Morphology- and phase-controlled synthesis of...

Morphology- and phase-controlled synthesis of monodisperselanthanide-doped NaGdF4 nanocrystals with multicolor photoluminescence†

Chenghui Liu, Hui Wang, Xinrong Zhang and Depu Chen*

Received 8th September 2008, Accepted 7th November 2008

First published as an Advance Article on the web 9th December 2008

DOI: 10.1039/b815682d

Monodisperse, regular-shaped and well-crystallized nanocrystals (NCs) of lanthanide-doped NaGdF4

with diverse shapes and structures are synthesized in high boiling organic solvents 1-octadecene and

oleic acid, through a competitive nucleation and growth pathway. The NCs can be manipulated to

different morphologies and phase structures by using controlled variations in the reaction conditions

such as composition of the solvent, temperature or reaction time. The NCs are thoroughly

characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), selected area

electron diffraction (SAED), high resolution TEM (HRTEM), field emission scanning electron

microscopy (FESEM), IR spectroscopy, and photoluminescence. Possible mechanisms of NC

nucleation and growth, size and shape evolution are proposed and tested. With different dopants, the

NCs can show intensive multicolor down-conversion emissions under 254 nm UV excitation or

up-conversion fluorescence under 980 nm NIR excitation, showing great promise in applications such

as multi-analyte biolabels, staining, displays and other optical technologies.

1. Introduction

In recent years, more and more attention has been focused on the

development of new luminescent nanomaterials such as lantha-

nide-doped nanocrystals (NCs) due to their many potential

applications in optics and optoelectronics. Such lanthanide ion-

based luminescent NCs show superior chemical and optical

properties, including low toxicity, large effective Stokes shifts,

multicolor emission, as well as high resistance to photobleaching,

blinking, and photochemical degradation.1 Thus they have been

widely used for numerous applications such as biolabels,1–3 light-

emitting devices,4,5 displays,6 optical amplifiers,7 and low-

threshold lasers.8

Amongst various host materials for lanthanide-doped NCs,

fluorides provide some distinct advantages over the conven-

tionally used oxide materials owing to the very low phonon

frequencies of their crystal lattices.9 The quenching of the excited

state of the rare-earth ions will be minimized when lanthanide

ions are doped into fluoride hosts, which will lead to long life-

times of their excited states and high luminescence quantum

yields even in the case of IR emitting ions. Hence, controlled

synthesis of lanthanide doped fluoride NCs has attracted a lot of

interest and has been a popular topic for several decades.

NaGdF4 NCs, existing either in the a-phase (cubic) or the

b-phase (hexagonal), are of particular interest. On one hand,

certain lanthanide ion doped NaGdF4 NCs can display down-

conversion emission under deep UV excitation, which is of

significant technological importance, in particular in the appli-

cations of mercury-free fluorescent tubes or plasma display

panels.10,11 On the other hand, lanthanide NCs can show multi-

color emissions by varying the dopants. For many cases it is

necessary to excite all these NCs by a single wavelength irradi-

ation. However, it is very difficult to do so since each lanthanide

activator has a special set of energy levels. So the same sensitizing

ion for all the lanthanide activators is usually utilized to solve this

problem. In this case, a single-wavelength irradiation can be used

to excite the sensitizers, followed by energy transfer to the

lanthanide activators and then the fluorescence can be generated.

In particular, the NaGdF4 NCs are known to be one kind of very

efficient host lattice for such luminescent processes12 since Gd3+

can act as an intermediate to allow energy to migrate over the

Gd3+ sublattice and consequently facilitate the energy transfer

process.13

Therefore, many works have been done to obtain such

lanthanide-doped NaGdF4 NCs with controlled sizes, shapes

and phases. Hitherto, many approaches including co-precipita-

tion method, hydrothermal method, reversed micelle method and

solid state reaction have been reported to synthesis lanthanide-

doped NaGdF4 crystals. These methods are well documented

and compared in the literature.14 Although crystals produced

using these methods were generally in the nano-scale, they were

either terribly aggregated or ill-shaped. More recently, Capo-

bianco and co-workers reported the synthesis of a-phaseNaGdF4:Ce

3+,Tb3+ NCs with a diameter of about 5 nm based on

the thermo-decomposition of corresponding lanthanide tri-

fluoroacetate precursors at high temperature.15 Zhang et al. have

synthesized b-phase NaGdF4 NCs by a hydro/solvothermal

technique using polyethyleneimine as a capping reagent.12 The

NCs obtained in their work have an elongated spherical shape

Department of Chemistry, Tsinghua University, Beijing, 100084, P. R.China. E-mail: [email protected]; Fax: +86 1062782485; Tel: +86 10 62781691† Electronic supplementary information (ESI) available: EDX spectrumof NaGdF4:Ce

3+,Tb3+ (Fig. S1); TEM images and XRD patternscorresponding to the NCs obtained under different ratios of F!/RE3+

(Fig. S2); TEM images and XRD results of NCs synthesized underdifferent ratios of TOP/1-octadecene and oleylamine/1-octadecene(Fig. S3, Fig. S4); and the discussion of self-assembly properties of theluminescent NCs (Fig. S5). See DOI: 10.1039/b815682d

This journal is ª The Royal Society of Chemistry 2009 J. Mater. Chem., 2009, 19, 489–496 | 489

PAPER www.rsc.org/materials | Journal of Materials Chemistry

with a wide diameter distribution between 25 nm and 45 nm. It

can be seen that only spherical or irregular-shaped NCs were

generally obtained in these approaches. However, little attention

has been paid to the morphology control of NaGdF4 NCs. So it

still remains a challenge to establish an efficient method for

growing monodisperse, high quality crystalline NCs with well-

defined sizes/shapes by a facile and economic strategy.

In this contribution, we report on the one-pot synthesis of

regular-shaped, monodisperse and well-crystallized NaGdF4:-

Ce(Yb)3+,Ln3+ (Ln " Tb, Eu, Dy, Sm, Er or Tm) multicolor

luminescent NCs through a competitive nucleation and growth

pathway. The NCs obtained are dispersible in nonpolar organic

solvents and the morphologies and phase structures of the

obtained NCs can be easily tuned by changing the experimental

conditions such as the solvent, the temperature or the reaction

time. The NCs were characterized using TEM, HRTEM, SAED,

FESEM, IR spectroscopy, XRD and photoluminescence and

possible mechanisms of the NC nucleation and growth, phase

selection, size and shape evolution are proposed.

2. Experimental

2.1. Reagents and materials

Sodium fluoride (NaF) was obtained from Yili Chemical Corp.

(Beijing, China). Rare earth (RE) oxides of SpecPure grade

(Gd2O3, Tb4O7, Eu2O3, Dy2O3, Sm2O3, Er2O3, Tm2O3, 99.99%)

were obtained from Grirem Advanced Materials Co., Ltd.

(Beijing, China). Oleic acid, sodium oleate and Ce(NO3)3$6H2O

were purchased from Sinopharm Chemical Reagent Co., Ltd.

(Shanghai, China). 1-Octadecene and oleylamine were purchased

from Acros Organics and tri-n-octyl phosphine was from Tokyo

Chemical Industry Co., Ltd. (TCI, Japan). RECl3 were prepared

by dissolving the corresponding rare earth oxides in hydrochloric

acid at elevated temperature followed by evaporating the solvent

under vacuum. RE(oleate)3 complexes were prepared by adopt-

ing the literature methods.16,17

2.2. Synthesis of the NaGdF4:Ce(Yb)3+,Ln3+ NCs

The synthesis of NaGdF4:Ce(Yb)3+,Ln3+ NCs was modified from

our previous work for the synthesis of NaYF4 nanoplates. In this

work, the procedures were further simplified by using a one-pot

synthetic approach, where no injection processes were needed.

Take the synthesis of uniform NaGdF4:Ce3+,Ln3+ short

nanorods as an example. Typically, 1 mmol of RE(oleate)3(Gd3+:Ce3+:Ln3+ " 85:10:5, Ln " Tb3+, Eu3+, Sm3+, Dy3+) and

0.21 g of NaF (5 mmol, F!/RE3+ molar ratio was kept at 5 unless

stated otherwise) were first added to the solvent of 15 mL oleic

acid/15 mL 1-octadecene simultaneously, then degassed under

vacuum for 30 min, and finally, heated rapidly to 280 #C and kept

at this temperature for 5 h under vigorous magnetic stirring in the

presence of nitrogen. Subsequently, the mixture was allowed to

cool to room temperature, and the NCs were precipitated by the

addition of ethanol and isolated via centrifugation.

NCs of different sizes, morphologies and structures were

synthesized following the similar procedures only by varying the

experimental conditions such as reaction temperature, time or

the components of the solvent.

2.3. Characterization

The as-prepared products were characterized by transmission

electron microscopy (TEM), high-resolution TEM (HRTEM)

and field emission scanning electron microscopy (FE-SEM)

using JEM-1200EX, JEM-2010 and JSM-7401F (JEOL, Japan)

microscopes with accelerating voltages of 100 kV, 120 kV and 3

kV, respectively. Samples for TEMmeasurements were prepared

by sonicating the precipitate products in hexane for 30 min and

evaporating a drop of the suspension onto a carbon-coated

copper grid. Powder X-ray diffraction (XRD) analysis was per-

formed on a D/max-2500 X-ray diffractometer (Rigaku, Japan)

with Cu Ka radiation at 1.5406 A. Fourier transform infrared

spectroscopic (FTIR) analysis was carried out by using a Perkin

Elmer spectrometer. The down-conversion luminescence spectra

of the NCs were measured on a FP-6500 fluorescence spectro-

photometer (Jasco, Japan), while the up-conversion fluorescence

spectra were measured using a LS-55 luminescence spectrometer

(Perkin-Elmer) with an external 980 nm laser diode (1 W,

continuous wave with 1 m fiber, Beijing Viasho Technology Co.)

as the excitation source. All measurements were performed at

room temperature.

3. Results and discussion

3.1. Morphology and structure of the synthesized NCs

Following the standard synthesis procedures stated above, the

NaGdF4:Ce3+,Ln3+ NCs obtained in the solvent of oleic acid/1-

octadecene (15mL/15mL) at 280 #C for 5 h were characterized by

TEM and FESEM, as shown in Fig. 1. The typical TEM images

(Fig. 1a–d, taken with different magnifications) demonstrate that

all the as-obtained NaGdF4 NCs are of single-crystalline nature

and display high crystallite size uniformity. All the NCs display

uniform short nanorod shapes. The average diameter of the NCs,

evaluated from 100 random particles from low resolution TEM

images, is (23.8 $ 1.6) nm (diameter) % (35.8 $ 1.5) nm (length),

which indicates a rather narrow size distribution. FESEM

characterizations (Fig. 1f, g) were also performed to further

illustrate the morphologies of the NCs, which also demonstrate

their size/shape-uniformity and monodispersity. It is notable that

the regular hexagonal shape of the top/bottom surfaces of the

NCs can be identified in Fig. 1e, where a higher concentration of

colloid NCs were dropped onto the TEM grid, because at high

concentration some NCs had the opportunity to stand on end

instead of lying on their sides. The hexagonal-shaped top/bottom

surfaces can also be observed in the FESEM images. The selected

area electron diffraction (SAED) pattern (Fig. 1h) shows the

spotty polycrystalline diffraction rings corresponding to the

specific (100), (110), (111), (201),(102) planes of the hexagonal

NaGdF4 lattice. The hexagonal phase structure of the NCs was

further identified by XRD analysis as shown in Fig. 1i. The peak

positions and intensities agree well with the data reported in the

JCPDS standard card (PDF 27-0699) for hexagonal NaGdF4

crystals. The atomic composition ratios of the obtained NCs

were determined by energy-dispersive X-ray analysis (EDX).

