hydrothermal growth of zno nanorods the role of kcl in controlling

Hydrothermal growth of ZnO nanorods: The role of KCl in controlling
rod morphology
Jonathan M. Downing
a
, Mary P. Ryan
b
, Martyn A. McLachlan
a,
⁎
a
Department of Materials & Centre for Plastic Electronics, Imperial College London, London SW7 2AZ, UK
b
Department of Materials & London Centre for Nanotechnology, Imperial College London, London SW7 2AZ, UK
abstractarticle info
Article history:
Received 17 August 2012
Received in revised form 19 April 2013
Accepted 22 April 2013
Available online 9 May 2013
Keywords:
Zinc oxide
Hydrothermal
Nanorod
Ionic additive
Alignment
Self-assembly
The role of potassium chloride (KCl) in controlling ZnO nanorod morphology of large area thin films prepared by
hydrothermal growth has been extensively investigated. The influence of KCl and growth time on the orienta-
tion, morphology and microstructure of the nanorod arrays has been studied with systematic changes in the
length, width, density and termination of the nanorods observed. Such changes are attributed to stabilization
of the high-energy (002) nanorod surface by the KCl. At low KCl concentrations (b 100 mM) c-axis growth

i.e. perpendicular to the polar surface, dominates, leading to nanorods with increased length over the control
sample (0 mM KCl). At higher concentrations (>100 mM) stabilization of the high-energy surface by KCl occurs
and planar (002) facets are observed accompanied by increased lateral (100) growth, at the highest KCl concen-
trations near coalesced (002) terminated rods are observed. Additionally we correlate the KCl concentration with
the uniformity of the nanorod arrays; a decrease in polydispersity with increased KCl concentration is observed.
The vertical alignment of nanorod arrays was studied using X-ray diffraction, it was found that this parameter
increases as growth time and KCl concentration are increased. We propose that the increase in vertical alignment
is a result of nanorod–nanorod interactions during the early stages of growth.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
There has been a significant volume of published work relating to
the growth of ZnO nanostructures over the past decade. The unique
electronic and optical properties of this material combined with the
library of 3-dimensional structures prepared, including; dendrites [1],
tetrapods [2],helixes[3], in addition to lower-dimensional structures
such as particles [4],nanorods[5] and plate lets [6] have fuelled contin-
ued research. Of the many available structures ZnO nanorods have been
proposed for a wide range of applications, including; photovoltaics [7],
light emitting diodes [8], piezoelectric generators [9] and chemical sen-
sors [10]. In addition to structural variations, ZnO continues to attract
interest owing to its outstanding electronic properties, including wide
and direct band gap ~3.3 eV, reported carrier concentrations of up to
2×10
16
cm
−3
, electron mo bilities up t o 205 cm
2
Vs
−1

,highoptical
transparency and a large exciton binding energy at room temperature
(60 meV) [11]. Further enhancement of many of these physical proper-
ties may be possible through intrinsic and extrinsic doping.
In many nanostructures the large surface-to-volume ratio, prefer-
ential growth of polar/non-polar surfaces and anisotropic charge trans-
port have been explored as means of improving device performance
[7,12,13]. For e xampl e, Z nO nanorods hav e bee n used to form pho tovol-
taic devices in which the power conversion efficiency was improved
by almost an order of magnitude compared with comparable devices
fabricated with planar ZnO layers [7,14]. Furthermore, improvements
in the turn-on field values of electroluminescent devices by a factor of
eight have been reported through control of nanorod morphology
[15]. Future improvements in emerging devices are likely if methods
for the reproducible fabrication of nanorod arrays can be achieved.
Using high temperature, vacuum techniques such as Pulsed Laser
Deposition [16], Vapor Liquid Solid Growth [17] and Chemical Vapor
Deposition [18] high quality nanorod arrays have been fabricated. In
comparison solution processing routes are more attractive owing to
their low cost, low deposition temperatures and compatibility with
large area deposition. However, morphological control and batch-to-
batch reproducibility continue to be problematic with such processes.
To date, the most widely reported solution processes for nanorod
growth are electrochemical deposition [19,20] and hydrothermal
growth [21]. The latter is particularly well suited to large area deposi-
tion, requires little capital investment of equipment and can be
achieved on electrically conducting and insulating substrates, includ-
ing flexible polymeric substrates [21].
For the growth of nanorod arrays, the hydrothermal technique
is outlined in detail by Vayssieres et al. [22], where nanorod growth

