Controlled synthesis and size-dependent polarization domain structure of colloidal germanium telluride nanocrystals

Article abstract of DOI:10.1021/ja108309s

Germanium telluride (GeTe) exhibits interesting materials properties, including a reversible amorphous-to-crystalline phase transition and a room-temperature ferroelectric distortion, and has demonstrated potential for nonvolatile memory applications. Here, a colloidal approach to the synthesis of GeTe nanocrystals over a wide range of sizes is demonstrated. These nanocrystals have size distributions of 10-20% and exist in the rhombohedral structure characteristic of the low-temperature polar phase. The production of nanocrystals of widely varying sizes is facilitated by the use of Ge(II) precursors with different reactivities. A transition from a monodomain state to a state with multiple polarization domains is observed with increasing size, leading to the formation of richly faceted nanostructures. These results provide a starting point for deeper investigation into the size-scaling and fundamental nature of polar-ordering and phase-change processes in nanoscale systems.

Full text of DOI:10.1021/ja108309s

COMMUNICATION
pubs.acs.org/JACS
Controlled Synthesis and Size-Dependent Polarization Domain
Structure of Colloidal Germanium Telluride Nanocrystals
||
||
Mark J. Polking,† Haimei Zheng,‡,§ Ramamoorthy Ramesh,*,†, and A. Paul Alivisatos*,‡,
†Department of Materials Science and Engineering and ‡Department of Chemistry, University of California, Berkeley, California 94720,
United States
§National Center for Electron Microscopy and Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley,
California 94720, United States
S Supporting Information
b
nanoparticles have been reported,12-16 the size scales of these
ABSTRACT: Germanium telluride (GeTe) exhibits inter-
esting materials properties, including a reversible amorphous-
to-crystalline phase transition and a room-temperature ferro-
electric distortion, and has demonstrated potential for non-
volatile memory applications. Here, a colloidal approach to
the synthesis of GeTe nanocrystals over a wide range of sizes
is demonstrated. These nanocrystals have size distributions of
10-20% and exist in the rhombohedral structure character-
istic of the low-temperature polar phase. The production of
nanocrystals of widely varying sizes is facilitated by the use of
Ge(II) precursors with different reactivities. A transition from
a monodomain state to a state with multiple polarization
domains is observed with increasing size, leading to the
formation of richly faceted nanostructures. These results
provide a starting point for deeper investigation into the
size-scaling and fundamental nature of polar-ordering and
phase-change processes in nanoscale systems.
materials are far from the quantum regime, and the formation of
crystalline GeTe of controlled sizes has remained elusive. In
addition, while colloidal chemistry has proven highly successful
for the synthesis of semiconducting and metallic nanomaterials
with tunable optical, magnetic, and other functionalities, few
syntheses of low-dimensional nanostructures of materials exhi-
biting spontaneous polar ordering exist,17-19 hindering funda-
mental study of polar phenomena at nanoscale dimensions.
Here, we describe a simple and highly adaptable synthesis of
GeTe nanocrystals of sizes ranging from 8 to 100 nm using
colloidal chemistry. These nanocrystals have narrow size dis-
tributions and exhibit the rhombohedral structure characteristic
of the polar phase down to particle sizes of less than 10 nm. We
observe a transition from a primarily monodomain state to a
multidomain state for length scales above ∼30 nm, indicating a
critical size scale for the emergence of a polarization domain
structure.
GeTe nanocrystals with average sizes of 8, 17, and 100 nm
were synthesized by reaction of the divalent germanium pre-
cursors Ge(II) chloride-1,4 dioxane complex and bis[bis-
(trimethylsilyl)amino]Ge(II) ((TMS2N)2Ge) with trioctyl-
phosphine-tellurium (TOP-Te). Phase-pure GeTe nanocrystals
can be prepared in a variety of solvents, including 1,2-dichlor-
obenzene, 1-octadecene, phenyl ether, and others. The basic
chemistry is also compatible with other surfactants, including
1-dodecanethiol, oleylamine, and phosphonic acids.