Fig. S1 (see ESI†) shows an EDX spectrum of NaGdF4:-

Ce3+,Tb3+. The atomic ratio of Gd:Ce:Tb was determined as

85.2:9.7:5.1, which is very close to the calculated value (85:10:5).

490 | J. Mater. Chem., 2009, 19, 489–496 This journal is ª The Royal Society of Chemistry 2009

In the present synthesis approach, the general process can be

depicted as follows: an anion-exchange reaction between F!

(provided by NaF), and RCOO! (provided by the RE(oleate)3)

would take place when the reaction system was settled at certain

high temperatures. Due to the strong coordination between RE

cations and carboxyl groups of the oleate anions, F! had to

compete with the RCOO! to form NaGdF4 precipitates during

the nucleation and growth processes. So the reaction was

significantly retarded by the strong chelation among the rare

earth cations and the bulky oleate anions, thereby achieving

a well-maintained balance between nucleation and growth

stages. For the oleate anions, the presence of a carboxylic group

with significant affinity to NCs surfaces together with a long

nonpolar tail group for sterical hindering makes it well bound on

the NC surfaces to prevent NCs from aggregating during the

reaction. As a result, monodisperse NCs can be obtained under

certain conditions and the as-obtained NCs can be well dispersed

in nonpolar solvents such as hexane and toluene.

3.2. Effects of reaction temperature and time on the phase

control and morphology of the NCs

It was found that the reaction temperature and time played

important roles in the synthesis of hexagonal phase NaGdF4:-

Ce(Yb)3+,Ln3+ NCs. The effects of reaction temperature and time

on the structure, morphology and sizes on the NCs were inves-

tigated by using XRD (Fig. 2) and TEM analysis (Fig. 3),

respectively. These results are summarized in Table 1.

It can be seen that high temperatures are preferred to form

b-phase NCs when other experimental conditions are

unchanged. For example, the products obtained at 280 #C

showed pure hexagonal phase even for a rather short reaction

time of 15 min (Fig. 2e), while the NCs synthesized below 260 #C

had dominant a-phase for as long as 5 h.

From the TEM images it can be concluded that with the

extension of the reaction time, the NCs tend to grow

anisotropically from mainly along the radial directions to along

the axial directions (c axis) which finally results in the rod-shaped

NCs.

As far as we know, selective adhesion of capping ligand onto

specific crystal planes is critical in the epitaxial growth of

NCs.18,19 Since 1-octadecene is a non-coordination reagent, we

speculate that oleic acid may play an important role in the shape

evolution in this work. Since oleic acid is a selective absorption

surfactant,20 it is supposed that during the growth of the NCs,

oleic acid is preferentially adsorbed onto the side surfaces that

are parallel to the c-axis of the growing crystallites, which results

in the epitaxial growth along the <001> directions and results in

nanorods. The HRTEM images of an individual nanorod show

the crystalline structures of its top/bottom surface (Fig. 3i) and

side surface (Fig. 3j, k). The HRTEM images clearly show that

Fig. 2 XRD patterns of the NCs obtained in the solvent of 15 mL oleic

acid/15 mL 1-octadecene under conditions of (a) 280 #C for 5 h; (b) 260#C for 5 h; (c) 230 #C for 5 h; (d) 200 #C for 5 h; (e) 280 #C for 15 min. The

peaks marked with asterisks are indexed to residual NaF. The excess NaF

can be removed by washing the products several times with water.

Fig. 1 (a–d) TEM images with different magnifications of the NaGdF4:Ce(Yb)3+,Ln3+ nanorods synthesized in 15 mL oleic acid/15 mL 1-octadecene at

280 #C for 5 h; (e) TEM images obtained by using colloid NCs sample with higher concentration than that of (a–d); (f–g) FESEM images of the

NaGdF4:Ce(Yb)3+,Ln3+ nanorods; (h) SAED pattern of the nanorods corresponding to the hexagonal NaGdF4 lattice. (i) XRD pattern of the

NaGdF4:Ce(Yb)3+,Ln3+ nanorods corresponding to the NCs of above TEM images.


the epitaxial growth is along the <001> direction. Meanwhile,

the oleic acid layer round the side surfaces of the nanorods can be

clearly observed in Fig. 3i highlighted by arrows. All these results

support our hypothesis. Furthermore, the shape evolution of the

NCs from nanoplates to nanorods with the extension of reaction

time also proves our theory.

Fig. 3 (a–d) TEM images of NaGdF4:Ce(Yb)3+,Ln3+ NCs synthesized in 15 mL oleic acid/15 mL 1-octadecene treated for 5 h at 200 #C, 230 #C, 260 #C,

and 280 #C, respectively; (e–h) TEM images of NaGdF4:Ce(Yb)3+,Ln3+ NCs synthesized in 15 mL oleic acid/15 mL 1-octadecene treated at 280 #C for 2.5

h, 1 h, 15 min, 10 h, respectively; (i) HRTEM image of the top/bottom surface of a single nanorod in image (d). The oleic acid layer (indicated by arrows)

adsorbed round the side surfaces can be clearly observed; (j) HRTEM image of the side surface of the nanorod as in image (i); (k) magnified picture of the

selected area in image (j). The scalar bars stand for 10 nm in images (i) and (j).

Table 1 Effects of reaction temperature and time on NC synthesisa

Temperature/#C Time/hour Phase Shape Size/nm

200 5 cubic cannot be identified very tiny230 5 cubic cannot be identified very tiny260 5 cubic/hexagonal a few regular-shaped NCs; most are

very tiny particles&20–30 nm for the regular-shaped

NCs280 1/4 hexagonal nanoplates diameter % length (or thickness)

(30.1 $ 2.3)nm%(14.0 $ 1.2)nm280 1 hexagonal dot-like nanorods diameter % length (or thickness)

(16.8 $ 0.6)nm%(18.7 $ 0.5)nm280 2.5 hexagonal nanorods diameter % length (or thickness)

(17.5 $ 1.1)nm%(28.3 $ 1.0)nm280 5 hexagonal nanorods diameter % length (or thickness)

(23.8 $ 1.6)nm%(35.8 $ 1.5)nm280 10 hexagonal sphere-like particles diameters among 20–25 nm

a Other conditions are the same: solvent is 15 mL oleic acid/15 mL 1-octadecene.


However, it is worth noting that when the reaction time

exceeded 10 h, the shape of the NCs changed from rod to sphere-

like with diameters in the range of 20–25 nm (Fig. 3h). The

reason will be discussed in the section ‘‘Formation and growth

mechanisms of the NCs’’.

3.3. Effects of the amount of oleic acid in the solvent

As mentioned above, oleic acid was speculated to play an

important role in the NC growth and shape evolution. So the

effect of oleic acid on the NC synthesis was explored with other

experimental conditions (280 #C, 5h) fixed.

The results of XRD analysis show that the NCs obtained

under different ratios of oleic acid/1-octadecene all exhibit pure

hexagonal phase (Fig. 4). But the morphologies of the

NaGdF4:Ce3+,Ln3+ NCs can be strongly affected by the amount

of oleic acid in the solvents. TEM images in Fig. 4 show that with

the ratio of oleic acid/1-octadecene of the solvent increased, NCs

grew more anisotropically and the monodispersity of the NCs

increased remarkably. In pure 1-octadecene without the presence

of oleic acid, only very agglomerated NCs as well as some small

particles were formed (Fig. 4a). Under low ratios of oleic acid/1-

octadecene (below 3/27), the agglomerated NCs disappeared and

ill-shaped nanoparticles with a rather wide size distribution were

obtained (Fig. 4b, c). When the ratio exceeded 6/24, the NCs

became more and more uniform and regular-shaped (Fig. 4d–g).

As can be seen from Fig. 4e, uniform nanoparticles with an

average diameter of 21.8 $ 1.4 nm were obtained for an oleic

acid/1-octadecene ratio of 10/20. Under higher oleic acid/1-

octadecene ratios such as 15/15 and 20/10, regular-shaped short

nanorods were synthesized. The results clearly indicate that the

morphology control in this approach is closely related to the

application of oleic acid as capping surfactant, since the amount

of oleic acid is the only parameter which was changed in this

study to obtain regular-shaped NCs.

3.4. Formation and growth mechanisms of the NCs

All the above experimental results suggest that the temperature

mainly determines the phase selection, while the reaction time

and the solvents contribute little to the phase control of the NCs

but play important roles in the anisotropic growth and shape

evolution.

Although the b-phase NCs are the thermodynamically stable

crystalline form compared with the a-phase, relatively high

energy is needed to overcome the dynamical energy barrier of the

a / b phase transition. In this work, it seems that high

temperature such as 280 #C can supply sufficient energy to obtain

hexagonal phase NCs, while the a / b energy barrier cannot be

completely surpassed at 260 #C or lower temperature which will

result in dominant a-phase NCs.

As regards the morphology control, according to previous

reports,21–23 the accelerated crystallization process in solutions

related to the high monomer concentrations is the main driving

force for anisotropic growth of the inorganic nanostructures.

Similarly, a monomer concentration-controlled kinetic model in

this work was proposed and can give reasonable explanations for

the shape evolution of NCs.

When the reaction was carried out in pure 1-octadecene, the

monomer concentration in the solution was very low due to the

insolubility of NaF and the rather slow nucleation and growth

process. Therefore, we cannot obtain monodisperse and regular-

shaped NCs in pure 1-octadecene.

But when oleic acid was added, the situation was quite

different. First, in the presence of oleic acid, the polarity of the

solvent is enhanced and as a result the solubility of NaF is

enhanced. Since oleic acid is a weak acid, even more F! may be

dissolved in the solvent through a NaF–HF pathway.

Second, after the addition of oleic acid, RE(oleate)3 can be

more easily dissolved in the solvent and their mobility in the

solution is enhanced compared with that in pure 1-octadecene.

All these factors benefit the nucleation and growth rate of the

Fig. 4 Effect of oleic acid/1-octadecene ratios on the morphology and structure of NaGdF4:Ce(Yb)3+, Ln3+ NCs. TEM images of NCs synthesized

under oleic acid/1-octadecene ratios: (mL/mL, total volume of 30mL) (a) 0/30; (b) 1/29; (c) 3/27; (d) 6/24; (e) 10/20; (f) 15/15; (g) 20/10. (h) XRD results of

the NCs obtained under different ratios of oleic acid/1-octadecene (OA/ODE). Other experimental conditions are the same: 280 #C, 5 h.