is promoted when a water-soluble zinc precursor, often zinc nitrate
(Eqs. (1)–(2)), is mixed with a suitable amine. At elevated tempera-
tures (b 100 °C) thermal decomposition of the amine (Eq. (3)) results
Thin Solid Films 539 (2013) 18–22
⁎ Corresponding author at: Department of Materials & Centre for Plastic Electronics,
Imperial College London, Royal School of Mines, South Kensington, London SW7 2AZ,
UK. Tel.: +44 20 7594 9692.
E-mail address: (M.A. McLachlan).
0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
/>Contents lists available at SciVerse ScienceDirect
Thin Solid Films
journal homepage: www.elsevier.com/locate/tsf
in a pH increase of the solution, producing thermodynamically unsta-
ble zinc hydroxide, which spontaneously precipitates as ZnO (Eq. (4)).
ZnðNO
3
Þ
2
→Zn
2þ
þ 2NO
À
3
ð1Þ
NO
À
3
þ H
2
O þ 2e

À
→NO
À
2
þ 2OH
À
ð2Þ
C
6
H
12
N
4
þ 10H
2
O⇌6CH
2
O þ 4NH
þ
4
þ 4OH
À
ð3Þ
Zn
2þ
þ 2OH
À
→ZnðOHÞ
2
→ZnO þ H

2
O ð4Þ
The formation of nanorods during hydrothermal growth occurs due
to the differing surface energies of the polar (002) and non-polar (100)
surfaces in wurtzite ZnO. Minimization of the high-energy polar surface
results in increased b002> growth and the subsequent formation of
nanorods [23].
The hydrothermal process is well-studied and the influence of
many growth parameters on nanorod morphology have been reported,
e.g. reducing [Zn
2+
]
(aq)
reduces rod diameter [21], increasing tempera-
ture increases growth rate and improves crystallinity [24].Alternative
Zn precursors and amines have been used to alter rod morphology
[25,26] and seed layers have been used to control areal density [26,27]
and vertical alignment [28,29].
Whilst the addition of chemical additives has been well studied as a
means of altering rod morphology i.e. the incorporation of surfactants to
minimize b100> growth through adsorption on the non-polar surfaces
[30,31]. There are few reports of the role of ionic additives in solution;
in alkali deposition baths, metal sulfates are reported to control rod
aspect ratios [15], while increasing Cd
2+
/Zn
2+
ratio in solution results
in the formation of bipyramidal structures [32]. In reactions using
hexamethylene triamine (HMT), citrate ions adsorb to (002) surface

with the effect to produce stacked plate like crystals [33].
We have recently reported on t he incorporation of polyethy leneimine
(PEI) and potassium chloride (KCl) in the hydrothermal growth of ZnO
and t heir influence on nanorod m orpholog y and vertical o rientation
[34].Herewereportasignificant advance of this preliminary work in
which exceptional control o f nanorod morphology and vertical alignment
are achi eved. The role o f KCl in co ntrolling nanorod morphology and its
influence on the grow th process i s described in deta il. The n anorod arrays
here are proposed as highly suitable and tunable active layers in emerg-
ing optoelectron ic devices, for example homojunctions in hybrid solar
cells and as electron injecting layers in organic light emitting diodes.
2. Experimental details
2.1. ZnO nanorod preparation
A two-stage deposition process for nanorod growth was adapted
from existing literature [21,24];
Seed lay er d eposition A 0.75 M solution o f z inc acetate dihydrate (Zn (O
2
CCH
3
)
2
.2(H
2
O)) and 2-am inoethanol (H
2
N(CH
2
)
2
OH) in 2-methoxyethano l (HO(CH

