Accepted Manuscript

Role of chemical potential and limitations of growth kinematics on stoichiometry of

indium antimonide nanowires

Zaina Algarni, Abhay Singh, U. Philipose

PII: S0254-0584(18)30640-0
DOI: 10.1016/j.matchemphys.2018.07.049
Reference: MAC 20834

To appear in: Materials Chemistry and Physics

Received Date: 29 November 2017

Accepted Date: 20 July 2018

Please cite this article as: Z. Algarni, A. Singh, U. Philipose, Role of chemical potential and limitations of
growth kinematics on stoichiometry of indium antimonide nanowires, Materials Chemistry and Physics
(2018), doi: 10.1016/j.matchemphys.2018.07.049.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Role of chemical potential and limitations of growth kinematics

on stoichiometry of indium antimonide nanowires

Zaina Algarni, Abhay Singh and U. Philipose1

1
Department of Physics, University of North Texas, Denton, Texas, 76203

PT
(Dated: October 31, 2017)

Abstract

RI
The conditions governing the growth of indium antimonide (InSb) nanowires by antimonidiza-

SC
tion of indium (In) droplets in a Chemical Vapor Deposition (CVD) system are examined . We
report on factors that affect the InSb nanowire morphology and stoichiometry when growth is cat-
alyzed by a low surface energy metal like In. Using thermodynamics and kinematics of nanowire

U
growth, theoretical predictions are compared to experimental data. Results detailing the depen-
AN
dence of morphology and composition of these nanowires on temperature and diameter of the
alloyed droplet are analyzed to show the dependence of these parameters on the chemical potential
M
between vapor, liquid and solid phases. The Gibbs-Thomson (GT) effect is used to explain lower
effective supersaturation and slower nucleation in small sized droplets, resulting in nanowires with
D

diameters below a critical diameter range to be comprised of pure In.

TE
C EP
AC

Typeset by REVTEX 1
 ACCEPTED MANUSCRIPT

I. INTRODUCTION

Group III-V semiconductors have many properties that make them attractive for electronic
and opto-electronic applications. These include narrow, direct energy band gap and high
carrier mobilities. Indium antimonide (InSb) belongs to this group of materials and its many

PT
technologically significant properties include a band gap of 0.17 eV at room temperature1,2 ,
high electron mobility (77000 cm2 V −1 s−1 )3 , small effective mass (m∗ = 0.013 me ), and

RI
ballistic length of 0.7 µm at room temperature4,5 . Because of these properties, this material
has a wide range of applications including infrared optical detectors and high speed de-

SC
vices6–8 . On account of its large Bohr radius (60 nm7,9,10 ), quantum confinement effects can
be observed in InSb at relatively large diameters. This makes it possible to tune the energy

U
band gap of InSb nanowires as a function of its diameter, enable fabrication of core-shell
heterostructure making them particularly attractive in the development of multi-spectrum
photon detectors and photovoltaics 7,11
AN . Recently, there have been several reports on the
synthesis of InSb nanowires12–16 . This effort is justified because when compared to bulk,
M
nanowires have unique size-dependent properties and can also be grown as heterostructures
with controlled dopant concentration and defect density.
D

In most cases, growth of InSb nanowires follows the vapor-liquid-solid (VLS) growth
TE

mechanism10,17,18 where growth is initiated by metal (gold) seeds, which function as a cata-
lyst3,4,19,20 . InSb nanowires have also been synthesized by a self-catalyzed growth process,
using a thin film of In as the seed layer2,21,22 . There are several details of the VLS growth
EP

mechanism and the role of the metal seed in determining the nanowire geometry and com-
position that are still under debate. The growth temperature for InSb nanowires grown in a
C

CVD system is above the congruent sublimation temperature of InSb (≈ 270 − 400◦ C) 23,24 .
AC

This implies that at the so-called ‘in-congruent sublimation point’, InSb will dissociate and
the more volatile Sb will first evaporate from the solid InSb source as tetramers (Sb4 ) or
as dimers (Sb2 ) 25
. In the temperature range of (≈ 430 − 600◦ C), Sb4 will be adsorbed in
the In droplet and dissociate into Sb2 molecules26 . The dissociated Sb2 molecules will then
react with In to form InSb. Above this temperature range, Sb4 molecules are desorbed
from the droplet surface, leaving behind a molten In droplet, which as growth progresses
will be supersaturated with its own vapor, resulting in growth of pure In nanowires. III-Sb
nanowires are also known to have larger diameters than III-As or III-P nanowires because

2
 ACCEPTED MANUSCRIPT

the nucleation from the seed particle occurs at high concentrations of In and Sb27 . This
work investigates these effects and addresses the difficulty in growing stoichiometric InSb
nanowires by low vacuum, inexpensive techniques like the CVD.