NCs. So it can be concluded that the higher oleic acid/1-octa-

decene ratio leads to a higher concentration of monomers and

thus a higher growth rate of the NaGdF4:Ce(Yb)3+,Ln3+ NCs in

the solution. Under this condition, the one-dimensional (1D)

growth was significantly promoted, and thus the shape evolution

from irregular shapes to nanorods was observed. Moreover,

although the crystallization speed is accelerated in the presence

of oleic acid, due to its long alkyl tail, weak acidity, strong

coordination effect with RE ions and the competitive anion

exchange process between F! and RCOO!, the growth rate of the

NCs remains slow enough to prevent the formation of aggre-

gated products.

Third, the preferred adsorption of oleic acid on the side

surfaces of the NaGdF4:Ce(Yb)3+,Ln3+ makes the NCs growing

more anisotropically along the <001> direction (c-axis direction),

which has been demonstrated above. The FTIR spectra (Fig. 5) of

the nanorods show that no characteristic group absorption peaks

from oleic acid can be seen but two strong bands centered at 1457

cm!1 and 1563 cm!1 are observed, which can be associated with

the asymmetric and symmetric stretching vibrations of carbox-

ylate anions.24 The FTIR results indicate that the adhesion of

oleic acid on the NCs is not a merely physical adsorption, while

strong interaction may exist between the RCOO! and the RE

atoms on the surfaces of the NCs, which well blocked the growth

along the radial directions but accelerated the growth rate along

the axial direction. Furthermore, according to the Le Chatelier’s

principle, increasing the amount of oleic acid in the solvent would

slow the decomposition/desorption of oleate anions from theNCs

surfaces and thus strengthen their interactions, which would

further facilitate the anisotropic growth of the NCs. In addition,

the slowed decomposition of RE-RCOO! with the increasing

oleic acid would make the RCOO! bind more easily on the

surfaces of the growing NCs, which can further protect the NCs

from aggregation.

All the experimental results for the shape evolution can be

reasonably explained by the monomer concentration-controlled

kinetic model and the selective adhesion of oleic acid on different

parts of the NCs surfaces. For example, the shape evolution with

extended reaction time mentioned above can be illustrated in

more detail. At the early stage, the monomer concentration was

relatively high, so the selective adhesion of oleic acid on specific

surfaces was the dominant factor, and regular-shaped NCs can

be obtained. With a prolonged time of more than 10 h, the

monomer concentration was gradually decreased to a critical

level. So intra-particle moves of atoms from high-energy facets to

other facets took place, and thus corners and tips were smoothed.

Eventually the transition from rod to sphere-like particles was

observed.21,25

Further experiments were designed to examine the important

role of the monomer concentration. As stated above, only

aggregated NCs were obtained in pure 1-octadecene due to the

low monomer concentration. But when the amount of NaF was

raised to F!/REmolar ratios of 16 or 24, monodisperse NCs were

obtained with nearly hexagonal shapes (Fig. S2a, S2b†).

Although the solubility of NaF is poor in 1-octadecene, high F!/

RE molar ratios will raise the F! concentration, and enhance the

chance of nucleation and growth of the NCs at the solid/liquid

interfaces. So the monomer concentration as well as the crys-

tallization speed are higher than that of the reaction system with

F!/RE molar ratio of 5. Therefore more regular-shaped and

uniform NCs are obtained (b-phase, XRD results see Fig. S2c†).

The results further supported the proposed monomer concen-

tration-controlled kinetic model.

To further prove the crucial effect of oleic acid on the nano-

crystal growth and crystallization, two other high boiling point

organic ligands, tri-n-octyl phosphine (TOP) and oleylamine,

mixed with the noncoordinating 1-octadecene were employed as

solvents for comparison with oleic acid.

TOP is a widely used capping reagent especially for the

synthesis of quantum dots.26,27 Although TOP has been used as

a stabilizer for the preparation of nanomaterials, it can not

play the same role as oleic acid in this study. First, TOP is

a Lewis base which cannot facilitate the dissolution of F! and

RE(oleate)3 and hence increase the monomer concentrations

and accelerate the crystallization process. Second, to the best of

our knowledge, TOP was seldom reported as a selective adhesion

surfactant for the anisotropic growth of NCs, which is quite

different from oleic acid. Therefore only spherical nanoparticles

were obtained under different TOP/1-octadecene ratios in this

study (Fig. S3a–c†), indicating the absence of anisotropic

growth. In this study, TOP shows a certain stabilization action

for NCs so that the nanoparticles synthesized are monodisperse

with diameters of 15–17 nm and no obvious size changes are

observed under different ratios of TOP/1-octadecene. The XRD

results (Fig. S3d†) indicate that the nanoparticles synthesized

under different TOP/1-octadecene ratios all exhibit hexagonal

phases.

Oleylamine is another widely used surfactant for NC

synthesis.28,29 Compared with oleic acid, oleylamine is a base, so

the results in oleylamine/1-octadecene are also quite different

from those in oleic acid/1-octadecene. The XRD patterns shown

in Fig. S4† demonstrated that both a-NaGdF4 and b-NaGdF4

were obtained in oleylamine/1-octadecene solvent. With

increasing the ratio of oleylamine/1-octadecene, a-NaGdF4

increased gradually and eventually only a-NaGdF4 was synthe-

sized in pure oleylamine. The same tendency could also be clearly

observed from the TEM images (Fig. S4†). According to the

previously reported literature for the synthesis of NaYF4,29 we

Fig. 5 FTIR spectrum of the NaGdF4:Ce(Yb)3+,Ln3+ nanorods

synthesized in 15 mL oleic acid/15 mL 1-octadecene at 280 #C for 5 h. The

FTIR spectra of pure oleic acid (OA) and sodium oleate were also

measured for comparison.


suppose that the energy barrier between a-NaGdF4 and

b-NaGdF4 may be significantly increased by the addition of

oleylamine, which makes a-NaGdF4 more stable. Therefore,

a-NaGdF4 NCs were hardly transformed into b-NaGdF4

nanorods or plates.

In summary, all these control experiments proved the crucial

role of oleic acid in this study and supported our monomer

concentration-controlled kinetic model for shape evolution of

the NCs.

3.5. Multicolor photoluminescence

Fig. 6a–d show the room-temperature excitation and emission

spectra of the NaGdF4:Ce3+, Ln3+ (Ln " Tb, Eu, Dy, Sm)

nanorods, which are dispersed as a 0.1 wt% (1 mg mL!1) colloid

in hexane. It can be seen that under a single irradiation of 254 nm

in the ultraviolet region, the NaGdF4 NCs doped with different

lanthanide ions show intensive multicolor visible emissions. Take

NaGdF4:Ce3+,Tb3+ as an example, electronic transitions within

4fn configurations of Tb3+ are strongly forbidden.9 But the

emission efficiency of Tb3+ can be greatly improved by exciting

a different ion (sensitizer, i.e., Ce3+ in this study) with an allowed

electronic transition which transfers the excitation energy to the

activator (i.e., Tb3+). The excitation spectra of all the samples are

all investigated and all show broad bands at around 250 nm

which is ascribed to the characteristic 4f–5d (or 2F5/2–5d) tran-

sition of Ce3+. During the fluorescence process, the excitation

energy is first absorbed by the 4f–5d transition of Ce3+ and

transferred to the Gd3+ and migrates over the Gd3+ lattice to the

Tb3+, where the energy is released as fluorescent emissions.12 For

example, the emission peaks of NaGdF4:Ce3+,Tb3+ are generated

from the 4f–4f transition of Tb3+ ions of 5D4–7F6 (488 nm),

5D4–7F5 (541 nm), 5D4–

7F4 (583 nm) and 5D4–7F3 (619 nm),

respectively. Moreover, by co-doping Yb-Er or Yb-Tm, the

NaGdF4 NCs can emit bright green or blue/violet up-conversion

fluorescence under a 980 nm NIR laser excitation, as shown in

Fig. 6e–f. Yb3+ acts as the sensitizer for the up-conversion

processes. Briefly, under the 980 nm excitation, an electron of

Yb3+ could be exited from the 2F7/2 to the 2F5/2 level. The energy

could be transferred to the activator ion (Er3+ or Tm3+) non-

radiatively to excite it to the corresponding excited level through

multi-photon processes and then the visible up-conversion

luminescence can be observed. Possible up-conversion mecha-

nisms for the Yb-Er and Yb-Tm co-doped nanomaterials have

been demonstrated in detail in our previous works.17,30,31

In this work, both the up-conversion and down-conversion

fluorescence are realized by using sensitizer–activator pairs.

Taking NaGdF4:Yb3+,Er3+ as an example, a relatively high Yb3+

doping level is required to obtain enough irradiation energy and

transfer it to Er3+ and to excite electrons of Er3+ to higher energy

levels. This energy transfer pathway requires a relatively higher

Yb3+ doping level than that of Er3+. In addition, if the doping

level of Er3+ is too high, cross-relaxation will happen and

concentration quenching will occur, which would decrease the

fluorescence intensity.31,32 The situations are similar for the

NaGdF4:Ce3+,Ln3+ NCs. That is why all observed activator

doping levels are generally very low and the sensitizer doping

levels are relatively high.

The multicolor up-/down-conversion luminescence of the

NaGdF4:Ce(Yb)3+,Ln3+ NCs shows their great promise in

applications in biolabels and optical technologies. More inter-

estingly, the as-synthesized NCs with regular shapes such as

hexagonal plates or nanorods tend to align in an orderly manner

to form superstructures via self-assembly, which was observed

Fig. 6 Room-temperature excitation (dashed line) and emission (solid line) spectra of 0.1wt% (a) NaGdF4:10%Ce3+,5%Tb3+; (b) NaGdF4:10%-

Ce3+,5%Eu3+; (c) NaGdF4:10%Ce3+,5%Dy3+; (d) NaGdF4:10%Ce3+,5%Sm3+ nanorods dispersed in hexane. All the emission spectra were recorded with

an excitation wavelength of 250 nm. (e–f) Room-temperature up-conversion fluorescence spectra of NaGdF4:18%Yb3+,2%Er3+ and NaGd-

F4:18%Yb3+,2%Tm3+ nanorods powders. (insets) Corresponding luminescence photographs of colloid NaGdF4:10%Ce3+,5%Ln3+ NCs excited at 254 nm

with a hand-held UV lamp and the NaGdF4:18%Yb3+,2%Er(Tm)3+ powder under irradiation with a 980 nm laser.


during the FESEM test (see Fig. S5 in the ESI†). It is well recog-

nized that the construction of highly ordered and densely packed

NCs arrays is very important for many technological applica-

tions.33,34 Combined with their multicolor photoluminescence,

the self-assembly properties of the highly luminescent NCs make

them especially promising for the construction of nanoarrays or

superlattices for applications in optical technologies.

4. Conclusions

To sum up, a one-pot synthesis approach was developed for

lanthanide-doped NaGdF4 NCs in this study. Multiple factors

including temperature, reaction time, and components of the

solvent all affected the formation and growth of the NaGdF4:-

Ce(Yb)3+,Ln3+ NCs and therefore NCs with diverse phase

structures and morphologies can be controllably synthesized by

controlling the appropriate experimental conditions.