2
)
2
OCH
3
)) [35]
was deposited on to tin-doped i ndium oxide coat-
ed glass s ubstrates (PsiOteC UK Ltd, 12–16 Ω/sq).
Dense films were created by loading at 500 rpm
before spinning at 2000 rpm for 30 s; three-coats
were applied with substrate heating (300 °C for
10 min) between coats. A final 60 min anneal at
450 °C was carr ied o ut, producing continuous
films of ~120 nm thic kness.
Nanorod gro wth 25 mM solutions of HMT (NH
2
(CH
2
)
6
NH(CH
2
)
6
NH
2
) and zinc nitrate ((Zn(NO
3
)
2

) were mixed in
a closed vessel immersed in a controlled tempera
ture water bath 95°C±1°C. The substrate
supported seed layers were suspended directly into
the hydrothermal solution. The additives, 0–500 mM
KCl and 10 mM PEI (H(NHCH
2
CH
2
)
n
NH
2
)were
added to the solution imme diate ly pr ior to seed
layer immersio n. Followin g rod depo sition the films
were rinsed thoroughly with deionized water and
allowed to dry at 95 °C.
The morphologies of the films were characterized using a LEO 1525
field emission scanning electron microscope. Surface images were
obtained on the as-prepared films whilst cross-sectional images were
obtained after the scratching with a diamond scribe (imaging typically
2–10 kV). Image analysis was carried out from the micrographs by
counting (areal density) and with use of ImageJ software to collate
rod length and width data. The crystal structures of the films were ana-
lyzed using Panalytical X'Pert MPD diffractometer equipped with an
Accelerator detector, operated at 40 kV/40 mA (Cu K
α
source, theta-
theta configu ratio n). X-ray diffraction (XRD) patterns were corrected

for K
α2
emission and adjusted for background in X'Pert Highscore Plus,
integrated peak intensity was found by peak fitting in X'Pert DataViewer.
3. Results
3.1. Role of KCl: film structure and rod morphology
Nanorod growth was carried out using equimolar concentrations
(25 mM) of Zn(NO
3
)
2
and HMT and 10 mM PEI, whilst the KCl con-
centration was systematically varied (0 – 500 mM). The influence of
growth time on rod length, Fig. 1, shows a linear increase in rod
length for growth times of 25 - 120 min. Further extensions of the
growth time do not lead to a significant increase in average rod length
owing to consumption of the reactants. Linear growth is a result of
the steady decomposition of HMT in solution, which acts as a buffer
to provide hydroxide ions that are subsequently consumed (Eq. (4))
to produce ZnO [36].
In the present study growth times of 40 and 120 min at each KCl
concentration were investigated in order to study growth early and
late in the linear growth regime. Fig. 2 shows cross-sectional and sur-
face SEM images of ZnO nanorod structures grown for 120 min
(40 min, not shown) over the 0–500 mM KCl concentration range.
Here KCl is seen to be influencing nanorod length, vertical alignment,
aspect ratio and tip termination. The introduction of KCl (10 mM) re-
sults in a significant increase in rod length (860 ± 250 nm) compared
with growth in the absence of KCl (370 ± 130 nm). Increasing the con-
centration of KCl to 50 mM results in a further increase in rod length

(1040 ± 180 nm), at higher KCl concentrations a gradual reduction
in nanorod length is observed (Fig. 2d-f). At KCl concentrations
(b 200 mM) the nanorods are terminated with sharp points, above
this concentration the rods are terminated by flat surfaces and show
clear hexagonal faceting.
Fig. 1. Showing measured average nanorod length plotted against growth time for
hydrothermal baths containing equimolar (25 mM) Zn(NO
3
)
2
/HMT in addition to
10 mM PEI and 100 mM KCl.
19J.M. Downing et al. / Thin Solid Films 539 (2013) 18–22
To further quantify the influence of KCl on nanorod growth detailed
analysis of films grown under each set of growth conditions was carried
out using SEM images. The calculated averages are accompanied by the
standard error of the mean, it should be noted that such errors are small
due to the large measured sample size. A minimum of two samples at
each KCl concentration were analyzed at a number of locations on
each substrate. In summary, and consistent with the trends shown in
Fig. 2, nanorod length increases on addition of KCl, reaching a maximum
at 50 mM (1040 nm (13 nm)); above this concentration nanorod
length is gradually reduced. The measured rod diameter increases
across the concentration series to a maximum of 90 nm (2 nm) at
500 mM KCl. The areal density is reduced on addition of small amounts
of KCl but increases with increasing KCl concentration. Fig. 3 shows the
processed results obtained from the image analysis for the calculated
deposition volume, and the measured rod length, diameter and areal
density as KCl concentration is varied.
Films grown over the entire KCl concentration range for both

