hollow, porous, and yttrium functionalized zno nanospheres with enhanced

Sensors

and

Actuators

B

178 (2013) 53–

62
Contents

lists

available

at

SciVerse

ScienceDirect
Sensors

and

Actuators

B:

Chemical

journa

l

h

o

mepage:

www.elsevier.com/locate/snb
Hollow,

porous,

and

yttrium

functionalized

ZnO

nanospheres

with

enhanced
gas-sensing

performances
Weiwei

Guo
a
,

Tianmo

Liu
a,∗
,

Rong

Sun
b
,

Yong

Chen
a,c
,

Wen

Zeng
a
,

Zhongchang

Wang
c,∗
a
College

of

Materials

Science

and

Engineering,

Chongqing

University,

Chongqing,

China
b
Institute

of

Engineering

Innovation,

The

University

of

Tokyo,

2-11-16

Yayoi,

Bunkyo-ku,

Tokyo

113-8656,

Japan
c
WPI

Research

Center,

Advanced

Institute

for

Materials

Research,

Tohoku

University,

2-1-1

Katahira,

Aoba-ku,

Sendai

980-8577,

Japan
a

r

t

i

c

l

e

i

n

f

o
Article

history:
Received

20

June

2012
Received

in

revised

form
18

December

2012
Accepted

20

December

2012
Available online 28 December 2012
Keywords:
ZnO
Nanospheres
Gas

sensor
Yttrium

doping
a

b

s

t

r

a

c

t
We

report

the

synthesis

of

a

hierarchical

nanostructure

of

hollow

and

porous

ZnO

nanospheres

with

a
high

specific

surface

area

as

a

novel

sensing

material

to

toxic

formaldehyde

by

a

simple

template-free
hydrothermal

technique

in

organic

solution.

We

demonstrate

that

the

liquid

mixture

ratio

and

hydro-
thermal

time

play

a

pivotal

role

in

forming

such

unique

morphology

and

propose

a

growth

mechanism
of

Ostwald

ripening

coupled

with

grain

rotation

induced

grain

coalescence.

Comparison

investigations
reveal

that

yttrium

allows

resistance

reduction

of

sensors

and

enhances

significantly

gas-sensing

per-
formances

of

ZnO

nanospheres

toward

the

formaldehyde

over

the

commonly

used

undecorated

ZnO
nanoparticles.

Such

hollow,

porous,

and

yttrium

functionalized

ZnO

nanospheres

could

therefore

serve
as

hybrid

functional

materials

for

chemical

gas

sensors.

The

results

represent

an

advance

of

hierarchical
nanostructures

in

enhancing

further

the

functionality

of

gas

sensors,

and

the

facile

method

presented
could

be

applicable

to

many

other

sensing

materials.
© 2012 Elsevier B.V. All rights reserved.
1.

Introduction
Inorganic

nanomaterials

with

hollow

and

porous

superstruc-
tures

find

numerous

technological

applications

where

morpholo-
gies

are

known

to

influence

functionality.

Gas

sensors

[1–3],
catalysts

[4,5],

drug

delivery

carriers

[6,7],

and

photoelectronic
building

blocks

[8–10]

are

just

a

few

significant

examples.

In

gen-
eral,

the

morphology

with

a

large

specific

surface

area

and

efficient
porosity

is

often

beneficial

for

the

catalytic,

gas-sensing

and

pho-

tovoltaic

applications

due

to

the

likelihood

to

enhance

surface
reactions.

In

this

respect,

the

active

search

of

unusual

morphology
is

currently

the

subject

of

intensive

research

in

the

nanomaterials
world

[11,12].

One

of

the

most

well-characterized

nanomaterials
in

terms

of

morphology

is

ZnO,

which

is

an

n-type

semiconductor
with

a

direct

wide

band

gap

(3.37

eV)

and

a

large

excitation

binding
energy

(60

meV)

[13,14].

To

date,

a

substantial

amount

of

exper-
iments

have

already

provided

definitive

evidence

that

size

and
morphology

of

ZnO

nanomaterials

can

affect

greatly

their

perform-
ances,

especially

gas-sensing

functionality

[15–17].