PT
II. EXPERIMENTAL DETAILS

The InSb nanowires discussed in this paper were grown by a self-catalyzed process using

RI
In as the metal seed layer in a CVD system. At a given growth temperature, the size of
these metal seeds were found to influence both the morphology and composition of the InSb

SC
nanowires that were grown in a quartz tube placed in a Lindberg Blue tube furnace. Details
21,28
of the growth process have been discussed in our earlier works . Solid InSb crystals

U
(Alpha Aesar,99.99%) were loaded into an alumina boat and the In coated Si substrates
were placed face-down directly above the source. The furnace was then flushed by flowing a
AN
mixture of H2 and Ar gas at (50 sccm) for 15 min, at room temperature. The source tem-
perature was ramped to 560◦ C and InSb nanowire growth was carried out at 530◦ C for 90
M
min at close to atmospheric pressure in a reducing environment of Ar:H2 gas at 50:50 sccm.
After growth, the as-grown InSb nanowires were characterized by scanning electron micro-
D

scope (SEM, Hitachi SU 1510), energy dispersive X-ray (EDX) spectroscopy and Raman
TE

spectroscopy to study their morphology and composition. Raman measurements were made
on these nanowires at ambient temperature using a Nicolet Almega XR–Raman spectrom-
eter equipped with a 532 nm green laser. To obtain information on carrier concentration in
EP

these nanowires, they were connected in a field-effect transistor type configuration, with the
highly doped p+ Si substrate (resistivity of 0.001–0.005 Ω.cm) covered with a 200 nm thick
C

SiO2 layer, serving as the back gate. Using a semiconductor parametric analyzer (Agilent
AC

Technologies B1500A), transport measurements were performed on individual nanowires

contacted by Cr/Au electrodes.

III. RESULTS AND DISCUSSION

During the initial stages of growth, as the solid InSb source is heated above its congruent
sublimation point, there is a much higher vapor pressure of Sb compared to In in the growth
tube. The VLS growth mechanism as explained by the In-Sb phase diagram29 requires that

3
 ACCEPTED MANUSCRIPT

there be a steady increase in the concentration of Sb in the alloy droplet for InSb nucleation
and subsequent nanowire growth. However, this is not possible if the vapor pressure of Sb
continuously decreases. Moreover, as the InSb source loses Sb it becomes rich in In and
hence during the later stages of growth, there is a higher In vapor pressure, causing In
vapors to condense in the alloy, making it an In-rich InSb droplet. These predictions were

PT
verified by experiments. There are several critical growth parameters that determine the
stoichiometry, growth rate and morphology of the InSb nanowires.

RI
A. Role of surface tension

SC
Considering the use of In as the metal catalyst, stable VLS growth of InSb nanowires is
possible only when the droplet shape and its surface tension satisfy criteria dictated by

U
the growth kinematics. The nanowire growth process involves formation of two-dimensional
AN
monolayers that nucleate and extend layer-by-layer from the In-Sb droplet to form the InSb
nanowire. The alloy droplet composition is governed by the In-Sb phase diagram29 and its
M
volume is expressed by the relation7 :

Valloy = xIn VIn + xSb VSb (1)

D

where xIn and xSb represent mole fractions of In and Sb respectively. Figure 1(a) shows a
TE

vector equilibrium diagram of a molten droplet resting on a solid surface. In the early stages
of growth, the inclination angle α is equal to zero and the contact angle θ is less than π/2
EP

[Figure 1(a)]. In this situation, the droplet is in equilibrium with its surroundings and the
sum of the net horizontal forces that acting on the triple phase line (TPL) is equal to zero.
C

Inset shows an SEM image of the droplet under this condition. Balancing the horizontal
components of surface tensions between liquid–solid (σls ) and liquid–vapor (σlv ) phases, the
AC

Neumann triangle equation that describes the relation between the contact angle (θ), the
inclination angle (α), and the surface tensions is expressed as30,31 :

σls = σsv cosα − σlv cosθ (2)

As growth proceeds, the vapor pressure of Sb increases and saturates the molten droplet.
During this stage of growth α and θ increases (Figure 1(b)), and eventually α = 90◦ . As
new layers of InSb form at the liquid-solid interface, the droplet is dragged upward to the

4
 ACCEPTED MANUSCRIPT

edge of the new layer. This is because there exists a net upward force (Figure 1(c)), which
causes the droplet to wet the nanowire growth front but not its sidewalls. An SEM image
of the growing InSb nanowire with the droplet riding on its tip is shown in the inset of
Figure 1(c). The inequality between the surface tensions, is expressed by equation 331,32 :

PT
σlv sinθ + σls > σsv (3)

RI
If the surface tension of the In-Sb alloy droplet is too low, there can be net surface forces
directed downward, tending to increase the part of the nanowire wetted by the liquid. As

SC
the nanowire penetrates into the droplet, the vertical forces will continue to increase and
the droplet will continue sliding down until the solid nanowire fully penetrates the droplet,
resulting in an unstable nucleation event and no nanowire growth as evidenced by the bulky