Nanocrystals with an average diameter of 8 nm were synthe-
sized by the reaction of (TMS2N)2Ge with TOP-Te in
the presence of 1-dodecanethiol and excess trioctylphosphine
at 230 ꢀC, and nanocrystals with a 17 nm average diameter were
prepared using the same precursors in the presence of oleylamine
at 250 ꢀC. The synthesis of nanocrystals with an average size of
100 nm was accomplished through the use of GeCl2-dioxane
complex and TOP-Te in the presence of 1-dodecanethiol at
180 ꢀC. Prior to the syntheses, solvents were dried and degassed
where appropriate, which was found to be crucial to the produc-
tion of phase-pure GeTe. Full details on all syntheses are
provided in the Supporting Information (SI).
emiconducting IV-VI nanocrystals have received attention
S
recently due to their strong quantum size effects1-3 and rich
array of phase transitions that influence their electronic, optical,
and phononic properties.4,5 The semiconductor germanium
telluride (GeTe) in particular has garnered interest due to
its reversible amorphous-to-crystalline phase transition6,7 and
ferroelectric phase transition, which leads to a spontaneous
polarization along a Æ111æ axis below ∼625 K.8-10 This polar
distortion also leads to the formation of polarization domain
boundaries, which influence its mechanical, electronic, ther-
mal, and other properties.11 The simplest possible ferroelec-
tric, GeTe provides a simple model system for the study of
polar-ordering phenomena at reduced dimensions, including
the question of a critical length scale for the emergence of a
polarization domain structure. Although the high bulk carrier
density of GeTe hinders direct measurement of the sponta-
neous polarization, the interplay of structural, optical, electro-
nic, and other properties makes this an important system for
further study.
The production of GeTe nanocrystals in two size regimes is
facilitated by the use of two precursors, GeCl2-dioxane complex
and (TMS2N)2Ge, with vastly different reaction kinetics. The
Despite interest in GeTe, methods for the synthesis of high-
quality nanomaterials of GeTe are relatively unexplored,
and little is currently known about its nanoscale properties.
While vapor-phase syntheses of GeTe nanowires and solution-
phase syntheses of micrometer-scale crystals and amorphous
Received: September 14, 2010
Published: January 31, 2011
r
2011 American Chemical Society
2044
dx.doi.org/10.1021/ja108309s J. Am. Chem. Soc. 2011, 133, 2044–2047
|

Journal of the American Chemical Society
COMMUNICATION
Figure 2. Powder X-ray diffraction patterns for 8, 17, and 100 nm GeTe
nanocrystals. The patterns indicate the presence of phase-pure GeTe in
the rhombohedral R3m (160) space group.
Figure 1. Transmission electron microscope (TEM) images and size
statistics for GeTe nanocrystals. (A-C) Low-resolution TEM image
(A), high-resolution TEM (HRTEM) image (B), and size statistics (C)
for 8 nm GeTe nanocrystals. (D-F) Low-resolution TEM image (D),
HRTEM image (E), and size statistics (F) for 17 nm GeTe nanocrystals.
(G-I) Low-resolution TEM image (G), HRTEM image (H), and size
statistics (I) for 100 nm GeTe nanocrystals. N is the number of particles
measured.
of the rhombohedral phase. Significant peak broadening for the
8 nm particles, however, complicates determination of the
material phase, necessitating analysis by Rietveld refinement
(SI, Figure S2). This analysis confirms the presence of the
rhombohedral phase with a rhombohedral angle of approxi-
mately 88.5ꢀ. Multiple trials consistently indicated a rhombohe-
dral distortion.
latter precursor often yielded nucleation within several seconds
in many different surfactant/solvent systems, whereas the former
reacted sluggishly, if at all, under the same reaction conditions.