On

the

other
hand,

doping

ZnO

with

various

elements,

e.g.,

noble

metals

[18–20],
rare-earth

metals

[21],

transition

metals

[22],

and

metal

oxides

[23]
∗
Corresponding

authors.

Tel.:

+81

22

217

5933;

fax:

+81

22

217

5930.
E-mail

addresses:



(T.

Liu),

,
(Z.

Wang).
has

been

suspected

to

enable

modulation

of

surface

charge

states
of

ZnO,

modifying

significantly

its

functionality.
A

general

approach

to

date

to

fabricate

nanomaterials

with

the
hollow

and

porous

morphologies

accompanies

the

use

of

remov-
able

or

sacrificial

templates,

including

either

the

hard

ones

such

as
monodisperse

silica

[24],

polymer

latex

spheres,

[25,26]

and

reduc-
ing

metal

nanoparticles

[27],

or

the

soft

ones

such

as

emulsion
micelles

[28]

and

gas

bubbles

[29].

The

disadvantages

for

the

use
of

templates

though

rest

with

the

high

cost

and

tedious

synthe-
sis

process,

posing

a

significant

hurdle

to

the

large-scale

industrial
applications.

Ideally,

one

would

prefer

a

one-step

template-free
method

to

synthesize

the

nanomaterials

with

hollow

and

porous

superstructures

in

a

size

tunable

manner.

Recent

breakthroughs
in

the

fabrication

of

nanomaterials

by

taking

full

advantage

of
known

physical

phenomena,

e.g.,

oriented

attachment

[30,31],

Ost-
wald

ripening

[32–34],

Kirkendall

effect

[35,36],

and

etching-based
inside-out

evacuation

[37,38],

has

brought

such

“ideal”

concept
closer

to

reality.

Among

all

the

fabrication

techniques,

the

etching
process

has

been

demonstrated

as

a

facile

choice

for

preparing

hol-
low

and

porous

nanomaterials

because

it

is

easy

to

dissolve

inner
nano-crystallites

via

adjusting

processing

time

and

temperature
[39–41].
Here,

we

report

a

technically

simple

and

flexible

route:

the

use
of

a

template-free

hydrothermal

process

to

prepare

the

hollow
and

porous

ZnO

nanospheres

with

a

large

specific

area

in

a

con-
trollable

manner.

We

investigate

in

detail

the

effect

of

the

liquid
mixture

ratio

and

hydrothermal

time

on

the

morphology

evolution
0925-4005/$

–

see

front

matter ©

2012 Elsevier B.V. All rights reserved.
/>54 W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62
and

propose

a

new

mechanism

that

is

responsible

for

the

unusual
nucleation

and

self-assembly

of

ZnO

building

blocks,

i.e.,

coupling
of

Ostwald

ripening

with

grain-rotation-induced

grain

coalescence
(GRIGC).

Such

a

unique

morphology

is

maintained

after

doping
yttrium

(Y)

to

produce

hybrid

functionality

of

ZnO

as

a

gas-sensing
material,

which,

to

the

best

of

our

knowledge,

has

rarely

been
reported.

Our

results

demonstrate

that

Y-doped

ZnO

nanospheres
lower

remarkably

resistance

and

enhance

gas-sensing

perform-
ances,

which

may

open

up

a

new

avenue

to

develop

advanced

gas
sensors.
2.

Experimental

All

ZnO

nanospheres

were

synthesized

by

the

hydrothermal
method.

Zinc

acetate

dehydrate

(Zn(CH
3
COOH)
2
·2H
2
O)

(4

mM)
was

first

dissolved

into

a

mixed

solution

of

ethanol

(40

mL)

and
monoethanolamine

(MEA)

(30

mL)

under

mechanical

stirring

for
1

h.

The

solution

was

then

transferred

in

autoclaves,

which

were
heated

to

160
◦
C

for

24

h

to

produce

precipitate.