U
cones of InSb in the SEM image shown in the inset of Figure 1(d). The choice of In as the
AN
metal catalyst as well as the composition of the alloy droplet (as determined by the phase
diagram) are thus critical in determining the surface tension of the droplet. 1-D growth is
M
possible only when the inequality expressed by equation 3 is satisfied.
To assess the appropriateness of the temperature used in this experiment, the surface
D

tensions of the liquid and solid phases (σlv and σsv ) at the growth temperature of 530◦ C
was estimated and found to be ≈ 0.45 N/m, 0.6 N/m respectively. From the SEM image
TE

in the inset of Figure 1(c), the contact angle is about 130◦ , which shows that at the growth
temperature, the surface tensions satisfy the inequality expressed by equation 3.
EP

B. Role of chemical potential

C
AC

Another factor that controls the nanowire growth kinematics and the stability of the var-
ious phases involved in the growth process is the Gibbs energy. Since, VLS growth is an
isothermal growth process, and since the pressure is kept constant during growth, the stable
phases are the ones where the Gibbs energy is a minimum. Under steady state conditions,
in unit time, the number of atoms condensing from the In and Sb vapor to the liquid In-Sb
droplet is equal to the number of atoms transferring from the liquid droplet to the growing
crystal. During the initial stages of growth, the droplet is mostly comprised of In. The
pressure inside this droplet (Pin ) and outside it (Pout ) are governed by the Young-Laplace

5
 ACCEPTED MANUSCRIPT

equation, expressed as33,34 :

2σlv
Pin = Pout + (4)
R
The significance of the droplet radius (R) is evident in equation 4, which shows that for
small radii droplets, the pressure difference becomes significant. The pressure outside the

PT
droplet is initially controlled by the vapor pressure of Sb, since when the source temperature
exceeds the in-congruent sublimation temperature of InSb (270◦ )C23 , there exists initially
a much higher vapor pressure of Sb than In in the growth chamber. Subsequently, the Sb

RI
vapor pressure decreases as growth proceeds35 and hence is considered to be a rate-limiting
factor during InSb nanowire growth. As growth progresses, the In seed is an alloyed droplet

SC
comprising of mole fractions of In and Sb and its Gibbs energy is a function of temperature,
pressure, composition as well as size of the seed. Due to the high surface to volume ratio of

U
the alloy seed, the total Gibbs energy will also include a contribution from its surface energy
AN
and is therefore added to the Gibbs energy of the bulk system. The Gibbs energy can thus
be expressed as7 :
G = Gbulk + Gsurf ace (5)
M
The change in Gibbs energy, dG, when small amounts of In and Sb are transferred between
D

the vapor and the liquid phase is expressed as a differential in the form33 :
TE

dG = −SdT + V dP + µv dnv + µlv−l dnlv−l + σlv dA (6)

where S represents entropy of the spherical droplet with surface area A, µlv−l represents the
EP

chemical potential of the alloy liquid phase when there is condensation of vapor to liquid
and µv is the chemical potential of the vapor phase. At a given growth temperature (T)
C

and pressure (Pout ) in the growth chamber, when the vapor and the liquid phases are in
equilibrium, using equation 4, the following condition must be satisfied33 :
AC

2σlv
µl,eq l,eq
v−l (Pin , T ) = µv−l (Pout + , T ) = µv (Pout , T ) (7)
Req

The superscript ‘eq’ is used to indicate the chemical potential of the liquid droplet with radius
Req that is in thermodynamic equilibrium with the surrounding vapor. During growth, the
surface curvature of the droplet (1/R) will adjust itself to an equilibrium surface curvature
(1/Req ). These changes are triggered by an attempt to attain equilibrium between the vapor
and liquid phases. Before reaching equilibrium, when µl is greater than µv , the droplet will

6
 ACCEPTED MANUSCRIPT

evaporate, causing its volume (Ωl ) to decrease. This decrease in volume of liquid droplet
leads to an increase in its diameter (reduction in its surface curvature 1/R). When R increases
above Req , equation 7 implies that the droplets chemical potential is less than that of the
vapor which causes vapor to condense into the liquid phase and eventually reach a state of
equilibrium. Defining pressure P∞ as the vapor pressure in the bulk limit or flat surface

PT
(infinite radius), the chemical potential of the liquid at P∞ is denoted as µl,∞
v−l and since for

a spherical droplet since dA/dn = 2Ωl /R, the change in chemical potential can be used to

RI
determine the critical radius of the droplet at a given temperature and pressure.