Both precursors are readily reduced in the presence of primary
amines and alkanethiols to yield Ge(0) nanocrystals. GeCl2-
dioxane reacts sluggishly with pure oleylamine at 300 ꢀC and
produces germanium nanocrystals with an average size of
approximately 15 nm (SI, Figure S1A); reduction of (TM-
S2N)2Ge under the same reaction conditions results in rapid
nucleation of ∼3 nm particles (Figure S1B). Similar trends are
observed at lower temperatures. Although the precise reaction
mechanism is not fully understood, this large increase in the
Ge(II) reduction rate may contribute to the increase in particle
nucleation rate and decrease in diameter from GeCl2-dioxane to
(TMS2N)2Ge, consistent with a previous study indicating the
vital role of Ge(II) reduction kinetics in GeTe formation.15
Typical transmission electron microscope (TEM) images for
8, 17, and 100 nm nanocrystals are shown in Figure 1. High-
resolution TEM (HRTEM) imaging indicates that particles of all
sizes are crystalline, and size statistics collected on samples of
over 250 particles demonstrate size distributions of 10-20%,
typical values for many colloidal syntheses.
Powder X-ray diffraction (XRD, Figure 2) was performed with
a Bruker AXS GADDS D-8 diffractometer (Co KR, 1.79026 Å).
All patterns confirm the presence of phase-pure GeTe. The polar
phase transition in GeTe from a rock salt structure to a
rhombohedral structure results in the splitting of the 111 and
220 diffraction lines (in the cubic indexing system) into 003-
021 and 024-220 doublets (in the rhombohedral indexing
system).9,10 The XRD patterns of the 17 and 100 nm particles
exhibit clear splitting of the 111 and 220 doublets characteristic
TEM and electron diffraction studies (Figures 3 and 4) reveal
the size-dependent evolution from a primarily monodomain
state to a state with multiple polarization domains. Many IV-VI
materials form in low-symmetry structures with weak bonding
along certain crystallographic directions, leading to stacking
faults, twin boundaries, and other defects.4,20 GeTe in the
rhombohedral phase forms {100} and {110} twin boundaries
as well as inversion boundaries separating domains with different
Æ111æ polarization axes.11 HRTEM investigations of the 8 and 17
nm particles indicate that these particles primarily consist of a
single domain. Electron diffraction patterns of the 100 nm
particles, in contrast, reveal splitting of diffraction spots consis-
tent with the formation of {100} and {110} twin boundaries
(Figure 3), and scanning electron microscope (SEM) and
Z-contrast TEM imaging (SI, Figure S3) show extensive faceting.
The change in lattice orientation across {100} and {110} domain
boundaries leads to pronounced diffraction contrast in dark-field
images.21 The breakdown of Friedel symmetry for noncentro-
symmetric crystals leads to differences in background contrast
between inversion domains in dark-field images, and observation
using complementary g and -g diffraction vectors results in
B
B
a reversal of the relative contrast.21 Dark-field TEM images of
the 100 nm particles (Figure 3) reveal the presence of all three
types of domain boundaries and the spatial relationships between
domains. These images suggest that the 100 nm particles are
largely bidomain. The electron diffraction patterns of the 8 nm
particles (SI, Figure S4), in contrast, contain a single set of spots,
and dark-field TEM images reveal largely uniform contrast,
confirming the monodomain structure.
To study the domain structure in still larger particles, addi-
tional precursor was injected into a solution containing seed
2045
dx.doi.org/10.1021/ja108309s |J. Am. Chem. Soc. 2011, 133, 2044–2047

Journal of the American Chemical Society
COMMUNICATION
Figure 4. Polarization domains in highly faceted GeTe nanostructures.
(A,B) Scanning electron microscope images of highly faceted GeTe
nanostructures. (C) TEM image and (D) corresponding electron
diffraction pattern of a single GeTe nanostructure illustrating splitting
of diffraction spots consistent with {100} twin boundary formation.
Inset: Splitting of the 024 diffraction spot along the 020 direction.
Figure 3. Polarization domains in 100 nm GeTe nanocrystals. (A-C)
ceramic materials.25 Ferroelastic domain formation is disfavored,
however, in unconstrained crystals26 but has nonetheless been
observed in free-standing BaTiO3 films due to the intrinsic
tension imposed by a surface layer with different structural
properties.26 Domain formation in our GeTe nanocrystals may
proceed via a similar mechanism. Although the high carrier
density in GeTe may significantly screen the polarization, an
additional electrostatic driving force for domain walls may be
present. The formation of purely ferroelectric 180ꢀ walls23
suggests an electrostatic influence on domain formation.