It is evident that lower reaction temperatures and the presence

of oleylamine in the solvent tend to result in the formation of

NCs with cubic phase, while higher temperatures will generate

hexagonal phase NCs. On the other hand, in regard to the

morphology control, hexagonal-shaped nanoplates, nanorods,

and spherical particles of b-phase NaGdF4:Ce(Yb)3+,Ln3+ as well

as small a-phase particles can be controllably formed by varying

the conditions of the solvents and the reaction time under certain

temperature. Based on the experimental results, the crucial effect

of oleic acid was illustrated in detail and a monomer concen-

tration-controlled kinetic model for the formation of the NCs

was proposed and examined by a series of further control

experiments.

Furthermore, by co-doping different lanthanide ions, the

obtained NaGdF4 NCs can exhibit intensive multicolor down-/

up-conversion luminescence in the visible range under excitation

with irradiation of 254 nm or 980 nm, which is promising for

applications in multiple biolabels, staining and displays.

Acknowledgements

Financial support from National Natural Science Foundation of

China (NSFC NO. 30671925 and 20535020) and Key Project

of Science and Technology of Beijing Municipal Commission of

Education (KZ200810005004) is gratefully acknowledged.

References

1 F. Meiser, C. Cortez and F. Caruso, Angew. Chem., Int. Ed., 2004, 43,5954.

2 S. Sivakumar, P. R. Diamente and F. C. J. M. van Veggel,Chem.-Eur.J., 2006, 12, 5878.

3 G. S. Yi, H. C. Lu, S. Y. Zhao, Y. Ge, W. J. Yang, D. P. Chen andL. H. Guo, Nano Lett., 2004, 4, 2191.

4 S. Sivakumar, F. C. J. M. van Veggel andM. Raudsepp, J. Am. Chem.Soc., 2005, 127, 12464.

5 J. W. Stouwdam and F. C. J. M. van Veggel, Nano Lett., 2002, 2,733.

6 E. Downing, L. Hesselink, J. Ralston and R. Macfarlane, Science,1996, 273, 1185.

7 G. A. Kumar, C.W. Chen, J. Ballato and R. E. Riman,Chem.Mater.,2007, 19, 1523.

8 J. W. Stouwdam, G. A. Hebbink, J. Huskens and F. C. J. M. vanVeggel, Chem. Mater., 2003, 15, 4604.

9 Z. L. Wang, Z. W. Quan, P. Y. Jia, C. K. Lin, Y. Luo, Y. Chen,J. Fang, W. Zhou, C. J. O’Connor and J. Lin, Chem. Mater., 2006,18, 2030.

10 P. Ptacek, H. Schafer, K. Kompe and M. Haase, Adv. Funct. Mater.,2007, 17, 3843.

11 R. T. Wegh, H. Donker, K. D. Oskam and A. Meijerink, Science,1999, 283, 663.

12 F. Wang, X. Fan, M. Wang and Y. Zhang,Nanotechnology, 2007, 18,025701.

13 G. Blasse, Mater. Chem. Phys., 1987, 16, 201.14 M. Karbowiak, A. Mech, A. Bednarkiewicz, W. Strek and

L. Kepinski, J. Phys. Chem. Solids, 2005, 66, 1008.15 J. C. Boyer, J. Gagnon, L. A. Cuccia and J. A. Capobianco, Chem.

Mater., 2007, 19, 3358.16 J. Park, K. J. An, Y. S. Hwang, J. G. Park, H. J. Noh, J. Y. Kim,

J. H. Park, N. M. Hwang and T. Hyeon, Nat. Mater., 2004, 3,891.

17 Y. Wei, F. Q. Lu, X. R. Zhang and D. P. Chen, Chem. Mater., 2006,18, 5733.

18 L. Manna, E. C. Scher and A. P. Alivisatos, J. Am. Chem. Soc., 2000,122, 12700.

19 Z. A. Peng and X. G. Peng, J. Am. Chem. Soc., 2002, 124, 3343.20 L. Wang and Y. Li, Chem. Mater., 2007, 19, 727.21 X. G. Peng, Adv. Mater., 2003, 15, 459.22 H. X. Mai, Y. W. Zhang, L. D. Sun and C. H. Yan, Chem. Mater.,

2007, 19, 4514.23 R. Si, Y. W. Zhang, H. P. Zhou, L. D. Sun and C. H. Yan, Chem.

Mater., 2007, 19, 18.24 J. Wang, J. Hu, D. Tang, X. Liu and Z. Zhen, J. Mater. Chem., 2007,

17, 1597.25 J. W. Mullin, Crystallization, 3rd ed., Butterworth-Heinemann,

Oxford, UK, 1997.26 C. B. Murray, D. J. Norris and M. G. Bawendi, J. Am. Chem. Soc.,

1993, 115, 8706.27 X. G. Peng, L. Manna, W. D. Yang, J. Wickham, E. Scher,

A. Kadavanich and A. P. Alivisatos, Nature, 2000, 404, 59.28 G. S. Yi and G. M. Chow, Adv. Funct. Mater., 2006, 16, 2324.29 H. X. Mai, Y. W. Zhang, R. Si, Z. G. Yan, L. D. Sun, L. P. You and

C. H. Yan, J. Am. Chem. Soc., 2006, 128, 6426.30 C. H. Liu and D. P. Chen, J. Mater. Chem., 2007, 17, 3875.31 Y. Wei, F. Q. Lu, X. R. Zhang and D. P. Chen, J. Alloys Compd.,

2007, 427, 333.32 J. H. Zeng, J. Su, Z. H. Li, R. X. Yan and Y. D. Li,Adv. Mater., 2005,

17, 2119.33 E. V. Shevchenko, D. V. Talapin, N. A. Kotov, S. O’Brien and

C. B. Murray, Nature, 2006, 439, 55.34 E. V. Shevchenko, D. V. Talapin, C. B. Murray and S. O’Brien,

J. Am. Chem. Soc., 2006, 128, 3620.


LETTERS

nature materials | VOL 3 | DECEMBER 2004 | www.nature.com/naturematerials 891

Ultra-large-scale syntheses of monodisperse nanocrystalsJONGNAM PARK1, KWANGJIN AN1, YOSUN HWANG2, JE-GEUN PARK2, HAN-JIN NOH3, JAE-YOUNG KIM3, JAE-HOON PARK3, NONG-MOON HWANG4 AND TAEGHWAN HYEON1*1National Creative Research Center for Oxide Nanocrystalline Materials and School of Chemical Engineering, Seoul National University, Seoul 151-744, Korea2Department of Physics and Institute of Basic Science, Sungkyunkwan University, Suwon 440-746, Korea3Department of Physics and Pohang Light Source, Pohang University of Science and Technology, Pohang 790-784, Korea4School of Materials Science & Engineering and Nano-Systems Institute (NSI-NCRC), Seoul National University, Seoul 151-744, Korea*e-mail: [email protected]

Published online: 28 November 2004; doi:10.1038/nmat1251

The development of nanocrystals has been intensively pursued, not only for their fundamental scienti! c interest, but also for many technological applications1–3. " e synthesis of

monodisperse nanocrystals (size variation <5%) is of key importance, because the properties of these nanocrystals depend strongly on their dimensions. For example, the colour sharpness of semiconductor nanocrystal-based optical devices is strongly dependent on the uniformity of the nanocrystals3–6, and monodisperse magnetic nanocrystals are critical for the next-generation multi-terabit magnetic storage media7–9. For these monodisperse nanocrystals to be used, an economical mass-production method needs to be developed. Unfortunately, however, in most syntheses reported so far, only sub-gram quantities of monodisperse nanocrystals were produced. Uniform-sized nanocrystals of CdSe (refs 10,11) and Au (refs 12,13) have been produced using colloidal chemical synthetic procedures. In addition, monodisperse magnetic nanocrystals such as Fe (refs 14,15), Co (refs 16–18), #-Fe2O3 (refs 19,20), and Fe3O4 (refs 21,22) have been synthesized by using various synthetic methods23. Here, we report on the ultra-large-scale synthesis of monodisperse nanocrystals using inexpensive and non-toxic metal salts as reactants. We were able to synthesize as much as 40 g of monodisperse nanocrystals in a single reaction, without a size-sorting process. Moreover, the particle size could be controlled simply by varying the experimental conditions. " e current synthetic procedure is very general and nanocrystals of many transition metal oxides were successfully synthesized using a very similar procedure.

The process conditions required for the synthesis of monodisperse particles of micrometre size24 are relatively well established, and a similar principle could be applied to the synthesis of uniform-sized nanocrystals. The inhibition of additional nucleation during growth, in other words, the complete separation of nucleation and growth, is critical for the successful synthesis of monodisperse nanocrystals. Our research group developed new procedures for the synthesis of monodisperse nanocrystals of metals19,23,25, metal oxides19,26,27, and metal sulphides28 without a laborious size-sorting process. In particular,

during the direct synthesis of monodisperse iron and iron oxide nanocrystals19,23, we were able to ascertain that the iron–oleate complex, which is generated in situ from the reaction of iron pentacarbonyl and oleic acid, is decomposed and acts effectively as a growth source in synthesizing monodisperse nanocrystals with increased particle size. From these results, we reasoned that a metal–surfactant complex would make an effective growth source for the synthesis of monodisperse nanocrystals. The overall synthetic procedure is depicted in Fig. 1 and the detailed experimental procedures are described in the Methods section. Instead of using toxic and expensive organometallic compounds such as iron pentacarbonyl, we prepared the metal–oleate complex by reacting inexpensive and environmentally friendly compounds, namely metal chlorides and sodium oleate. The Fourier transform infrared spectrum of the iron–oleate complex, which was prepared by reacting iron chloride (FeCl3·6H2O) and sodium oleate, shows a C=O stretching peak at 1,700 cm–1, which is a characteristic peak for a metal–oleate complex (see Supplementary Information, Fig. S1). The iron–oleate complex in 1-octadecene was slowly heated

Metalchloride

Na–oleate Metal–oleatecomplex

Metal–oleatecomplex

+ + NaCl

Thermal decompositionin high boiling solvent

Monodispersenanocrystals

20 nm

Figure 1 The overall scheme for the ultra-large-scale synthesis of monodisperse nanocrystals. Metal–oleate precursors were prepared from the reaction of metal chlorides and sodium oleate. The thermal decomposition of the metal–oleate precursors in high boiling solvent produced monodisperse nanocrystals.

nmat1251-print.indd 891nmat1251-print.indd 891 10/11/04 11:43:36 am10/11/04 11:43:36 am

© 2004 Nature Publishing Group


LETTERS

892 nature materials | VOL 3 | DECEMBER 2004 | www.nature.com/naturematerials

to 320 °C, and was aged at that temperature for 30 min, generating iron oxide nanocrystals. The amount of the separated nanocrystals produced was as large as 40 g with a yield of >95%. The nanocrystals could easily be re-dispersed in various organic solvents including hexane and toluene.