2Ωl σlv

SC
∆µl = Ωl (Pout − P∞ ) + (8)
R

Since ∆µl can be expressed as kB T ln(Pout /P∞ ), the critical radius can be expressed as:

U
2Ωl σlv
Rc = (9)
AN kB T ln(Pout /P∞ )

This is the Gibbs-Thomson equation that gives the relation between supersaturation and the
radius of liquid alloy31,36 . The radius (R) of the spherical droplet is related to the nanowire
M
radius(r) by the relation r = R sinθ. The nanowire growth rate (dL/dt) can be defined in
terms of the critical radius as37 :
D

dL Rc
= C(1 − ) (10)
dt r
TE

where C is a constant that depends on growth conditions like supersaturation and temper-
ature. As the nanowire radius tends to the critical radius Rc (equation 9), the growth rate
EP

dL/dt (equation 10) tends to zero, which implies that Rc is the kinetically allowed minimum
radius of the nanowire.
C

As growth progresses and the droplet tries to attain thermodynamic equilibrium, some
of the liquid at the liquid-solid interface will solidify since its chemical potential is higher
AC

than that of the solid phase. Applying the same analysis (as that described for the vapor-
liquid system) to the liquid-solid system, at constant pressure and temperature, the chemical
potential between liquid and solid phases can be expressed as33 :

σls
µll−s = µs − (11)
Al−s

Where σls is defined by equation 2 at α = 90◦ . Al−s is the areal density of atoms at the
liquid-solid interface. Comparing equations 8 and 11, nanowire growth is triggered when

7
 ACCEPTED MANUSCRIPT

there exists the following relationship of chemical potentials between the vapor, liquid and
solid phases33 :
µlv−l > µll−s (12)

If µll−s is greater than µlv−l , solidification cannot be approached and nanowire growth will not

PT
occur. For nanowire growth, the critical stage is the onset of the supersaturation nucleation
process. During this stage, the liquid in the droplet will start to solidify only when the
difference in chemical potentials between the phases satisfy equation 12. Figure 2 shows

RI
a plot of µlv−l and µll−s at different contact angles for InSb nanowire of radius 60 nm.
The chemical potentials were estimated based on the values of surface tension obtained

SC
at the growth temperature of 530◦ C, with the molar volume of liquid phase set at 1.8 ×
10−5 m3 /mole7 . The numerical plots were obtained using the equations:

U
300
µlv−l = 1000[12 + sinθ] (13)
r
and
AN
µll−s = 1000[35 + 30cosθ] (14)
M
As seen in the plot (Figure 2), the inequality expressed by equation 12 is satisfied when
angle θ approaches 130◦ . At this angle, the chemical potential is (≈ 16 kJ/mole) and
D

solidification is energetically feasible. InSb nanowire growth resulting from solidification

TE

from the liquid droplet will only commence when the inequality expressed by equation 12
is met. As explained in the previous section, at this stage of growth, there is a decrease in
R and an increase in θ. Though the inequality criteria is satisfied for all θ >130◦ ,there are
EP

limitations on θ imposed by the surface tension considerations expressed in equation 3 and

equation 2.
C
AC

C. The role of carrier gas on nanowire composition

During the synthesis of InSb nanowires by CVD, the biggest challenge is contamination with
oxygen. In experiments where Ar is used as the carrier gas, it might be virtually impossible
to remove all traces of oxygen. Argon gas is distilled from air and since its vapor pressure
is comparable to that of O2 it always contains trace amounts of O2 as impurity. During
the VLS growth of InSb nanowires, any trace of O2 in the growth chamber will result in
In or Sb oxide formation. The only way to circumvent this problem is to create a reducing

8
 ACCEPTED MANUSCRIPT

environment in the growth chamber by using a mixture of H2 and Ar as the carrier gas.
However, a high flow rate of hydrogen will lower the vapor pressure of Sb and drive out
vapors of Sb from the tube38 . This will result in a high In content to accumulate in the
droplet over the time, resulting in formation of In droplets or In nanowires. It is therefore
essential to lower the flow rate of H2 to achieve a high Sb vapor pressure.

PT
RI
D. Maintaining stoichiometry of InSb

During the initial growth phase, the In film breaks up into droplets with sizes varying from

SC
about 50 - 200 nm, as shown in the SEM images of Figure 3. The non-uniformity in the size
of these droplets was found to have a direct relationship to the nanowire composition. Post-

U
growth analysis of these nanowires showed that many of the droplets with radius exceeding

AN
100 nm and less than 40 nm were pure In (Figure 4(a)), while nanowires with radius in the
range of 50–60 nm were InSb (Figure 4(b)). As can be seen from the diameter-composition
plot in Figure 5, nanowires grown at 530◦ C with diameter in the range of 100-130 nm have
M
an In/Sb ratio of ≈ 1. However, the thin nanowires (diameters in the range of 60-80 nm)
have a very high concentration of In and in some cases, they are purely In. The same was
D

also true for very thick nanowires. The variance in stoichiometry is explained on the basis
TE

of the Gibbs-Thomson effect. In equation 9 defined above, the output pressure (Pout ) can
be interpreted as the partial pressure of Sb. In this case equation 9 requires that there
are limitations on the critical radius, beyond which nucleation and growth of InSb is not
EP

possible. In other words, the Gibbs-Thomson effect limits the Sb supersaturation in the
seed particle. It was found that InSb nanowires cannot nucleate from small radii droplets,
C

as an increased Sb pressure in the droplet will reduce the driving force required to transfer
AC

more Sb molecules from vapor to liquid thus preventing supersaturation. Also, as indicated
by equation 10, the growth rate is slow for thinner nanowires. This slow growth rate could
result in the loss of Sb from the growing crystal if the growth temperature is high enough for
dissociation to occur. On the other hand, if the source temperature is lowered sufficiently
to lower Sb vapor pressure, equation 9 predicts that this will result in an increase in Rc and
the droplet will grow radially beyond the critical values required for nucleation. Due to the
flow of carrier gases, these vapors have a short residence time in the growth tube and so the
ramp rate of temperature to its set-point is also critical.