Bright-field TEM image (A), dark-field TEM image taken with g = 002
B
(B), and corresponding electron diffraction pattern (C) consistent with
a (100) twin boundary (109ꢀ domain wall). (E-G) Bright-field TEM
image (E), dark-field TEM image taken with g = 422 (F), and
B
corresponding electron diffraction pattern (G) consistent with a (110)
twin boundary (71ꢀ domain wall). (I-K) Bright-field TEM image (I)
and dark-field TEM images taken with g = 002 (J) and g =002 (K),
B
B
respectively, indicating reversal of domain contrast characteristic of an
inversion domain boundary (180ꢀ domain wall). (D,H,L) Schematic
illustrations of the relationships between polarization vectors across
109ꢀ, 71ꢀ, and 180ꢀ domain walls.
This study demonstrates the synthesis of nanocrystals of the
semiconductor GeTe with controlled sizes and narrow size
distributions using colloidal techniques. Preliminary analysis
indicates the presence of a rhombohedral distortion and polar-
ization domains with a characteristic size scale of ∼30-50 nm.
Detailed studies to elucidate the size-dependent polar ordering in
these materials are currently underway. These particles should
serve as a basis for future fundamental studies of ferroelectric
ordering and phase transitions at reduced dimensions and the
complex interrelationships between structural, electronic, opti-
cal, and other parameters in IV-VI materials.
particles, resulting in the formation of highly faceted nanostruc-
tures of regular diameter exhibiting surface features with a length
scale of approximately 30 nm (Figure 4). Electron diffraction
patterns of these crystals reveal similar splitting of diffraction
spots, indicative of {100} and {110} twin boundary formation
(Figure 4C,D, and SI, Figure S5), and exhibit no evidence of
diffraction rings characteristic of polycrystals. Dark-field TEM
imaging (SI, Figure S6) shows a complex contrast pattern with
many strongly diffracting regions, suggesting the presence of
numerous polarization domains in the particles. This is consis-
tent with the observed splitting of diffraction spots along multiple
directions (Figures S5 and S6), indicating a polydomain state.
Sequential addition of precursor to these structures thus results
in the formation of networks of polarization domains rather than
simple conformal addition of material to the particle surface.
These observations suggest an average domain size of approxi-
mately 20-50 nm, consistent with literature reports on GeTe
thin films.22
’ ASSOCIATED CONTENT
S
Supporting Information. Materials and methods for all
b
syntheses and experiments, TEM images of Ge nanocrystals,
additional SEM images, TEM images, electron diffraction pat-
terns, and diffraction results for GeTe nanomaterials. This
material is available free of charge via the Internet at http://
pubs.acs.org.
The formation of domain boundaries in polar-ordered materi-
als arises from a balance among electrostatic energy, elastic
energy, and the energy of domain wall formation.23 Periodic
arrays of ferroelastic domain walls form to alleviate epitaxial
strains in thin films24 or strains imposed by surrounding grains in
’ AUTHOR INFORMATION
Corresponding Author
rramesh@berkeley.edu; alivis@berkeley.edu
2046
dx.doi.org/10.1021/ja108309s |J. Am. Chem. Soc. 2011, 133, 2044–2047

Journal of the American Chemical Society
COMMUNICATION
’ ACKNOWLEDGMENT
The authors thank Dmitri Talapin, Jonathan Owen, Ronald
Gronsky, Jeffrey Urban, and Delia Milliron for helpful discussions
and Paul Trudeau for technical assistance. Transmission electron
microscopy work performed at the National Center for Electron
Microscopy, Lawrence Berkeley National Laboratory, was sup-
ported by the Office of Science, Office of Basic Energy Sciences, of
the U.S. Department of Energy under Contract No. DE-AC02-
05CH11231. The remainder of this work was supported under the
Physical Chemistry of Nanocrystals Project of the Director, Office
of Science, Office of Basic Energy Sciences, Materials Sciences and
Engineering Division, of the U.S. Department of Energy under
Contract No. DE-AC02-05CH11231. M.J.P. was supported by a
National Science Foundation Graduate Research Fellowship and
by a National Science Foundation Integrated Graduate Education
and Research Traineeship (IGERT) fellowship.