The

pure

ZnO
powder

was

prepared

by

centrifuging

the

precipitate,

washing
it

with

distilled

water

and

ethanol

to

remove

unwanted

ions,
and

drying

at

60
◦
C

in

air.

The

obtained

powder

(0.03

g)

was

dis-
persed

in

deionized

water

(20

mL),

and

1.5

mL

mixed

solution

of
ethanol

and

yttrium

nitrate

hexahydrate

(N
3
O
9
Y·6H
2
O)

(0.01

M)
was

then

added.

The

solution

was

stirred

thoroughly

for

1

h

and
dried

at

80
◦
C

in

air

before

annealing

at

400
◦
C

for

2

h

to

elimi-
nate

NO
3
–
ions.

The

Y-doped

ZnO

powder

with

a

mass

ratio

of

Y
to

Zn

of

4%

was

harvested.

To

make

a

straightforward

compari-
son,

the

ZnO

nanoparticles

were

also

prepared

by

dissolving

4

mM
Zn(CH
3
COOH)
2
·2H
2
O

and

20

mM

NaOH

in

70

mL

distilled

water,
which

was

then

transferred

in

autoclaves

and

heated

at

160
◦

C

for
20

h.
Microstructure

analysis

was

conducted

by

the

X-ray

diffrac-
tion

(XRD),

scanning

electron

microscopy

(SEM),

and

transmission
electron

microscopy

(TEM).

For

the

XRD,

a

Rigaku

D/Max-
1200X

diffractometry

with

Cu

K˛

radiation

operated

at

40

kV
and

200

mA

was

applied.

Surface

morphologies

of

the

sam-
ples

were

observed

using

a

Hitachi

S-4300

SEM.

Microstructures
and

chemical

composition

were

analyzed

using

the

JEOL

JEM-
2010F

electron

microscope

operated

at

an

accelerating

voltage

of
200

kV.

Specific

surface

area

was

measured

upon

the

multipoint
Brunauer–Emmett–Teller

(BET)

analysis

of

nitrogen

adsorption
isotherms,

which

were

recorded

on

a

surface

area

analyzer
(Micromeritics,

ASAP

2020M).
The

powders

upon

harvest

were

mixed

with

diethanolamine
and

ethanol

to

form

pastes,

which

were

subsequently

coated

onto
an

alumina

ceramic

tube

pre-loaded

with

a

pair

of

gold

electrodes
at

each

end.

Next,

the

tube

was

dried

at

400
◦
C

for

2

h

in

order
to

eliminate

organic

binder

as

well

as

strengthen

the

bonding
between

the

pastes

and

tube.

A

Ni–Cr

wire

was

placed

inside

the
tube

as

a

heater.

The

heating

wire

and

tube

were

soldered

on

the
pedestals

to

fabricate

gas

sensors.

The

sensors

were

finally

aged
at

200
◦
C

for

240

h

in

order

to

improve

stability

and

repeatabil-
ity.

Gas-sensing

measurements

were

conducted

using

a

computer
controlled

measurement

system

(HW-30A,

Hanwei

Electronics

Co.,
Ltd.)

at

room

temperature

at

a

humidity

of

40%.

The

sensor

was
first

connected

to

the

circuit

board

of

measurement

system,

and
then

the

tested

gas

was

introduced

into

the

glass

chamber

through
injecting

a

given

amount

of

gas.

The

operating

temperature

of

sen-
sors

can

be

adjusted

precisely

via

altering

the

current

flow

across
the

Ni–Cr

heater.

Resistance

(R
s
)

of

the

sensors

was

estimated

by
R
s
=

R
L
(V
c
−

V
out
)/V
out
,

where

the

R
L
was

resistance

of

a

load

resistor
(R
L
=

47

k),

and

the

V
c
and

V
out
were

circuit

and

output

volt-
age

(V
c
=

6

V),

respectively.

The

sensor

response

(S)

was

defined

as
S

=

R
a
/R
g
at

reductive

atmosphere,

while

as

S

=

R
g
/R
a
at

oxidative
Fig.