9
 ACCEPTED MANUSCRIPT

The edges of the growth substrate were also found to contain a high density of very thin
nanowires that were pure In. The most likely cause for this is that during the temperature
ramp-up phase of growth, after the more volatile Sb has evaporated from the source, excess
liquid In nucleates on the growth substrate and as growth proceeds, the continuous supply
of In from the evaporating source results in the synthesis of In nanowires. The substrate

PT
also showed evidence of growth of multiple nanowires from a single nucleation site in a
‘flower-shaped configuration’ as shown in Figure 6(a). The nanowires were comprised of In,

RI
(Figure 6(b)), while the island from which it grew was Sb (Figure 6(c)). This suggests that
though the initial nucleation event did not result in 1-D growth but rather in an island of

SC
Sb. A continuous supply of In from the evaporating source then resulted in formation of In
nanowires. It was also verified that nanowires grown on the same substrate under identical

U
conditions showed different composition; a direct verification of the fact that Sb dissolution
in the alloy droplet is extremely critical for InSb nucleation and InSb nanowire growth.
AN
IV. CHARACTERIZATION OF InSb NANOWIRES
M

To assess the optical quality of the InSb nanowires, they were characterized by Raman
D

spectroscopy. Room temperature Raman spectrum obtained from In and InSb nanowires
TE

(Figure 7) (a), (b) shows distinct peaks associated with pure In and InSb respectively.
The Raman spectrum for In nanowires (Figure 7)(a) shows a peak at 118 cm−1 , which
has been observed and reported on previously and is attributed to In21 .The TO and LO
EP

phonon modes of InSb at 179 cm−1 and at 187 cm−1 (Figure 7)(b) with narrow full width
at half maximum (FWHM) attests to good crystalline quality of the InSb nanowires39 . In
C

comparison with bulk InSb, the Raman spectrum for InSb nanowires show some striking
difference. The prominent additional peak at 150 cm−1 is attributed to TO-TA modes40 ;
AC

this has been reported in other works and are attributed to the presence of intrinsic defects
in the nanowire41 . These defects are most likely related to a higher concentration of In in
the nanowires which results in Sb vacancies.
To ascertain the electronic quality of these nanowires, electron transport measurements
were performed on individual In and InSb nanowires contacted by Cr/Au electrodes with
a back-gate are shown in Figure 8(a) and (b) respectively. Figure 8 (a) shows the back-gate
measurement results on a single 80 nm diameter In nanowire; inset shows the linear Ids - Vds

10
 ACCEPTED MANUSCRIPT

plot, evidence that the Cr/Au provides ohmic contact to the metallic nanowire. It is clear
that there is no gating effect on the metallic nanowires; an expected result attributed to
high electron concentration in the In nanowires. Figure 8(b) shows transport measurement
results obtained from a InSb nanowire. The current Ids flowing through the InSb nanowire is
about two orders of magnitude lower than that through the metallic In nanowire. The back-

PT
gate effect on Ids at Vds = 0.7 V was analyzed to realize that the electron concentration in the
InSb nanowires is of the order of 1017 cm−3 , about an order of magnitude higher than the

RI
intrinsic carrier concentration of 1016 cm−3 . The higher electron concentration is attributed
to the presence of intrinsic defects either on the surface or inside the nanowire.

SC
V. CONCLUSION

U
We can thus conclude that there are several energetic and kinematic factors that govern
AN
the VLS nanowire growth. The evaporation of solid InSb source to achieve antimonidiza-
tion of In droplets will not result in stoichiometric InSb nanowires unless several critical
M
parameters are well controlled. Unless these conditions are met, the synthesized nanowires
is mostly comprised of metallic In since the nanowire growth commenced past the stage of
D

in-congruent sublimation where the InSb source is almost entirely made of the less volatile
group-III element. The modified composition of the InSb source material also results in
TE

un-reproducible experimental results and maintaining stoichiometry of InSb nanowires be-

comes a big challenge. The influence of the Sb vapor environment, the surface energies
EP

and chemical potential of the droplet are all critical parameters for growth of stoichiometric
InSb nanowires. During the initial growth phase, the varying sizes of the molten In droplets
C

results in nanowires with composition varying from pure metal In to semiconducting InSb.
A single substrate can exhibit variation in nanowire composition. The optical and electronic
AC

quality of the In and InSb nanowires show that they have good crystal quality.