growth times were characterized using XRD, the diffraction patterns
are shown in Fig. 4a, the integrated intensity of the (002) diffraction
peak (~34.4 °2θ) has been used to quantify rod alignment. Diffraction
intensity for a given peak is linearly dependent on the quantity (sample
volume) and orientation of nanorods on the substrate, i.e. for a given
volume (002) diffraction intensity is increased if nanorods are aligned.
To deconvolute the orientation and volume effects the integrated
(002) peak areas are corrected for the sample volume. Finally to quan-
tify rod alignment, data were normalized to the (002) peak intensity of
the 0 mM KCl sample, providing a direct comparison of KCl addition on
vertical alignment, Fig. 4b. Samples grown for 40 min show little varia-
tion in vertical alignment as the KCl concentration is varied. In contrast
those grown for 120 min all show improvements in alignment that
reaches a maximum at around 400 mM KCl.
4. Discussion
A simple method for controlling ZnO nanorod morphology through
addition of KCl to the hydrothermal growth bath is presented. Our pro-
posed mechanisms to explain the observed changes in morphology as a
function of KCl concentration are outlined below.
4.1. Solution chemistry
It is first necessary to consider the free energy of hydration for each
of the ionic species in the hydrothermal bath i.e. NH
4
+
,K
+
,NO
3
-
,NO

2
-
,Cl
-
,
OH
-
. Typically these are between − 285 and − 430 kJ mol
−1
[37],indi-
cating that these species will be heavily solvated during nanorod
growth hence reactions between these species in solution are unlikely.
Furthermore all ionic species are in low concentrations, the maximum
([KCl] 500 mM), is significantly lower than the solubility limit at
100 °C (7.6 M).
Analysis of the SEM micrographs of short (40 min) growth time
samples yields some information about nanorod nucleation and early-
stage growth behavior. In our experiments similar areal densities
were measured without KCl and across the whole KCl concentration
series, on average 149 μm
−2
indicating that the nucleation density is
independent of KCl concentration.
Increasing the KCl concentration results in a marked reduction in
the standard deviation of rod length, showing improved uniformity
as KCl concentration is increased (Table 1).
In the presence of KCl the observed nanorod growth behavior can
be explained by considering nanorod surface termination (Fig. 2i-j).
In order to minimize the area of high energy (002) surfaces nanorod
growth proceeds primarily along the b002> direction with slow growth

in the b100> directions. Fig. 2 shows that at KCl concentrations
≤ 100 mM the nanorods are terminated by (101) faceted points (con-
sistent with ref. [23]), and as the KCl concentration increases a planar
(002) surface is stabilized. We propose that the observed increase in
nanorod length at 10 mM KCl i.e. rapid b002> growth, is attributed to
the partial stabilization of a small (002) surface at the tip of the growing
rod and not lateral (100) adsorption as previously speculated [25].The
proposed structure and morphology of the growing rods are shown
schematically in Fig. 5. In the absence of KCl, nanorods are terminated
by sharp points and the (002) surface is fully minimized (Fig. 5a). Addi-
tion of small amounts of KCl, results in the partial stabilization of the
(002) surface Fig. 5b, which promotes rapid b002> growth and explains
the observed increase in nanorod length and accompanying reduction
in diameter. Further increases in KCl concentration completely stabilize
the (002) surface resulting in a promotion of lateral growth resulting in
planar faceted rods, consistent with previous reports [34]. Under alka-
line conditions (pH 11) where it has been proposed that changes in
nanorod aspect ratio are attributed to cation adsorption to the lateral
(100) surfaces [15], our hydrothermal method growth occurs at a mea-
sured pH of 5.5 [38], under these conditions the non-polar (100) surface
is thought to remain neutral, with any charging of the (100) surface
being negligible in comparison to the effect of additive stabilization
to the (002) growth surface, growth rates on the relative planes are
shown schematically in Fig. 5.
Fig. 2. Cross-section (upper) and surface (lower) SEM images showing ZnO nanorods grown from hydrothermal baths containing equimolar (25 mM) Zn(NO
3
)
2
/HMT and 10 mM
PEI. The modification of nanorod morphology with the addition of a) 0, b) 10, c) 50, d) 100, e) 200, f) 500 mM KCl for a growth time of 120 min is shown, c) nanorod–nanorod