To understand the mechanism of monodisperse nanoparticle formation, we obtained transmission electron microscope (TEM; JEOL EM-2010) images and conducted in situ infrared spectroscopy on the reaction mixture after heating at various temperatures and for various times. We also investigated the thermal decomposition behaviour of the solid-state iron–oleate precursor using thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and temperature-programmed infrared spectroscopy (see Supplementary Information). The TGA/DSC patterns and infrared spectra revealed that one oleate ligand dissociates from the precursor at 200–240 °C and the remaining two oleate ligands dissociate at ~300 °C by a CO2 elimination pathway. The TEM image of the sample taken at 310 °C without aging showed that nanoparticles were not produced, whereas the TEM image taken at 320 °C revealed the formation of relatively uniform nanoparticles with sizes ranging from 8 nm to 11 nm. All the TEM images taken after aging at 320 °C for 10, 20 and 30 min showed monodisperse 12 nm nanoparticles. Aging at 260 °C for one day produced polydisperse and poorly crystalline 9 nm nanoparticles, and aging at the same temperature for three days generated monodisperse 12 nm nanocrystals. When aged at 240 °C for one day, no nanoparticles were formed, and aging at 240 °C for three days produced highly polydisperse ~14 nm nanoparticles. When aged at 200 °C for three days, no nanoparticles were formed.

From these TEM, DSC/TGA and infrared data, we could propose the following mechanism for the nanocrystal formation. Nucleation occurs at 200–240 °C triggered by the dissociation of one oleate ligand from the Fe(oleate)3 precursor by CO2 elimination. The major growth occurs at ~300 °C initiated by the dissociation of the remaining two oleate ligands from the iron–oleate species, although

slow growth seems to occur at <250 °C. Consequently, we were able to synthesize monodisperse nanocrystals from the separation of nucleation and growth processes, which tend to take place at different temperatures. Because the growth process is time-dependent, when the precursor was aged at a low temperature of 240 °C (close to the nucleation temperature) for three days, we obtained polydisperse nanocrystals. These results imply that the current successful synthesis of monodisperse nanocrystals can be attributed to the effective separation of nucleation and growth processes, which result from the different temperature dependence of nucleation and growth kinetics.

0.2 mµ

Figure 2 12-nm magnetite nanocrystals. The TEM image clearly demonstrates that the nanocrystals are highly uniform in particle-size distribution. Inset is a photograph showing a Petri dish containing 40 g of the monodisperse magnetite nanocrystals, and a US one-cent coin for comparison.

a

b

c

d

e

f

j

i

h

g

20 nm

20 nm

20 nm

20 nm

20 nm 5 nm

5 nm

5 nm

5 nm

5 nm

Figure 3 TEM images (a–e) and HRTEM images (f–j) of monodisperse iron oxide nanocrystals. (a, f) 5 nm; (b, g) 9 nm; (c, h) 12 nm; (d, i) 16 nm; and (e, j) 22 nm nanocrystals. TEM images showed the highly monodisperse particle size distributions and HRTEM images revealed the highly crystalline nature of the nanocrystals.




LETTERS


The nanocrystals were characterized using TEM, X-ray diffraction (XRD; Rigaku D/Max-3C)), X-ray absorption spectroscopy (XAS), and X-ray magnetic circular dichroism spectroscopy (XMCD). The TEM image of the iron oxide nanocrystals synthesized at the 40-g scale in a single reaction using 200 g of solvent, shown in Fig. 2,

exhibited an extensive 2D assembly of uniform 12-nm nanocrystals, demonstrating their monodisperse particle size distribution (size variation = 2.3%, see Supplementary Information). A photograph of a Petri dish containing 40 g of the nanocrystals is shown in the inset of Fig. 2. The particle size of the iron oxide nanocrystals could

b

16 nm

14 nm

12 nm

9 nm

5 nm

Fe3O4

22 nm

–Fe2O3!

710 720 730Photon energy (eV)

Inte

nsity

(Arb

.uni

ts)

A

B

C 710 720 730

5 nm

22 nm

XMCD

!+

!–

!+ !–"! = –MCD

a

Fe3O4

22 nm

16 nm

14 nm

12 nm

9 nm

5 nm

–Fe2O3!L3 L2

L2

710 720 730Photon energy (eV)

Inte

nsity

(Arb

.uni

ts)

Fe2+

720 725

XAS

c

1.2

1.0

0.8

0.6

0.4

0.2

0.0

M/M

(TB)

–50 0 50 100 150 200 250 300 350 400

Temperature (K)

5 nm9 nm

12 nm16 nm22 nm

d

0 10 20 30

Diameter (nm)

25

20

15

10

5

0

K (1

05 erg

cm

–3)

K

TB

300

200

100

0

TB (K)

Figure 4 Characterization of monodisperse iron oxide nanocrystals. a, Fe L2,3-edge XAS and b, XMCD spectra of iron oxide nanocrystals in comparison with those of reference bulk materials, !-Fe2O3 and Fe3O4. The magnifi ed L2 region XAS spectra and the XMCD spectra of the 5 nm and 22 nm nanocrystals are shown in the insets of a and b, respectively. In the inset of b, !+ and !– represent the absorption coeffi cients for the photon helicity vector parallel and antiparallel to the magnetization direction of the nanocrystals, respectively. c, Temperature dependence of magnetization measured after zero-fi eld cooling (ZFC) using 100 Oe. For the sake of presentation, we have normalized the magnetization data with respect to the value at the maximum of ZFC magnetization, M(TB), for individual samples. d, Size dependence of TB, obtained from M(T) shown in c.




LETTERS

894 nature materials | VOL 3 | DECEMBER 2004 | www.nature.com/naturematerials

be controlled by using various solvents with different boiling points. As shown in Fig. 3, 5 nm (a, TEM; and f, high-resolution TEM), 9 nm (b and g), 12 nm (c and h), 16 nm (d and i), and 22 nm (e and j) iron oxide nanocrystals were synthesized using 1-hexadecene (b.p. 274 °C), octyl ether (b.p. 287 °C), 1-octadecene (b.p. 317 °C), 1-eicosene (b.p. 330 °C) and trioctylamine (b.p. 365 °C), respectively. All the nanocrystals are highly monodisperse in particle-size distribution (size variation <4.1%, see Supplementary Information). As the boiling point of the solvent increased, the diameter of the iron oxide nanocrystals increased. This result can be explained by the higher reactivity of the iron–oleate complex in the solvent with a higher boiling point. High-resolution TEM (HRTEM) images of these iron oxide nanocrystals showed distinct lattice fringe patterns, indicating the highly crystalline nature of the nanocrystals (Fig. 3f–j). The size of the iron oxide nanocrystals can be further fi ne-tuned by varying the concentration of oleic acid. For example, 11 nm, 12 nm and 14 nm iron oxide nanocrystals were synthesized using solutions with oleic acid concentrations of 1.5 mM, 3 mM and 4.5 mM, respectively (see Supplementary Information).

The XRD pattern of the 12-nm iron oxide nanocrystals revealed a cubic spinel structure of magnetite (see Supplementary Information). For the quantitative identifi cation of the compositions of the iron oxide nanocrystals, XAS and XMCD measurements at the Fe L2,3-edges were carried out at the EPU6 beamline at the Pohang Light Source. Figure 4a,b shows XAS spectra and XMCD results of the iron oxide nanocrystals with diameters of 5 nm, 9 nm, 12 nm, 16 nm and 22 nm, in comparison with those of two reference materials, bulk !-Fe2O3 (maghemite) and bulk Fe3O4 (magnetite), which have nearly the same spinel crystal structure with only ~1% difference in the cubic lattice constant. Both the XAS and MCD spectra of the 5 nm nanoparticles are very similar to those of !-Fe2O3, which contains only Fe3+. From the XAS and XMCD results, we made a quantitative estimation of the compositions for the iron oxide nanocrystals in the form of (!-Fe2O3)1–x(Fe3O4)x. The estimations are x = 0.20, 0.57, 0.68, 0.86 and 1.00 for the 5, 9, 12, 16 and 22 nm nanocrystals, respectively. Therefore, !-Fe2O3 is the dominant phase of the small 5-nm iron oxide nanocrystals, whereas the proportion of the Fe3O4 component gradually increases on increasing the particle size.

The magnetic properties of these iron oxide nanocrystals were studied using a commercial superconducting quantum interference device magnetometer (Quantum Design, MPMS5XL). Figure 4c shows the temperature dependence of magnetization measured with an applied magnetic fi eld of 100 Oe from 380 K to 5 K. All of our nanocrystals show superparamagnetic behaviour at high temperatures. However, on cooling, the zero-fi eld-cooled magnetization begins to drop and deviate from the fi eld-cooled magnetization at blocking temperature, TB. TB is at 40 K for the 5 nm sample, TB increases continuously as the diameter of the nanocrystals increases: for example, to 260 K for the 22 nm nanocrystals (Fig. 4d). From the measured TB, we calculated the magnetic anisotropy constant, K, using the equation: K = 25kBTB/V, where kB is Boltzmann’s constant and V is the volume of a single nanocrystal. As is typical of nanocrystals, the calculated magnetic anisotropy constant was found to increase with decreasing particle size (Fig. 4d).

The current synthetic procedure turned out to be widely applicable, and we successfully synthesized nanocrystals of many

a

b

c

20 mµ

20 mµ

(511)(422)(420)(331)(400)

(222)(311)(220)(200)(111)

<111><110>

d 2.56

d 2.02

50 nm 5 nm

100 nm

20 mµ

(112)(103)(110)(102)(101)(002)(100)

100 nm

<100>

d 2.78

5 nm5 nm

<002>d 2.6

10 n

m16

nm

8 nm

(001){011}

{010}

(001)–

<200>d 1.43

5 nm

Figure 5 TEM images, HRTEM images and electron diffraction patterns of monodisperse nanocrystals. a, MnO, b, CoO and c, Fe. TEM images show the highly uniform characteristics of the nanocrystals in terms of both particle size and particle shape. HRTEM images and electron diffraction patterns revealed the highly crystalline nature of the nanocrystals.




LETTERS


transition metal oxides from the thermolysis of metal–oleate complexes. For example, when the Mn–oleate complex was refl uxed in a solution containing octyl ether and oleic acid, uniform 12-nm f.c.c. MnO nanocrystals were synthesized (Fig. 5a). For the Co–oleate complex, short pencil-shaped CoO nanorods (Fig. 5b) were obtained. The CoO nanorods are uniform in diameter and form self-assembled superlattices (Fig. 5b). The XRD pattern of the CoO nanorods revealed an interesting Würtzite structure, similar to that of ZnO (see Supplementary Information). Furthermore, the (002) peak is narrower than the other peaks, demonstrating that the nanocrystals grow preferentially along the c axis. These results were confi rmed by the subsequent HRTEM, which shows the (002) lattice spacing value of 2.6 Å (Fig. 5b). When the iron–oleate complex was heated at a higher temperature of 380 °C, novel cube-shaped 20-nm iron nanocrystals were produced. XRD and HRTEM analyses revealed that the surface of these Fe nanocubes is passivated by a thin FeO layer (Fig. 5c). Nanocrystals of manganese ferrite and cobalt ferrite were synthesized from the thermal decomposition of the reaction mixtures composed of 1:2 molar ratio of the corresponding metal–oleate complex and iron–oleate complex (see Supplementary Information).

The synthetic procedures developed in the present study offer several very important advantageous features over the conventional methods for the synthesis of monodisperse nanocrystals. First, this process allows monodisperse nanocrystals to be obtained on an ultra-large scale of 40 g in a single reaction and without a further size-sorting process. When the reactors are set up in parallel, multi-kilograms of monodisperse nanocrystals can be readily obtained. Second, the synthetic process is environmentally friendly and economical, because it uses non-toxic and inexpensive reagents such as metal chlorides29. Third, the synthetic method is a generalized process that can be used to synthesize different kinds of monodisperse nanocrystals.