11
 ACCEPTED MANUSCRIPT

1 Yang Y, Li L, Huang X, Li G, and Zhang L 2007 Fabrication and optical property of single-

PT
crystalline InSb nanowire arrays J. Mater. Sci.42 2753-2757
2 Vaddiraju S, Sunkara M K, Chin A H , Ning C Z, Dholakia G R, and Meyyappan M 2007
Synthesis of Group III Antimonide Nanowires J. Phys. Chem. C.111 7339-7347

RI
3 Plissard S R , Slapak D R , Verheijen M A, Hocevar M, Immink G W G , Weperen I Van,

SC
Nadj-Perge S, Frolov S M, Kouwenhoven L P and Bakkers E P A M 2012 From InSb Nanowires
to Nanocubes: Looking for the Sweet Spot Nano Lett. 12 1794-1798
4 Yang X, Wang G, Slattery P, Zhang J Z and Li Y 2010 Ultrasmall Single-Crystal Indium

U
Antimonide Nanowires Cryst. Gr. and Design. 10 2479-2482
5 AN
Khan M I, Wang X, Bozhilov K N and Ozkan C S 2008 Templated Fabrication of InSb Nanowires
for Nanoelectronics J. Nanomater. 2008 698759
M
6 Kuo C-H, Wu J-M and Lin S-J 2013 Room temperature-synthesized vertically aligned InSb
nanowires: electrical transport and field emission characteristics Nanoscale Res. Lett.8 69
D

7 Ghasemi M, Zanolli Z, Stankovski M and Johansson J 2015 Size- and shape-dependent phase
diagram of InSb nano-alloys Nanoscale 7 17387-17396
TE

8 Kuo C-H, Wu J-M, Lin S-J and Chang W-C 2013 High sensitivity of middle-wavelength infrared
photodetectors based on an individual InSb nanowire Nanoscale Res. Lett. 8 327
EP

9 Vogel A T, de Boor J, Wittemann J V, Mensah S L, Werner P and Schmidt V 2011 Fabrication

of High-Quality InSb Nanowire Arrays by Chemical Beam Epitaxy Crys. Growth Des. 11 1896-
C

1900
10 Park H D, Prokes S M, Twigg M E, Ding Y and Wang Z L 2007 Growth of high quality, epitaxial
AC

InSb nanowires J.Cryst. Growth. 304 399-401

11 Nilsson H A, Samuelsson P, Caroff P and Xu H Q 2012 Supercurrent and Multiple Andreev
Reflections in an InSb Nanowire Josephson Junction Nano Lett. 12 228-233
12 So H, Pan D, Li L and Zhao J 2017 Foreign-catalyst-free growth of InAs/InSb axial heterostruc-
ture nanowires on Si (111) by molecular-beam epitaxy Nanotech. 28 135704
13 Singh A, Algarni Z and Philipose U 2017 Template-Assisted Electrochemical Synthesis of p-
Type InSb Nanowires ECS J. Solid State Sci. Technol. 6 N39-N43

12
 ACCEPTED MANUSCRIPT

14 Gul O, Zhang H, de Vries F K, van Veen J, Zuo K, Mourik V, Conesa-Boj S, Nowak M P, van
Woerkom D J, Quintero-Perez M, Cassidy M C, Geresdi A, Koelling S, Car D, Plissard S R,
Bakkers E P A M and Kouwenhoven L P 2017 Hard Superconducting Gap in InSb Nanowires
Nano Lett. 17 2690-2696
15 Li S , Huang G Y, Guo J K , Kang N, Caroff P and Xu H Q 2017 Ballistic transport and

PT
quantum interference in InSb nanowire devices Chin. Phys. B. 26 027305
16 Li S, Kang N, Caroff P and Xu H Q 2017 0-π phase transition in hybrid superconductor-InSb

RI
nanowires quantum dot devices Phys. Rev. B. 95 014515
17 Shafa M, Akbar S, Gao L, Fakhar-e-Alam M and Wang Z M 2016 Indium Antimonide Nanowires:

SC
Synthesis and Properties Nanoscale Research Letters 11 164
18 Wagner R S and Ellis W C 1964 Vapor-liquid-solid mechanism of single crystal growth

U
App.Phys.Lett 4 89-90
19
AN
Wang Y, Chi J, Banerjee K, Grutzmacher D, Schapers T and G. Lu J 2011 Field effect transistor
based on single crystalline InSb nanowire J. Mater. Chem. 21 2459-2462
20 Algarni Z, George D, Singh A, Lin Y and Philipose U 2016 Hole-dominated transport in InSb
M
nanowires grown on high-quality InSb films J. Nanopart. Res. 18 361
21 Sapkota G and Philipose U 2014 Synthesis of metallic, semiconducting, and semi-metallic
D

nanowires through control of InSb growth parameters Semicond. Sci. Tech. 29 035001
TE