interaction is highlighted (all SEM scale markers 200 nm). TEM images highlighting the change in nanorod tip termination at i) 10 mM and j) 300 mM KCl.
20 J.M. Downing et al. / Thin Solid Films 539 (2013) 18–22
The longest rods are formed at 50 mM KCl but the greatest volume
of ZnO is deposited at 100 mM KCl—owing to the marked increase in
lateral growth. Control of rod termination may be advantageous in
some device applications e.g. photovoltaics, where a smaller (002)
surface may reduce the polar barrier for electron transfer [13].
4.2. Physical nanorod interaction
As the growth time is extended from 40 to 120 min there is a dis-
tinct reduction in areal density of nanorods (cf. Fig. 3d), indicating
that many of the nucleated nanorods observed early in the linear
growth regime do not continue to grow. We propose that this phe-
nomenon results from the physical interaction occurring when adja-
cent nanorods grow at angles whereby their growth paths intersect,
Fig. 2c, resulting in the termination of some nanorods as growth
time increases. This process is supported by the observed increase
in (002) diffraction intensity at longer growth times (cf. 40 vs.
120 min, Fig. 4b). Very recent work shows the validation of this phe-
nomenon by application of the geometrical selection model [39],
where three distinct growth regimes are outlined, namely isolated,
competitive and aligned.
The variation in nanorod alignment between different KCl concen-
trations is ascribed to the differences in growth rate between lateral
and vertical directions. In regimes where c-axis growth is hindered
(>100 mM KCl), nanorod diameters are increased; hence rod-to-rod
interactions occur at an earlier growth stage. Under these conditions
(>100 mM KCl), misaligned nanorods are more likely to interact early
in the growth process and terminate, i.e. growth in the competitive re-
gime is reduced, resulting in the highly orientated films at lower time
periods. The XRD data support this hypothesis; diffraction intensity

per unit volume is at a maximum at 400 mM KCl for 120-min growth,
showing increased alignment despite the reduced nanorod length
(315 nm) of these films.
Table 1
Measured nanorod length and calculated statistical information for samples grown
over the KCl concentration range 0 – 500 mM.
KCl concentration
(mM)
Length
(nm)
Standard
deviation (nm)
Sample
size
Standard
deviation/Length
0 370 130 70 0.35
10 860 250 150 0.30
50 1040 180 191 0.18
100 960 120 106 0.12
200 540 80 147 0.15
300 390 40 134 0.10
400 310 30 97 0.11
500 310 20 28 0.08
Fig. 4. XRD patterns and calculated diffraction data for ZnO nanorods films grown with
varying KCl concentrations, a) volume corrected and normalized ZnO (002) peaks for
growth times of 40 (left) and 120 min (right), b) integrated (002) peak intensity
corrected for volume and normalized to 0 mM KCl sample where values >1 indicate
increased nanorod vertical alignment.
Fig. 3. Morphological data calculated from SEM micrographs, showing a) volume,

