢀC under vacuum for 30 min. (IV) Oleylamine and Se after heating at
200 ꢀC under Ar for 30 min. (V) Oleylamine and Se after heating at 200 ꢀC under Ar for 20 h.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6879–6886 | 6883

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 8 13C-NMR (a) and 1H-NMR (b) analyses of the Se–1-ODE mixture
at room temperature (step I) and after loading AIBN and heating at 200
ꢀC for 1 (step II), 2 (step III) and 3 h (step IV) under Ar. HSQC 2D NMR
spectra of the mixture at steps I (c) and IV (d). FTIR spectra of the
solutions at various stages (e). The main features of the various spectral
responses are identical to those highlighted in Fig. 1, although AIBN
containing solution promotes Se solubilisation within 1 h.
Fig. 7 1H and 77Se HMBC NMR spectra of the Se precursor in oleyl-
amine. As indicated in the image, 77Se (multiplet, d ¼ 336.17 ppm)
correlates with protons 1H (multiplet) at d ¼ 2.81, 1.55 and 1.32 ppm.
initiators.25 It should be mentioned that, in the absence of Se,
the migration of the olen moiety is not observed when AIBN is
present. Consequently, it could be concluded that the radical
initiator assists the generation of Se radicals which subse-
quently attack g-H in 1-ODE as suggested in Scheme 1.
Reported times for Se dissolution in 1-ODE at various
concentrations and temperatures are summarised in Table S2
(ESI†). The time required for solubilisation of 0.1 mol dmꢁ3
signicantly decreases from 48 h at 150 ꢀC to between 1 and 2 h
ꢀ
at 200 C. Our measurements in the presence of 1 mg of AIBN
show slightly shorter times for this Se concentration. The
remarkable point is observed at concentrations in the range of
1.0 mol dmꢁ3, where the dissolution time is decreased from 20
to 3 h in the presence of AIBN. This temperature is still below
the melting point of Se (221 C).26 Furthermore, the main by-
ꢀ
products of AIBN decomposition are traces of short chain
volatile alkanes which do not compromise the purity of the nal
material.
3.4. CdSe QD synthesis employing Se precursors generated
in the presence of AIBN
Scheme 2 Proposed mechanism of the thermal induced Se radical
attack on oleylamine, leading to the dissolution of Se and the disap-
pearance of the olefin group.
CdSe QDs showing a high degree of monodispersity and strong
absorption-luminescence overlap were obtained employing Se
precursors in 1-ODE in the presence of AIBN as shown in Fig. 9.
The TEM micrograph in Fig. 9a shows highly monodispersed
3.0 ꢂ 0.2 nm dots obtained aer 15 min growth time. The inset
at 200 ꢀC in the presence of AIBN. Indeed, the 13C NMR signals in Fig. 9a displays absorption and luminescence spectra
at d ¼ 124.4 and 131.7 ppm (Fig. 8a) as well as the 1H NMR featuring a strong spectral overlap and a progressive decrease of
multiplet at d ¼ 5.42 ppm (Fig. 8b) can be observed aer 1 h as the optical band gap with increasing growth time.
opposed to 20 h in the absence of AIBN (Fig. 1). The migration
Selected area electron diffraction pattern rings in Fig. 9b are
of the double bond is further conrmed by HSQC NMR (Fig. 8c characteristic of the zinc blende (ZB) unit cell, with the 111, 220
and d) as well as FTIR spectroscopy (Fig. 8e), suggesting that and 311 rings clearly identiable although broadened by the
AIBN effectively promotes the dissolution of Se. AIBN decom- nanoscopic dimension of the crystals. The experimental data
poses at temperatures above 65 ꢀC, eliminating molecular correlate closely with simulations of the diffraction patterns for
nitrogen and generating 2-cyanoprop-2-yl radicals acting as 3 nm CdSe nanocrystals.27 X-ray powder diffraction (Fig. 9c) is
6884 | J. Mater. Chem. A, 2014, 2, 6879–6886
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
In the case of 1-ODE, the most plausible intermediates in
this process are identied as 1,3-di(octadec-3-en-2-yl)mono-
selenide, 1,3-di(octadec-3-en-2-yl)diselenide and 1,3-di(o-
ctadec-3-en-2-yl)triselenide. On the other hand, the intrinsic
nucleophilic properties of oleylamine facilitate the reaction
with the double bond during Se solubilisation, inducing the
formation of polyalkyl aminoselenides besides the polyalkyl
polyselenide species.
For the rst time, we show that addition of trace amounts of
AIBN signicantly accelerates the solubilisation of Se in 1-ODE
in the molar concentration range. Studies carried out at various
concentrations and temperatures indicate that Se solubilisation
requires temperatures above 150 ꢀC, even in the presence of
AIBN. Although AIBN promotes faster Se dissolution in the 0.1
mol dmꢁ3 concentration range, the strongest effects are
observed at signicantly higher Se concentrations.