METHODS

SYNTHESIS OF IRON–OLEATE COMPLEXThe metal–oleate complex was prepared by reacting metal chlorides and sodium oleate. In a typical synthesis of iron–oleate complex, 10.8 g of iron chloride (FeCl3·6H2O, 40 mmol, Aldrich, 98%) and 36.5 g of sodium oleate (120 mmol, TCI, 95%) was dissolved in a mixture solvent composed of 80 ml ethanol, 60 ml distilled water and 140 ml hexane. The resulting solution was heated to 70 °C and kept at that temperature for four hours. When the reaction was completed, the upper organic layer containing the iron–oleate complex was washed three times with 30 ml distilled water in a separatory funnel. After washing, hexane was evaporated off, resulting in iron–oleate complex in a waxy solid form.

SYNTHESIS OF IRON OXIDE NANOCRYSTALSThe following is a typical synthetic procedure for monodisperse iron oxide (magnetite) nanocrystals with a particle size of 12 nm. 36 g (40 mmol) of the iron-oleate complex synthesized as described above and 5.7 g of oleic acid (20 mmol, Aldrich, 90%) were dissolved in 200 g of 1-octadecene (Aldrich, 90%) at room temperature. The reaction mixture was heated to 320 °C with a constant heating rate of 3.3 °C min–1, and then kept at that temperature for 30 min. When the reaction temperature reached 320 °C, a severe reaction occurred and the initial transparent solution became turbid and brownish black. The resulting solution containing the nanocrystals was then cooled to room temperature, and 500 ml of ethanol was added to the solution to precipitate the nanocrystals. The nanocrystals were separated by centrifugation.

Received 11 June 2004; accepted 20 September 2004; published 28 November 2004.

References1. Schmid, G. Nanoparticles: From Theory to Application (Wiley-VCH, Weinheim, 2004).2. Klabunde, K. J. Nanoscale Materials in Chemistry (Wiley-Interscience, New York, 2001).3. Alivisatos, A. P. Semiconductor clusters, nanocrystals, and quantum dots. Science 271, 933#937

(1996).4. Nirmal, M. & Brus, L. Luminescence photophysics in semiconductor nanocrystals. Acc. Chem. Res.

32, 407#414 (1999).5. Murray, C. B., Kagan, C. R. & Bawendi, M. G. Synthesis and characterization of monodisperse

nanocrystals and close-packed nanocrystal assemblies. Annu. Rev. Mater. Sci. 30, 545#610 (2000).6. Rogach, A. L. et al. Organization of matter on different size scales: monodisperse nanocrystals and

their superstructures. Adv. Funct. Mater. 12, 653#664 (2002).7. Sun, S., Murray, C. B. Weller, D., Folks, L. & Moser A. Monodisperse FePt nanoparticles and

ferromagnetic FePt nanocrystal superlattices. Science 287, 1989#1992 (2000).8. Speliotis, D. E. Magnetic recording beyond the fi rst 100 (invited). J. Magn. Magn. Mater. 193, 29#35

(1999).9. O’Handley, R. C. Modern Magnetic Materials (Wiley, New York, 1999).10. Murray, C. B., Norris, D. J. & Bawendi, M. G. Synthesis and characterization of nearly monodisperse

CdE (E = S, Se, Te) semiconductor nanocrystallites. J. Am. Chem. Soc. 115, 8706#8715 (1993).11. Peng, X., Wickham, J. & Alivisatos, A. P. Kinetics of II-VI and III-V colloidal semiconductor

nanocrystal growth: “focusing” of size distributions. J. Am. Chem. Soc. 120, 5343#5344 (1998).12. Stoeva, S., Klabunde, K. J., Sorensen, C. M. & Dragieva, I. Gram-scale synthesis of monodisperse

gold colloids by the solvated metal atom dispersion method and digestive ripening and their organization into two- and three-dimensional structures. J. Am. Chem. Soc. 124, 2305#2311 (2002).

13. Jana, N. R. & Peng, X. Single-phase and gram-scale routes toward nearly monodisperse Au and other noble metal nanocrystals. J. Am. Chem. Soc. 125, 14280#14281 (2003).

14. Park, S.-J. et al. Synthesis and magnetic studies of uniform iron nanorods and nanospheres. J. Am. Chem. Soc. 112, 8581#8582 (2000).

15. Dumestre, F., Chaudret, B. Amiens, C., Renaud, P. & Fejes P. Superlattices of iron nanocubes synthesized from Fe[N(SiMe3)2]2. Science 303, 821#823 (2004).

16. Sun, S. & Murray, C. B. Synthesis of monodisperse cobalt nanocrystals and their assembly into magnetic superlattices (invited). J. Appl. Phys. 85, 4325#4390 (1999).

17. Puntes, V. F., Krishnan, K. M. & Alivisatos, A. P. Colloidal nanocrystal shape and size control: the case of cobalt. Science 291, 2115#2117 (2001).

18. Dumestre, F. et al. Shape control of thermodynamically stable cobalt nanorods through organometallic chemistry. Angew. Chem. Int. Edn 41, 4286#4289 (2002).

19. Hyeon, T., Lee, S. S., Park, J., Chung, Y. & Na, H. B. Synthesis of highly crystalline and monodisperse maghemite nanocrystallites without a size-selection process. J. Am. Chem. Soc. 123, 12798#12801 (2001).

20. Rockenberger, J., Scher, E. C. & Alivisatos, A. P. A new nonhydrolytic single-precursor approach to surfactant-capped nanocrystals of transition metal oxides. J. Am. Chem. Soc. 121, 11595#11596 (1999).

21. Sun, S. et al. Monodisperse MFe2O4 (M = Fe, Co, Mn) nanoparticles. J. Am. Chem. Soc. 126, 273#279 (2004).

22. Pileni, M. P. The role of soft colloidal templates in controlling the size and shape of inorganic nanocrystals. Nature Mater. 2, 145#150 (2003).

23. Hyeon, T. Chemical synthesis of magnetic nanoparticles. Chem. Comm. 927#934 (2003).24. Sugimoto, T. Monodispersed Particles (Elsevier Science, Amsterdam, 2001).25. Kim, S. W. et al. Synthesis of monodisperse palladium nanoparticles. Nano Lett. 3, 1289#1291

(2003).26. Hyeon, T. et al. Synthesis of highly crystalline and monodisperse cobalt ferrite nanocrystals.

J. Phys. Chem. B 106, 6831#6833 (2002).27. Joo, J. et al. Multigram scale synthesis and characterization of monodisperse tetragonal zirconia

nanocrystals. J. Am. Chem. Soc. 125, 6553#6557 (2003).28. Joo, J. et al. Generalized and facile synthesis of semiconducting metal sulfi de nanocrystals. J. Am.

Chem. Soc. 125, 11100#11105 (2003).29. Peng, X. Green chemical approaches toward high-quality semiconductor nanocrystals. Chem. Eur. J.

8, 334#339 (2002).

AcknowledgementsT.H. would like to thank the fi nancial support from the Korean Ministry of Science and Technology through the National Creative Research Initiative Program. J.G.P. would like to thank the fi nancial support by the KOSEF through the Center for Strongly Correlated Materials Research at the Seoul National University. J.H.P. would like to thank the fi nancial support by KISTEP through X-ray/particle-beam Nanocharacterization Program.Correspondence and requests for materials should be addressed to T. H.

Competing fi nancial interestsThe authors declare that they have no competing fi nancial interests.




Subscriber access provided by NEW YORK UNIV

Chemistry of Materials is published by the American Chemical Society. 1155 SixteenthStreet N.W., Washington, DC 20036

ArticleSynthesis of Oil-Dispersible Hexagonal-Phase

and Hexagonal-Shaped NaYF4:Yb,Er NanoplatesYang Wei, Fengqi Lu, Xinrong Zhang, and Depu Chen

Chem. Mater., 2006, 18 (24), 5733-5737• DOI: 10.1021/cm0606171 • Publication Date (Web): 25 October 2006Downloaded from http://pubs.acs.org on May 11, 2009

Subscriber access provided by NEW YORK UNIV

Chemistry of Materials is published by the American Chemical Society. 1155 SixteenthStreet N.W., Washington, DC 20036

More About This Article

Additional resources and features associated with this article are available within the HTML version:

• Supporting Information• Links to the 15 articles that cite this article, as of the time of this article download• Access to high resolution figures• Links to articles and content related to this article• Copyright permission to reproduce figures and/or text from this article

Synthesis of Oil-Dispersible Hexagonal-Phase and Hexagonal-ShapedNaYF4:Yb,Er Nanoplates

Yang Wei, Fengqi Lu, Xinrong Zhang, and Depu Chen*Department of Chemistry, Tsinghua UniVersity, Beijing 100084, People’s Republic of China

ReceiVed March 14, 2006. ReVised Manuscript ReceiVed July 20, 2006

Oil-dispersible R-NaYF4 spherical nanoparticles and !-NaYF4 hexagonal-shaped nanoplates weresynthesized by the liquid-solid two-phase approach at different reaction temperatures. The TEM andFE-SEM images reveal that the nanoplates have a relatively narrow size distribution. In comparison withother methods, pure !-NaYF4 hexagonal-shaped nanoplates were prepared under a relatively mild condition.The nanoplates grew at the liquid-solid interface with slow crystallization rate, which may be preferablefor achieving !-NaYF4.

Introduction

Upconversion luminescent materials have received greatattention for a few decades.1 Because of their unique anti-Stokes optical property, upconversion luminescent materialshave a number of potential applications, including lasers,2three-dimensional displays,3 light emitting devices,4 biologi-cal detection,5,6 and many others.7 In recent years, severalresearch groups have reported the upconversion luminescencein nanomaterials.8-11 Besides the dried powdered nanocrys-talline upconversion phosphors (UCPs), there is a growinginterest to prepare UCPs that have a good dispersibility inorganic solvents.12-17 Spherical cubic-phase NaYF4 nano-particles that could be dispersed in nonpolar solvent havebeen prepared in homogeneous solution13 and by a newlydeveloped liquid-solid-solution (LSS) process.15 Veryrecently, the methods to prepare NaYF4:Yb,Er/Tm nano-

crystals based on the co-thermolysis of sodium trifluoroac-etate and rare earth trifluoroacetate were independentlyreported by Yan et al.16 and Capobianco et al.17 NaYF4 existsin two polymorphs at ambient pressure: cubic R-phase andhexagonal !-phase. !-NaYF4 has been reported as the mostefficient host material for green and blue UCPs.18 In ourprevious research,6 the cubic-to-hexagonal phase transitionprocess of NaYF4 is an exothermic process. Correspondingly,the hexagonal-to-cubic phase transition process of NaYF4has been reported as an endothermic process.18 From theresults provided by differential scanning calorimetric mea-surement, it is reasonable to conclude that for NaYF4, thehexagonal phase is more thermodynamically stable than thecubic phase. However, in most cases for preparation ofNaYF4, R-NaYF4 nanocrystals were obtained. The cubicphase could transfer into hexagonal phase under heattreatment. Annealing treatment18 and hydrothermal or sol-vothermal treatment19,20 were adopted to produce the !-NaYF4.!-NaYF4 could also be formed under drastic conditionsreported by Yan et al.16 In this paper, a liquid-solid two-phase approach was used to synthesize !-NaYF4 directly.The results revealed that not only !-NaYF4 hexagonal-shapednanoplates but also R-NaYF4 spherical nanoparticles couldbe synthesized by the liquid-solid two-phase approach atdifferent reaction temperatures. The possible mechanism wasdiscussed in this paper.