b) length, c) diameter and d) areal density of ZnO nanorods grown for 120 min with
varying concentrations of KCl (error bars show standard error).
21J.M. Downing et al. / Thin Solid Films 539 (2013) 18–22
5. Conclusions
A reproducible method for preparing tailored ZnO nanorods from
aqueous solution has been developed through incorporation of KCl into
the hydrothermal growth bath. KCl acts as a growth modifier through
stabilization of the polar (002) nanorod surfaces. The range of KCl con-
centrations inve stigat ed (0–500 mM) spans growth modes where par-
tial adsorption results in the formation of high aspect ratio nanorods
and complete adsorption results in the formation of near-coalesced
rods. In comparison to electrochemical growth, where 60 mM KCl is
reported to be sufficient to stabilize the (002) surface and change the
morphology from rods to platelets [40], platele t d e po si tio n has not
been reported by the hydrothermal method, furthermore in our work
only rod-like structures were prepared. At high KCl concentrations
(>100 mM), the growth of short er and wider Z nO nanorods is o bserved,
consistent with reduced b 002> growth due to stabilizat ion of the (0 02)
surfaces by KCl.
The incorporation of simple ionic additive into the hydrothermal
growth bath provides a convenient method affording control of the
nanorod dimensions, areal density and surface termination. This rep-
resents a significant step in the controlled and reproducible low-cost
solution processing of tailored nanostructures and should facilitate
the uptake of these structures into relevant device architectures.
Acknowledgments
The authors acknowledge useful discussions withDr.JosephFranklin,
Imperial College. JMD is supported by the EPSRC, EP/J016039/1, and in
part by the Energy Futures Lab Imperial College Lo ndon. M AM is grateful
for the support of a Royal Academy of Engineering/EPSRC Research

Fellowship that supported him during this work.
References
[1] G.R. Li, X.H. Lu, D.L. Qu, C.Z. Yao, F. Zheng, Q. Bu, C. Dawa, Y. Tong, J. Phys. Chem. C
111 (18) (2007) 6678.
[2] Z. Chen, Z.W. Shan, M.S. Cao, L. Lu, S.X. Mao, Nanotechnology 15 (3) (2004) 365.
[3] X. Kong, Z. Wang, Nano Lett. 3 (12) (2003) 1625.
[4] S.D. Oosterhout, M.M. Wienk, S.S. Van Bavel, R. Thiedmann, L. Jan Anton Koster,
J.Gilot,J.Loos,V.Schmidt,R.A.J.Janssen, Nat. Mater. 8 (10) (2009) 818.
[5] S. Peulon, D. Lincot, J. Electrochem. Soc. 145 (3) (1998) 864.
[6] B.N. Illy, B. Shollock, J. MacManus-Driscoll, M.P. Ryan, Nanotechnology 16 (2) (2005)
320.
[7] D.C. Olson, S.E. Shaheen, Collings, D.S. Ginley, J. Phys. Chem. C 111 (2007) 16670.
[8] R. Könenkamp, R.C. Word, C. Schlegel, Appl. Phys. Lett. 85 (24) (2004) 6004.
[9] Z.L. Wang, J. Song, Science 312 (5771) (2006) 242.
[10] J. Zhang, S. Wang, M. Xu, Y. Wang, B. Zhu, S. Zhang, W. Huang, S. Wu, Cryst.
Growth Des. 9 (8) (2009) 3532.
[11] D. Look, D. Reynolds, J. Sizelove, R. Jones, C. Litton, G. Cantwell, W. Harsch, Solid
State Commun. 105 (6) (1998) 399.
[12] X. Jiaqiang, C. Yuping, C. Daoyong, S. Jianian, Sens. Actuators B 113 (1) (2006) 526.
[13] S. Schumann, R. Da Campo, B. Illy, A.C. Cruickshank, M.A. McLachlan, M.P. Ryan,
D.J. Riley, D.W. McComb, T.S. Jones, J. Mater. Chem. 21 (7) (2011) 2381.
[14] D. Olson, Y J. Lee, M.S. White, N. Kopidakis, S.E. Shaheen, D.S. Ginley, J.A. Voigt,
J.W.P. Hsu, J. Phys. Chem. C 112 (26) (2008) 9544.
[15] J. Joo, B.Y. Chow, M. Prakash, E.S. Boyden, J.M. Jacobson, Nat. Mater. 10 (8) (2011)
596.
[16] Y. Sun, G. Fuge, M. Ashfold, Chem. Phys. Lett. 396 (1–3) (2004) 21.
[17] M.H. Huang, Y.Y. Wu, H. Feick, N. Tran, E. Weber, P.D. Yang, Adv. Mater. 13 (2)
(2001) 113.
[18] W.I. Park, G.C. Yi, M. Kim, S.J. Pennycook, Adv. Mater. 14 (24) (2002) 1841.
[19] H.E. Belghiti, T. Pauporte, D. Lincot, Phys. Status Solidi (a) 205 (10) (2008) 2360.