’ REFERENCES
(1) Baumgardner, W. J.; Choi, J. J.; Lim, Y.-F.; Hanrath, T. J. Am.
Chem. Soc. 2010, 132, 9519.
(2) Hickey, S. G.; Waurisch, C.; Rellinghaus, B.; Eychmueller, A.
J. Am. Chem. Soc. 2008, 130, 14978.
(3) Kovalenko, M. V.; Heiss, W.; Shevchenko, E. V.; Lee, J.-S.;
Schwinghammer, H.; Alivisatos, A. P.; Talapin, D. V. J. Am. Chem. Soc.
2007, 129, 11354.
(4) Littlewood, P. B. J. Phys. C: Solid State Phys. 1980, 13, 4855.
(5) Waghmare, U. V.; Spaldin, N. A.; Kandpal, H. C.; Seshadri, R.
Phys. Rev. B 2003, 67, 125111.
(6) Andrikopoulos, K. S.; Yannopoulos, S. N.; Voyiatzis, G. A.;
Kolobov, A. V.; Ribes, M.; Tominaga, J. J. Phys.: Condens. Matter
2006, 18, 965.
(7) Lencer, D.; Salinga, M.; Grabowski, B.; Hickel, T.; Neugebauer,
J.; Wuttig, M. Nat. Mater. 2008, 7, 972.
(8) Steigmeier, E. F.; Harbeke, G. Solid State Commun. 1970, 8, 1275.
(9) Goldak, J.; Barrett, C. S.; Innes, D.; Youdelis, W. J. Chem. Phys.
1966, 44, 3323.
(10) Chattopadhyay, T.; Boucherle, J. X.; Von Schnering, H. G.
J. Phys. C: Solid State Phys. 1987, 20, 1431.
(11) Snykers, M.; Delavignette, P.; Amelinckx, S. Mater. Res. Bull.
1972, 7, 831.
(12) Tuan, H.-Y.; Korgel, B. A. Cryst. Growth Des. 2008, 8, 2555.
(13) Meister, S.; Peng, H.; McIlwrath, K.; Jarausch, K.; Zhang, X. F.;
Cui, Y. Nano Lett. 2006, 6, 1514.
(14) Yu, D.; Wu, J.; Gu, Q.; Park, H. J. Am. Chem. Soc. 2006, 128, 8148.
(15) Buck, M. R.;Sines, I. T.;Schaak, R. E. Chem. Mater. 2010, 22, 3236.
(16) Caldwell, M. A.; Raoux, S.; Wang, R. Y.; Philip Wong, H.-S.;
Milliron, D. J. J. Mater. Chem. 2010, 20, 1285.
(17) Urban, J. J.; Yun, W. S.; Gu, Q.; Park, H. J. Am. Chem. Soc. 2002,
124, 1186.
(18) Adireddy, S.; Lin, C.; Cao, B.; Zhou, W.; Caruntu, G. Chem.
Mater. 2010, 22, 1946.
(19) O’Brien, S.; Brus, L.; Murray, C. B. J. Am. Chem. Soc. 2001, 123,
12085.
(20) Enders, P. Phys. Stat. Sol. B 1983, 120, 735.
(21) Snykers, M.; Serneels, R.; Delavignette, P.; Gevers, R.; Van
Landuyt, J.; Amelinckx, S. Phys. Stat. Sol. A 1977, 41, 51.
(22) Yashina, L. V.; Puettner, R.; Neudachina, V. S.; Zyubina, T. S.;
Shtanov, V. I.; Poygin, M. V. J. Appl. Phys. 2008, 103, 094909.
(23) Lines, M. E.; Glass, A. M. Principles and Applications of Ferro-
electrics and Related Materials; Clarendon Press: Oxford, 1977; pp
87-102.
(24) Roitburd, A. L. Phys. Stat. Sol. A 1976, 37, 329.
(25) Arlt, G. J. Mater. Sci. 1990, 25, 2655.
(26) Luk’yanchuk, I. A.; Schilling, A.; Gregg, J. M.; Catalan, G.; Scott,
J. F. Phys. Rev. B 2009, 79, 144111.
2047
dx.doi.org/10.1021/ja108309s |J. Am. Chem. Soc. 2011, 133, 2044–2047