22 Laura Ye Q, Yamada T, Liu H, Scheffler R, Mingo N and Leverenz R 2006 Single Crystal InSb
Nanowires: Synthesis, Characterization, Properties and Applications Mater. Res. Soc. Symp.
EP

Proc. 940
23 Tsao J Y 1991 Phase equilibria during InSb molecular beam epitaxy J.Cryst. Growth 110
C

595-603
24 Yano M, Yokose H, Iwai Y and Inoue M 1991 Surface reaction of III-V compound semiconductors
AC

irradiated by As and Sb molecular beams J. Cryst. Growth 111 609-613

25 Wang Y Q, Wang Z L, Brown T, Brown A and May G 2002 Thermodynamic analysis of anion
exchange during heteroepitaxy J. Cryst. Growth 242 5-14
26 Yata M 1986 Growth kinetics of InSb thin films on Si(100) surfaces by In1 and Sb4 molecular
beams Thin Sol. Films 137 79-87
27 Ek M, Borg B M, Johansson J and Dick K A 2013 Diameter Limitation in Growth of III-Sb-
Containing Nanowire Heterostructures ACS Nano 7 3668-3675

13
 ACCEPTED MANUSCRIPT

28 Philipose U, Sapkota G, Salfi J and Ruda H E 2010 Influence of growth temperature on the
stoichiometry of InSb nanowires grown by vapor phase transport Semicond. Sci.Technol. 25
075004
29 Sharma R C, Ngai T L and Chang Y A 1989 The In-Sb (Indium-Antimony) system Bulletin of
Alloy Phase Diagrams. 10 657-664

PT
30 Crawford S, Lim S K and Gradecak S 2013 Fundamental Insights into Nanowire Diameter
Modulation and the Liquid/Solid Interface Nano Lett. 13 226-232

RI
31 Schmidt V, Wittemann J V, Senz S and Gosele U 2009 Silicon Nanowires: A Review on Aspects
of their Growth and their Electrical Properties Adv. Mater. 21 2681-2702.

SC
32 Dubrovskii V G, Cirlin G E, Sibirev N V, Jabeen F, Harmand J C and Werner P 2011 New
Mode of Vapor-Liquid-Solid Nanowire Growth. Nano Lett. 11 1247-1253.

U
33 Wang H, Wang J-T, Cao Z-X, Zhang W-J, Lee C-S, Lee S-T and Zhang X-H 2015 A surface

AN
curvature oscillation model for vapour-liquid-solid growth of periodic one-dimensional nanos-
tructures Nat. Commun. 6 6412.
34 Dhalluin F, Desre P J, den Hertog M I, Rouviere J-L, Ferret P, Gentile P and Baron T 2007
M
Critical condition for growth of silicon nanowires J. Appl. Phys. 102 094906
35 Hall R N 1963 Solubility of III-V Compound Semiconductors in Column III Liquids J. Elec-
D

trochem. Soc. 110 385-388

TE

36 Givargizov E I 1975 Fundamental aspects of VLS growth J. Cryst. Growth 31 20-30.

37 Chen Z and Cao C 2006 Effect of size in nanowires grown by the vapor-liquid-solid mechanism
EP

Appl. Phys. Lett. 88 143118

38 Zi Y, Jung K, Zakharov D and Yang C 2013 Understanding Self-Aligned Planar Growth of InAs
C

Nanowires Nano Lett. 13 2786-2791.

39 Feng Z C, Perkowitz S, Rao T S and Webb J B 1990 Resonance Raman scattering from epitaxial
AC

InSb films grown by metalorganic magnetron sputtering J. Appl. Phys. 68 5363-5365

40 Horiuchi T and Wada N 2007 AntiStokes Photoluminescence and Enhanced Raman Scattering
from InSb Quantum Dots and Nanowires. AIP Conference Proceedings 893 927-928.
41 Das S R, Akatay C, Mohammad A, Khan M R, Maeda K, Deacon R S, Ishibashi K, Chen Y
P, Sands T D, Alam M A and Janes D B 2014 Electrodeposition of InSb branched nanowires:
Controlled growth with structurally tailored properties J. Appl. Phys. 116 083506

14
 ACCEPTED MANUSCRIPT

(a ) (b )

PT
RI
(c ) (d )

U SC
AN
M

FIG. 1. Schematic of the VLS growth mechanism as applied to the InSb nanowires involving the
D

balance of the surface tensions at different growth stages. (a) The alloy droplet in the early stage
TE

of growth, showing the contact angle (θ ) is less than π/2 and the inclination angle (α) is equal to
zero. Inset shows an SEM image of alloyed droplet under this condition. (b) The saturated alloyed
EP

droplet shows an increase in (θ) and (α). (c) Late stage of the growth showing a net force directed
upward. Inset shows an SEM image of a single InSb nanowire with the alloyed droplet under
equilibrium condition of surface tensions. (d) Net forces directed downward causing the nanowire
C

to penetrate into the droplet. Inset shows an SEM image of the nanostructures -resembling cones
AC

grown under this condition.