The utilisation of AIBN as a promoter in the generation of Se
precursors allows the facile phosphine-free synthesis of high-
quality zinc-blende CdSe QDs without the need for further
additives. The crystal structure of the dots contrasts with the
strong wurtzite character of the materials obtained in the
presence of phosphine groups.28 We strongly believe that opti-
Fig. 9 Characteristic TEM micrograph (scale bar 20 nm) of 3.0 ꢂ 0.2
nm CdSe dots after 15 min growth (a). The inset shows the absorption
and luminescence spectra recorded after 3 (blue), 8 (grey) and 15 min mization of the chalcogenide precursor's composition by
(red) growth times. Experimental and simulated selected area electron
diffraction patterns (b), powder XRD (c) and high resolution TEM image
(d, scale bar 2 nm) of CdSe dots. The XRD pattern of cubic CdSe (PDF
adjusting the concentration of radical initiators, such as AIBN,
can have a signicant impact on scaling up the synthesis of
these materials and their deployment in solution-processable
photovoltaic technologies.
19-0191) is also shown in (c). The inset in (d) shows a 3.0 nm dot,
highlighting the 70 ꢂ 2ꢀ cross-grating originating from adjoining (111)
planes in fcc crystals with a lattice spacing of 3.5 ꢂ 0.2 A˚ (parallel lines).
Acknowledgements
fully compatible with phase pure ZB CdSe dots. Fig. 9d shows a
The authors are grateful to Mr Jonathan Jones, Mr Paul Law-
rence and Dr Craig Butts for their support to this research. This
research is supported by the Engineering and Physical Sciences
Research Council (EP/G031088/1, SUPERGEN Consortium on
Excitonic Solar Cells (DJF and BH) and EP/J002542/1 (MCG and
DBA)).
typical high resolution TEM image obtained from as-prepared 3
nm CdSe QDs, featuring a cross-grating pattern generated by
adjoining (111) planes with the 70 ꢂ 2ꢀ angle. The measured
˚
lattice spacing of the (111) plane of 3.5 ꢂ 0.2 A is close to the
value expected for cubic CdSe (the inset image of Fig. 9d).
These results show that the presence of the trace amount of
AIBN does not affect the nucleation of CdSe, demonstrating its
suitability for the preparation of Se precursors for QD synthesis.
The ability of preparing chalcogenide precursors with high
concentration and short times is crucial for deployment of
scalable routes for the production of thin-lm photovoltaic
devices. The introduction of radical initiators in phosphine-free
solvents, along with the temperature and concentration,
provides an additional optimisation parameter that can be
exploited in this direction.
References
1 Y. Shirasaki, G. J. Supran, M. G. Bawendi and V. Bulovic, Nat.
Photonics, 2013, 7, 13–23.
2 (a) F. A. Esteve-Turrillas and A. Abad-Fuentes, Biosens.
Bioelectron., 2013, 41, 12–29; (b) F. Wang, X. Liu and
I. Willner, Adv. Mater., 2013, 25, 349–377; (c) X. Michalet,
F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li,
G. Sundaresan, A. M. Wu, S. S. Gambhir and S. Weiss,
Science, 2005, 307, 538–544.
4. Conclusions
3 (a) L. Etgar, Materials, 2013, 6, 445–459; (b) P. V. Kamat,
J. Phys. Chem. C, 2008, 112, 18737–18753.
Our studies have unveiled the processes taking place during the
dissolution of Se in 1-ODE and oleylamine, an important
bottleneck in the scale-up of phosphine-free synthesis of CdSe
QDs. The process is initiated by the generation of Se radicals
which attack the allylic proton in both solvents. This reaction
manifests itself by the migration of the olen group in 1-ODE
and isomerization in oleylamine.
4 (a) T. Trindade, P. O'Brien and N. L. Pickett, Chem. Mater.,
2001, 13, 3843–3858; (b) C. B. Murray, D. J. Norris and
M. G. Bawendi, J. Am. Chem. Soc., 1993, 115, 8706–8715.
5 (a) Z. A. Peng and X. Peng, J. Am. Chem. Soc., 2000, 123, 183–
184; (b) L. Qu, Z. A. Peng and X. Peng, Nano Lett., 2001, 1,
333–337.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6879–6886 | 6885

View Article Online
Journal of Materials Chemistry A
Paper
´
6 (a) C. de Mello Donega, P. Liljeroth and D. Vanmaekelbergh, 17 (a) C. Bullen, J. van Embden, J. Jasieniak, J. E. Cosgriff,
Small, 2005, 1, 1152–1162; (b) K. L. Sowers, B. Swartz and
T. D. Krauss, Chem. Mater., 2013, 25, 1351–1362.
7 B. Stewart, A. Harriman and L. J. Higham, Organometallics,
2011, 30, 5338–5343.