[20] T. Pauporte, G. Bataille, L. Joulaud, F.J. Vermersch, J. Phys. Chem. C 114 (1) (2009)
194.
[21] L. Vayssieres, Adv. Mater. 15 (5) (2003) 464.
[22] Vayssieres, K. Keis, S. Lindquist, A. Hagfeldt, J. Phys. Chem. B 105 (17) (2001)
3350.
[23] W J. Li, E W. Shi, W Z. Zhong, Z W. Yin, J. Cryst. Growth 203 (1–2) (1999) 186.
[24] M. Guo, P. Diao, S. Cai, J. Solid State Chem. 178 (6) (2005) 1864.
[25] K. Govender, D.S. Boyle, P.B. Kenway, P. O'Brien, J. Mater. Chem. 14 (2004) 2575.
[26] M. Kokotov, G. Hodes, J. Mater. Chem. 19 (23) (2009) 3847.
[27] Y. Lee, Y. Zhang, S.L.G. Ng, F.C. Kartawidjaja, J. Wang, J. Am. Ceram. Soc. 92 (2009)
1940.
[28] J.Y. Lee, T. Sounart, D. Scrymgeour, J. Voigt, J. Hsu, J. Cryst. Growth 304 (1) (2007)
80.
[29] D. Boyle, K. Govender, P. O'Brien, Chem. Commun. 1 (2002) 80.
[30] M. Law, L.E. Greene, J. Johnson, R. Saykally, P. Yang, Nat. Mater. 4 (6) (2005) 455.
[31] J J. Qiu, X. Li, F. Zhuge, X. Gan, X. Gao, W. He, S J. Park, H K. Kim, Y H. Hwang,
Nanotechnology 21 (19) (2010) 195602.
[32] R. Zhang, L.L. Kerr, J. Solid State Chem. 180 (3) (2007) 988.
[33] Z.R. Tian, J.A. Voigt, J. Liu, B. McKenzie, M.J. McDermott, M.A. Rodriguez, H.
Konishi, H. Xu, Nat. Mater. 2 (12) (2003) 821.
[34] J. Downing, M.P. Ryan, N. Stingelin, M.A. McLachlan, J. Photon. Energy 1 (2011)
011117.
[35] M. Ohyama, H. Kouzuka, T. Yoko, Thin Solid Films 306 (1) (1997) 78.
[36] K. Govender, J. Mater. Chem. 14 (2004) 2575.
[37] Y. Marcus, J. Chem. Soc. Faraday Trans. 87 (18) (1991) 2995.
[38] M.N.R. Ashfold, R.P. Doherty, N.G. Ndifor-Angwafor, D.J. Riley, Y. Sun, Thin Solid
Films 515 (24) (2007) 8679.
[39] T.Y. Olson, A.A. Chernov, B. Drabek, J.H. Satcher, T.Y J. Han, Chem. Mater. 25 (8)
(2013) 1363.
[40] L. Xu, Y. Guo, Q. Liao, J. Zhang, D. Xu, J. Phys. Chem. B 28 (109) (2005) 13519.

Fig. 5. Schematic illustrations showing the modification of nanorod morphology with
KCl addition, a) 0 mM, b) b 100 mM, c) >>100 mM.
22 J.M. Downing et al. / Thin Solid Films 539 (2013) 18–22