15
 ACCEPTED MANUSCRIPT

7 0
m l

m
v -l
6 0 l
l-s

PT
5 0
( k J /m o le )

4 0

RI
3 0

SC
l

2 0
m

U
1 0

0
AN
q
0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0
(d e g r e e )
M
D

FIG. 2. Variation of the chemical potential of liquid phase µl with contact angle θ during the
vapor to liquid and liquid to solid phase transitions. The dependence was obtained using : µlv−l =
TE

300
1000[12 + r sinθ] and µll−s = 1000[35 + 30cosθ].
C EP
AC

16
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D

FIG. 3. SEM image of In droplets on Si substrate during the initial growth phase. The droplet
diameters range from 50-200 nm.
TE
C EP
AC

17
 ACCEPTED MANUSCRIPT

(a ) (b )

PT
RI
U SC
AN
M

FIG. 4. Composition variation in nanowires with different radii. (a) EDX spectrum of pure In
D

nanowire. Inset is an SEM image of a single In nanowire with radius of 120 nm. (b) EDX spectrum
TE

of a stoichiometric InSb nanowire. Inset is an SEM image of a single InSb nanowire with radius
of 60 nm.
C EP
AC

18
 ACCEPTED MANUSCRIPT

In
1 0 0
S b

PT
W e ig h t p e r c e n ta g e (% )

8 0

RI
6 0

SC
4 0

U
2 0

0
AN
6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0
M
D ia m e te r (n m )
D

FIG. 5. The diameter- composition plot of nanowires grown at 530◦ C . Plot shows that for a given
growth temperature there is a range of diameters for which the In:Sb ratio is close to 1. Outside
TE

this range for d <100 nm and d>130 nm, the In content increases.
C EP
AC

19
 ACCEPTED MANUSCRIPT

(b )

(a )

PT
RI
(c )

U SC
AN
M

FIG. 6. (a) SEM image of nanowires grown from a single nucleation site in a flower shaped
D

configuration. (b) EDX spectrum collected from a selected region on the nanowires shown in (a).
TE

The nanowires are mostly comprised of In. (c) EDX spectrum collected from the bare of the island
showing the composition to be Sb.
C EP
AC

20
 ACCEPTED MANUSCRIPT

PT
(a ) -1
(b ) T O - T A :1 5 0 c m
-1
1 1 8 c m

RI
-1
T O :1 7 9 c m

In te n s ity (a r b .u n its )
In te n s ity (a r b .u n its )

-1

SC
L O :1 8 7 c m

U
1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 2 4 0 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 2 4 0
-1 -1
R a m a n S h ift (c m )
AN R a m a n S h ift (c m )
M

FIG. 7. Raman spectrum for In and InSb nanowires, measured at room temperature. (a) Raman
D

spectrum for In nanowires which shows a broad Raman peak at 118 cm−1 , attributed to pure
TE

metallic In. (b) The spectrum for InSb nanowires shows two Raman active modes attributed to
TO and LO modes of phonon vibration in InSb. The additional peak at 150 cm−1 is attributed
EP

to TO-TA modes and is believed to originate from defects in the nanowire.

C
AC

21
 ACCEPTED MANUSCRIPT

8 0
(b )

PT
3 .4 0 5 4 V = 7 0 0 m V
(a ) 1 0 0 V = 5 0 m V 6 0 d s

Id s (n A )
d s
)

4 0
m A

5 0
3 .3 5
(

Id s(n A )
2 0 5 2

Id s(m A )
Id s

0 0

I
-1 .0 -0 .5 0 .0 0 .5 1 .0 3 .3 0 -1 .0 -0 .5
-2 0
0 .0 0 .5 1 .0
V d s (V o lts ) V d s (V o lts )
-5 0 5 0

RI
-4 0
3 .2 5 -6 0
-1 0 0
-8 0
3 .2 0 4 8

3 .1 5 4 6

SC
3 .1 0
4 4
3 .0 5

3 .0 0 4 2

U
-1 0 -5 0 5 1 0 -1 5 -1 0 -5 0 5 1 0 1 5
V g s
(V o lts ) V (V o lts )
g s

AN
M
FIG. 8. Electron transport measurements made on In and InSb nanowires contacted by Cr/Au
electrodes. (a) Back-gate measurement made on of a single metallic In nanowire shows no gate
D

effect. Inset shows the linear variation of Ids with Vds . (b) Gate measurments made on a single
InSb nanowire at Vds = 0.7 V. The increase in Ids with Vgs shows n-type behavior of InSb nanowire.
TE

Inset shows Ids vs Vds for an InSb nanowire at zero gate bias. As seen in the plot Ids for InSb is
much lower than that for an In nanowire.
C EP
AC

22