Experimental Section

Materials. Sodium fluoride (NaF), sodium hydroxide (NaOH),ethanol, and hydrochloric acid (HCl) were obtained from BeijingChemical Corp. (Beijing, China). Yttrium oxide (Y2O3, 99.99%),ytterbium oxide (Yb2O3, 99.99%), and erbium oxide (Er2O3,99.99%) were obtained from Grirem Advanced Materials Co., Ltd.(Beijing, China), and were of SpecPure grade. Oleic acid was

* To whom correspondence should be addressed. Tel.: +86 10 62781691.Fax: +86 10 62782485. E-mail: [email protected].(1) Auzel, F. Chem. ReV. 2004, 104, 139.(2) Scheps, R. Prog. Quantum Electron. 1996, 20, 271.(3) Downing, E.; Hesselink, L.; Ralston, J.; Macfarlane, R. Science 1996,

273, 1185.(4) Sivakumar, S.; van Veggel, F. C. J. M.; Raudsepp, M. J. Am. Chem.

Soc. 2005, 127, 12464.(5) van de Rijke, F.; Zijlmans, H.; Li, S.; Vail, T.; Raap, A. K.; Niedbala,

R. S.; Tanke, H. J. Nat. Biotechnol. 2001, 19, 273.(6) Yi, G. S.; Lu, H. C.; Zhao, S. Y.; Yue, G.; Yang, W. J.; Chen, D. P.;

Guo, L. H. Nano Lett. 2004, 4, 2191.(7) Shalav, A.; Richards, B. S.; Trupke, T.; Kramer, K. W.; Gudel, H. U.

Appl. Phys. Lett. 2005, 86, 013505.(8) Yi, G. S.; Sun, B. Q.; Yang, F. Z.; Chen, D. P.; Zhou, Y. X.; Cheng,

J. Chem. Mater. 2002, 14, 2910.(9) Matsuura, D. Appl. Phys. Lett. 2002, 81, 4526.(10) Patra, A.; Friend, C. S.; Kapoor, R.; Prasad, P. N. J. Phys. Chem. B

2002, 106, 1909.(11) Vetrone, F.; Boyer, J. C.; Capobianco, J. A.; Speghini, A.; Bettinelli,

M. J. Phys. Chem. B 2003, 107, 1107.(12) Heer, S.; Lehmann, O.; Haase, M.; Gudel, H. U. Angew. Chem., Int.

Ed. 2003, 42, 3179.(13) Heer, S.; Kompe, K.; Gudel, H. U.; Haase, M. AdV. Mater. 2004, 16,

2102.(14) Yi, G. S.; Chow, G. M. J. Mater. Chem. 2005, 15, 4460.(15) Wang, X.; Zhuang, J.; Peng, Q.; Li, Y. D. Nature 2005, 437, 121.(16) Mai, H. X.; Zhang, Y. W.; Si, R.; Yan, Z. G.; Sun, L. D.; You, L. P.;

Yan, C. H. J. Am. Chem. Soc. 2006, 128, 6426.(17) Boyer, J. C.; Vetrone, F.; Cuccia, L. A.; Capobianco, J. A. J. Am.

Chem. Soc. 2006, 128, 7444.

(18) Kramer, K. W.; Biner, D.; Frei, G.; Gudel, H. U.; Hehlen, M. P.;Luthi, S. R. Chem. Mater. 2004, 16, 1244.

(19) Wang, L. Y.; Yan, R. X.; Hao, Z. Y.; Wang, L.; Zeng, J. H.; Bao, H.;Wang, X; Peng, Q.; Li, Y. D. Angew. Chem., Int. Ed. 2005, 44, 6054.

(20) Zeng, J. H.; Su, J.; Li, Z. H.; Yan, R. X.; Li, Y. D. AdV. Mater. 2005,17, 2119.

5733Chem. Mater. 2006, 18, 5733-5737

10.1021/cm0606171 CCC: $33.50 © 2006 American Chemical SocietyPublished on Web 10/25/2006

purchased from Beijing Chemical Reagents Co. (Beijing, China).1-Octadecene was purchased from Acros Organics (NJ). Rare earthchlorides (RECl3‚6H2O with RE ) Y, Yb, Er) were prepared bydissolving the corresponding rare earth oxides in hydrochloric acidat elevated temperature, and then evaporating the solvent in avacuum. Sodium oleate was prepared by reacting oleic acid andsodium hydroxide using ethanol as reaction medium. Sodiumhydroxide was dispersed in ethanol under vigorous stirring, andthen equal molar oleic acid was added dropwise. After theneutralization reaction was completed, ethanol and water wereevaporated in a vacuum.Synthesis of Rare Earth Oleate Complexes.A literature method

for the synthesis of iron-oleate complex21 was adopted to preparethe rare earth oleate complexes. In a typical synthesis of yttrium-oleate complex, 20 mmol of yttrium chloride (YCl3‚6H2O) and 60mmol of sodium oleate were dissolved in a mixture solventcomposed of 40 mL of ethanol, 30 mL of distilled water, and 70mL of hexane. The resulting solution was added into a 250-mLround-bottomed flask with a reflux condenser, and then heated to70 °C and kept at that temperature for 4 h. After the reaction wascompleted, the reaction mixture was transferred into a separatoryfunnel. The upper organic layer was separated and washed threetimes with 30 mL of distilled water. After being washed, yttrium-oleate complex was produced in a waxy solid form by evaporatingoff the remaining hexane. Ytterbium-oleate complex and erbium-oleate complex were synthesized in the same way.Synthesis of NaYF4:Yb,Er Nanocrystals. The reaction tem-

perature is the only difference between the synthetic proceduresfor preparation of R-NaYF4:Yb,Er spherical nanoparticles and!-NaYF4:Yb,Er hexagonal-shaped nanoplates. For the synthesis of!-NaYF4:Yb,Er hexagonal-shaped nanoplates, 0.2 g of NaF solidpowder and 30 mL of 1-octadecene were added into a 100 mLthree-necked flask, and NaF was dispersed in the 1-octadecene withvigorous magnetic stirring. The mixture was degassed under vacuumfor about 30 min, and flushed periodically with N2. The temperatureof the reaction flask was then stabilized at 260 °C under N2atmosphere. Next, 0.8 mmol of yttrium-oleate complex, 0.17 mmolof ytterbium-oleate complex, and 0.03 mmol of erbium-oleatecomplex were dissolved in 30 mL of 1-octadecene to form anoptically transparent solution. This solution was bubbled with N2gas for 10 min, and then was quickly delivered to the vigorouslystirring reaction flask in a single injection with a 50-mL syringethrough a rubber septum. The reaction was kept at 260 °C for 6 hin N2 atmosphere under vigorous stirring. As the reaction mixturewas cooled to !60 °C, it was washed three times with 30 mL of!60 °C hot deionized water in a separatory funnel. Next, 100 mLof ethanol was added to the resulting solution containing thenanocrystals. The nanocrystals were separated by centrifugation.The as-prepared nanocrystals could be easily redispersed in variousnonpolar organic solvents, such as hexane and toluene. When thereaction temperature was lowered to 210 °C, R-NaYF4:Yb,Erspherical nanoparticles were obtained.Characterization. Investigations on the size and morphology

of the nanocrystals were performed using a JEM-1200EX transmis-sion electron microscope (TEM) (JEOL, Japan) operating ataccelerating voltages up to 100 kV and a JSM-7401F field emissionscanning electron microscope (FE-SEM) (JEOL, Japan) operatingat accelerating voltages up to 1 kV. Powder X-ray diffractionpatterns were obtained on a D/max-RB X-ray diffractometer(Rigaku, Japan) and a D/max-2500 X-ray diffractometer (Rigaku,Japan). Upconversion fluorescent spectra were measured on a LS-55 fluorescence spectrophotometer (Perkin-Elmer Corp.) with an

external 980 nm laser (Beijing Hi-Tech Optoelectronic Co., China)instead of internal excitation source. A 62.5/125 (core/claddingdimensions, which are given in micrometers) multimode opticalfiber with the numerical aperture 0.22 was used to conduct the laserinto the spectrophotometer. The distance between the fiber headand the samples is about 3 mm.

Results and Discussion

Synthesis of NaYF4:Yb, Er Nanocrystals.A liquid-solidtwo-phase approach was used to synthesize the NaYF4:Yb,-Er nanocrystals. Because of its high boiling point, 1-octa-decene was chosen as the solvent for the high-temperaturegrowth and annealing of NaYF4 crystallites. Rare earth oleatecomplexes were dissolved in the organic solvent as the liquidphase. Because of the insolubility of NaF in 1-octadecene,NaF was dispersed in the same organic solvent as the solidphase. NaYF4:Yb,Er nanocrystals grew at the liquid-solidinterface, and oleic acid released from the rare earth oleatecomplexes could bind to the growing nanocrystals with thealkyl chains outward, through which the nanocrystals gainedhydrophobic surfaces.TEM was used to map the shape and size of the

nanocrystals dispersed on a carbon-coated copper grid fromhexane solutions, and FE-SEM was also used to give thethree-dimensional morphological observation of the nano-crystals. The TEM images of the nanocrystals prepared at210, 230, and 260 °C are shown in Figure 1A,B and Figure2, respectively. The TEM images with higher magnificationfor these nanocrystals are shown in Figures S1-S3. Asshown in Figure 1A and Figure S1, the nanoparticles obtainedat 210 °C have roughly spherical shapes with the averagesize of about 7 nm. Increasing the reaction temperature to230 °C, much bigger nanocrystals were formed. It could befound from the TEM images shown in Figure 1B and FigureS2 that the shapes of these nanocrystals are between roundand hexagonal. In fact, from the FE-SEM image shown inFigure S5, it is clear that the nanocrystals obtained at 230°C were mixtures composed of hexagonal-shaped nanoplatesand small nanoparticles glued to the nanoplates. After thereaction temperature reached 260 °C, pure hexagonal-shapednanoplates were obtained. The TEM images of these nano-plates were shown in Figure 2 and Figure S3, and the FE-SEM images of them were shown in Figure S6. The TEMimages shown in Figure 2 reveal that most of the nanoplatesare lying flat on the face, and a small quantity of thenanoplates are standing on the edge. The edge lengths ofthe nanoplates shown in the TEM images are not equal,which is probably due to the nanocrystals being tilted bydifferent angles with respect to the carbon films. Thenanoplates are characterized by !35 nm in edge length and!20 nm in thickness. The TEM and FE-SEM images revealthat the nanoplates have a relatively narrow size distribution.In the liquid-solid two-phase approach, both nucleation andgrowth of nanoplates could only occur at the interface ofthe two phases. After the nuclei formed at the interface, theycould enter the liquid phase, and thus the nanocrystalsstopped growing. Only if the nanocrystals returned to theinterface could they continue to grow. The bigger sizenanocrystals had, the more slowly they moved. Smaller

(21) Park, J.; An, K. J.; Hwang, Y. S.; Park, J. G.; Noh, H. J.; Kim, J. Y.;Park, J. H.; Hwang, N. M.; Hyeon, T. Nat. Mater. 2004, 3, 891.