R. J. Mulder, E. Rizzardo, M. Gu and C. L. Raston, Chem.
´
´
Mater., 2010, 22, 4135–4143; (b) R. Garcıa-Rodrıguez,
M. P. Hendricks, B. M. Cossairt, H. Liu and J. S. Owen,
Chem. Mater., 2013, 25, 1233–1249.
8 J. Jasieniak, C. Bullen, J. van Embden and P. Mulvaney, 18 W. W. Yu and X. Peng, Angew. Chem., Int. Ed., 2002, 41, 2368–
J. Phys. Chem. B, 2005, 109, 20665–20668. 2371.
9 (a) H. Shen, H. Wang, Z. Tang, J. Z. Niu, S. Lou, Z. Du and 19 J. D. Odom, W. H. Dawson and P. D. Ellis, J. Am. Chem. Soc.,
L. S. Li, CrystEngComm, 2009, 11, 1733–1738; (b) M. Sun 1979, 101, 5815–5822.
and X. Yang, J. Phys. Chem. C, 2009, 113, 8701–8709; (c) 20 H. Eggert, O. Nielsen and L. Henriksen, J. Am. Chem. Soc.,
L. Liu, Z. Zhuang, T. Xie, Y.-G. Wang, J. Li, Q. Peng and
Y. Li, J. Am. Chem. Soc., 2009, 131, 16423–16429.
10 S. Sapra, A. L. Rogach and J. Feldmann, J. Mater. Chem.,
2006, 16, 3391–3395.
1986, 108, 1725–1730.
21 R. M. Hoyte and D. B. Denney, J. Org. Chem., 1974, 39, 2607–
2612.
22 B. Hou, D. Benito-Alifonso, N. Kattan, D. Cherns,
´
11 (a) Z. Deng, H. Yan and Y. Liu, J. Am. Chem. Soc., 2009, 131,
17744–17745; (b) B. Xing, W. Li, X. Wang, H. Dou, L. Wang,
M. C. Galan and D. J. Fermın, Chem. – Eur. J., 2013, 19,
15847–15851.
K. Sun, X. He, J. Han, H. Xiao, J. Miao and Y. Li, J. Mater. 23 J. Lu, Y. Xie, F. Xu and L. Zhu, J. Mater. Chem., 2002, 12,
Chem., 2010, 20, 5664–5674. 2755–2761.
12 (a) S. Mourdikoudis and L. M. Liz-Marzan, Chem. Mater., 24 L. Yu, Y. Wu, T. Chen, Y. Pan and Q. Xu, Org. Lett., 2012, 15,
2013, 25, 1465–1476; (b) Y. Wei, J. Yang, A. W. H. Lin and
J. Y. Ying, Chem. Mater., 2010, 22, 5672–5677.
13 S. C. Riha, B. A. Parkinson and A. L. Prieto, J. Am. Chem. Soc.,
2011, 133, 15272–15275.
144–147.
25 M. B. Smith and J. March, March's Advanced Organic
Chemistry: Reactions, Mechanisms, and Structure, WILEY,
6th edn, 2007.
14 Y. Liu, D. Yao, L. Shen, H. Zhang, X. Zhang and B. Yang, 26 Y. A. Yang, H. Wu, K. R. Williams and Y. C. Cao, Angew.
J. Am. Chem. Soc., 2012, 134, 7207–7210. Chem., Int. Ed., 2005, 44, 6712–6715.
15 (a) S. Flamee, M. Cirillo, S. Abe, K. De Nolf, R. Gomes, 27 P. A. Stadelmann, Ultramicroscopy, 1987, 21, 131–145.
T. Aubert and Z. Hens, Chem. Mater., 2013, 25, 2476–2483; 28 (a) W. W. Yu, Y. A. Wang and X. Peng, Chem. Mater., 2003, 15,
(b) S. Flamee, R. Dierick, M. Cirillo, D. Van Genechten,
T. Aubert and Z. Hens, Dalton Trans., 2013, 42, 12654–12661.
16 M. Ibanez, R. Zamani, A. LaLonde, D. Cadavid, W. Li,
A. Shavel, J. Arbiol, J. R. Morante, S. Gorsse, G. J. Snyder
and A. Cabot, J. Am. Chem. Soc., 2012, 134, 4060–4063.
4300–4308; (b) B. Hou, D. Parker, G. P. Kissling, J. A. Jones,
´
D. Cherns and D. J. Fermın, J. Phys. Chem. C, 2013, 117,
6814–6820; (c) Y. C. Li, H. Z. Zhong, R. Li, Y. Zhou,
C. H. Yang and Y. F. Li, Adv. Funct. Mater., 2006, 16, 1705–
1716.
6886 | J. Mater. Chem. A, 2014, 2, 6879–6886
This journal is © The Royal Society of Chemistry 2014