5734 Chem. Mater., Vol. 18, No. 24, 2006 Wei et al.

nanocrystals could move into the liquid phase and went backto the interface more quickly than bigger nanocrystals, andso the smaller nanocrystals had more chance to grow thanthe bigger nanocrystals; thus the narrow particle sizedistribution was obtained.The efficiency of the reaction that occurred at the liquid-

solid interface is much lower than that in homogeneoussolution. Several factors, such as the reaction time, stirringspeed, reaction temperature, have strong impacts on theoutput of this reaction. Vigorous stirring and enough highreaction temperature were required. To avoid the thermolysisof rare earth oleate complexes at high temperature, thereaction temperature for preparation of hexagonal-shapednanoplates was chosen at 260 °C. The output of the reactionincreased greatly with increasing the reaction time. From theTEM images shown in Figure S4, there were no obviousdifference found in the size and morphology of the nanoplatesobtained from different reaction time. This result could begood evidence for the reaction mechanism provided above.

The XRD results reveal that the spherical NaYF4 nano-particles synthesized at 210 °C crystallized in the cubicR-phase and the hexagonal-shaped NaYF4 nanoplates syn-thesized at 260 °C crystallized in the hexagonal !-phase.The crystallographic phase of the initial seed during thenucleation processes is critical for directing the nanocrystalshapes due to its characteristic unit cell structure.22 The cubicR-NaYF4 seeds have isotropic unit cell structures, whichgenerally induce the isotropic growth of nanocrystals, andtherefore spherical R-NaYF4 nanoparticles were observed.In contrast, hexagonal !-NaYF4 seeds have anisotropic unitcell structures, which can induce anisotropic growth alongcrystallographically reactive directions, and thus hexagonal-shaped nanoplates were obtained.Structural Characterization of NaYF4:Yb,Er Nano-

crystals. X-ray powder diffraction patterns of NaYF4:Yb,-

(22) Jun, Y. W.; Lee, J. H.; Choi, J. S.; Cheon, J. J. Phys. Chem. B 2005,109, 14795.

Figure 1. TEM images of NaYF4:Yb,Er nanocrystals obtained at 210 °C (A) and 230 °C (B).

Figure 2. TEM images of !-NaYF4:Yb,Er nanoplates obtained at 260 °C.

Synthesis of Oil-Dispersible NaYF4:Yb,Er Nanoplates Chem. Mater., Vol. 18, No. 24, 2006 5735

Er nanocrystals obtained at 210, 230, and 260 °C are shownin Figure 3 and Figure 4. Deduced from the XRD data, allof the samples were well-crystallized. Except two peaks thatwere marked with asterisks, all of the other peaks of thenanoparticles prepared at 210 °C and of the nanoplatesprepared at 260 °C could be readily indexed to the cubicR-NaYF4 phase or hexagonal !-NaYF4 phase, respectively.For the sample synthesized at 230 °C, the !-NaYF4 phasewas the major species; however, a small quantity ofR-NaYF4phase could be detected. The selected-area electron diffrac-tion (SAED) patterns shown in Figure S7, which were takenfrom a single nanoplate obtained at 260 °C, demonstratedthe single-crystalline nature of the sample; it could be readilyindexed as hexagonal phase, in good agreement with its XRDdata.The average particle size estimated by line-broadening was

6.4 nm for the R-NaYF4 spherical nanoparticles. For!-NaYF4 hexagonal-shaped nanoplates, the particle sizesestimated by line-broadening from different diffraction peaksvaried from 26.2 to 63.2 nm.

These two asterisked peaks in the XRD patterns could bewell indexed to the reactant NaF, which means the nano-crystals obtained were mixed with some residual NaF. Duringthe reaction process, the polar head of oleic acid could absorbto the NaF particles with the nonpolar tail on the outside,through which the NaF particles gained hydrophobic sur-faces. After the reaction was completed, to remove theresidual NaF, the reaction mixture was washed three timeswith deionized water in a separatory funnel; however, itseemed that this procedure is not effective enough to removeall of the residual NaF. NaF could be removed completelyfrom the products; the detailed procedure is described in theSupporting Information.For bulk NaYF4, R-phase is the metastable high-temper-

ature phase, while !-phase is the thermodynamically stablelow-temperature phase.16 At temperatures above approxi-mately 700 °C, the !-phase is unstable.18 However, in mostcases for preparation of nanosized NaYF4, it tends tocrystallize in the cubic phase, the less thermodynamicallystable phase, first. In their newly reported work, Yan et al.demonstrated a general one-step synthesis of NaREF4 (RE) Pr to Lu, Y) nanocrystals via the co-thermolysis of Na-(CF3COO) and RE(CF3COO)3 precursors.16 Pure R-NaYF4could be obtained at a low temperature (280 °C) and a lowratio of Na/RE with a relatively short reaction time, while!-NaYF4 was formed only under drastic conditions (highNa/RE, 330 °C, and long reaction time).It is interesting that both pure R-NaYF4 nanocrystals and

!-NaYF4 nanoplates could be obtained in this liquid-solidtwo-phase approach, and the only difference between thesynthetic procedures is the reaction temperature. In com-parison with the method to prepare the R-NaYF4 sphericalnanoparticles in homogeneous solution,13 the liquid-solid-solution (LSS) process,15 and the thermal decompositionmethod,16,17 the reaction rate and crystallization rate are muchslower in the liquid-solid two-phase approach, for bothnucleation and growth of nanocrystals could only occur atthe liquid-solid interface. To overcome the energy barrierfor the formation of !-NaYF4, appropriate high temperaturewas needed to prepare !-NaYF4. It is why only R-NaYF4nanoparticles were obtained at 210 °C and mixtures com-posed of !-NaYF4 hexagonal-shaped nanoplates and R-NaYF4spherical nanoparticles were obtained at 230 °C in thisapproach. Pure !-NaYF4 hexagonal-shaped nanoplates couldbe obtained at 260 °C, a relatively mild condition, in thisliquid-solid two-phase approach. In contrast, drastic condi-tions were needed for synthesis of !-NaYF4 via the thermaldecomposition method. In addition, in a glycerol-mediatedsynthesis of NaYF4:Yb,Er nanoparticles in homogeneoussolution at 260 °C, pure R-NaYF4 was obtained even after 6h of reaction. However, pure hexagonal phase was detectedin the products obtained after 2 h of reaction via the liquid-solid two-phase approach. (The details are described in theSupporting Information.)In the structure of R-NaYF4, the cation sites are occupied

randomly by Na+ and Y3+ cations, while in !-NaYF4 thecation sites are of three types: a 1-fold site occupied by Y3+,a 1-fold site occupied randomly by 1/2Na+ and 1/2Y3+, and a2-fold site occupied randomly by Na+ and vacancies. Thus,

Figure 3. Calculated line pattern (A) and experimental powder XRD data(B) for the R-NaYF4:Yb,Er nanoparticles obtained at 210 °C.

Figure 4. Calculated line pattern (A) and experimental powder XRD datafor the !-NaYF4:Yb,Er nanoplates obtained at 230 °C (B) and 260 °C (C).

5736 Chem. Mater., Vol. 18, No. 24, 2006 Wei et al.

for NaYF4, the cubic-to-hexagonal phase transformation isof a disorder-to-order character with respect to cations.16From the results described above, we presume that a slowcrystallization process may be preferable for achieving!-NaYF4, the order phase. A similar phenomenon was foundin the recently reported synthesis of lanthanide fluoridenanoparticles by a reverse microemulsion method.23 Sphericalamorphous YF3 particles were obtained by the classicalmicroemulsion method (mixing of two microemulsionscontaining fluoride and YCl3, respectively); conversely,single-crystal particles with regular hexagonal and triangularshape were obtained by the single microemulsion method(direct addition of a fluoride solution to a microemulsioncontaining YCl3). The particle growth and crystallizationwere slower in the single microemulsion method than in theclassical microemulsion method, and, as a result, single-crystal nanoparticles rather than the amorphous particles wereobtained in the single microemulsion.Upconversion Fluorescent Properties of !-NaYF4:Yb,-

Er Nanoplates. Room-temperature upconversion fluores-cence spectra of NaYF4:Yb,Er nanocrystals obtained at 210,230, and 260 °C in the wavelength region of 400-700 nmare shown in Figure 5. As compared to the fluorescentintensity of R-NaYF4 nanoparticles prepared at 210 °C, thatof !-NaYF4 nanoplates prepared at 260 °C is enhancedgreatly. This result is in good agreement with the reportedresult that !-NaYF4 is the most efficient host material forgreen and blue upconversion phosphors known today.18-20It should be pointed out that the enhancement of thefluorescent intensity may be partly due to the increase of

the particle size. There are three major bands in the curve,centered at 522, 542, and 654 nm, respectively. Themechanisms responsible for the upconversion fluorescenceare shown in the Supporting Information.The 542 nm-to-654 nm emission ratio of !-NaYF4

nanoplates prepared at 260 °C is about 1.29 to 1. Severalfactors such as doping levels, excitation power, and impuri-ties have impacts on the green-to-red emission ratio. In thispaper, the doping levels for Yb3+ and Er3+ were notoptimized. The actual molar ratio of the rare earth metals(Y:Yb:Er) in the !-NaYF4 nanoplates obtained at 260 °Cexamined by ICP-OES was 0.816:0.156:0.028. This resultis in good agreement with the planned molar ratios for Y,Yb, and Er (0.80:0.17:0.03). Furthermore, the organiccapping groups binding on the surface of the !-NaYF4nanoplates may increase the multiphonon relaxation ratesbetween the metastable states, which reduce the green-to-red emission ratio.

Acknowledgment. Financial support from National NatureScience Foundation of China (20535020) is gratefully acknowl-edged.

Supporting Information Available: TEM and FE-SEM imagesof NaYF4:Yb,Er nanocrystals, SAED and XRD patterns of !-NaYF4:Yb,Er nanoplates obtained at 260 °C, procedures for removingresidual NaF from products, brief description for the glycerol-mediated synthesis of R-NaYF4:Yb,Er nanoparticles, and mecha-nisms responsible for the upconversion fluorescence of !-NaYF4:Yb,Er nanoplates obtained at 260 °C (PDF). This material isavailable free of charge via the Internet at http://pubs.acs.org.

CM0606171(23) Lemyre, J. L.; Ritcey, A. M. Chem. Mater. 2005, 17, 3040.

Figure 5. Upconversion emission spectrum of NaYF4:Yb,Er nanocrystals obtained at 210, 230, and 260 °C (the excitation conditions were the same forthese samples: the laser power is 437 mW, and the emission slit is 3 nm).

Synthesis of Oil-Dispersible NaYF4:Yb,Er Nanoplates Chem. Mater., Vol. 18, No. 24, 2006 5737

Morphology- and phase-controlled synthesis of...

Documents

Transcript of Morphology- and phase-controlled synthesis of...