1.

XRD

spectra

of

Y-free

and

Y-doped

ZnO

nanospheres

with

a

series

of

Y/Zn

ratios.

Textural

orientations

of

detected

matters

are

given

as

well

for

easy

reference.
atmosphere,

where

R

a
and

R
g
were

resistance

in

air

and

target

gas,
respectively.

The

response

and

recovery

time

was

defined

as

the
interval

between

when

response

reached

90%

of

its

maximum

and
dropped

to

10%

of

its

maximum.
3.

Results

and

discussion
3.1.

Chemical

composition

and

morphology
To

identify

chemically

the

prepared

samples,

we

first

conducted
XRD

analyses,

as

shown

in

Fig.

1,

where

textural

orientation

of

the
detected

matters

is

shown

as

well

for

easy

reference.

For

the

ZnO,
2%

and

4%

Y-doped

ZnO

samples,

all

of

the

peaks

are

identified

as
belonging

to

the

wurtzite

(hexagonal)

structure

of

ZnO

(JCPDS

(36-
1451)).

No

secondary

phase

is

detected

although

the

lattice

of

the
Y-doped

sample

is

found

to

be

somewhat

expanded

as

compared
to

the

Y-free

sample.

In

addition

to

ZnO,

Y
2
O
3
is

also

detected

in

the
6%

and

8%

Y-doped

ZnO

samples.

This

suggests

that

the

Y

atoms

fill
the

lattice

sites

of

ZnO

at

the

low

doping

concentration,

but

tend
to

form

a

new

Y
2
O
3
phase

at

high

doping

concentration

(over

6%).
However,

there

are

no

characteristic

secondary-phase

XRD

peaks
in

the

2%

and

4%

Y-doped

ZnO

samples,

indicating

that

the

sec-
ondary

phase

is

very

scarce

or

highly

dispersed.

This

is

because
there

appear

well

defined

XRD

peaks

if

size

of

the

crystallites

is
above

1–3

nm

[42].

This

case

is

also

recognized

in

two-phase

sys-
tems,

in

which

the

secondary

phase

with

a

small

concentration

is
highly

dispersed

on

surfaces

of

the

basic

oxide’s

grains.

These

indi-
cate

that

the

secondary

oxide

phase,

if

have,

should

have

a

smaller
grain

size

than

the

basic

oxide

in

the

samples

with

Y

doping

con-
centrations

of

2%

and

4%.

When

the

doping

concentration

is

over
6%,

a

new

Y
2
O
3
phase

is

formed

with

a

grain

size

larger

than

3

nm.
Table

1

lists

the

lattice

constants

of

both

the

undoped

and

Y-doped
samples

obtained

from

XRD

data

and

the

crystallite

size

calculated
using

the

Scherrer

formula.

The

lattice

constants

(a

and

c)

and

grain
sizes

increase

with

the

rise

of

the

amount

of

Y,

suggesting

that

the
introduction

of

Y

distorts

the

crystal

structure

of

the

host

oxide.

Table

1
The

lattice

constants

of

the

Y-doped

ZnO

sphere

and

ZnO

nanoparticle,

and

grain
sizes

calculated

using

the

Scherrer

formula.
ZnO

Lattice

constant

Grain

size

(nm)
a

(Å)

c

(Å)
0%

Y

doped 3.24926

5.20505

18.5
2%

Y

doped

3.25261

5.20983

19.3
4%

Y

doped

3.25689

5.21532

21.6
6%

Y

doped 3.25795

5.21823

22.4
8%

Y

doped

3.25948

5.21993

23.5
Nanoparticle 3.26889

5.22845

28.9
W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62 55
This

is

because

the

radius

of

Y
3+
ion

(0.92
˚
A)

is

larger

than

that

of
Zn
2+
(0.74
˚
A),

which

should

increase

the

lattice

constants

of

ZnO
by

Y

doping

and

hence

result

in

a

shift

in

diffraction

peak

toward
lower

2␪

angle.
To

unveil

morphologies

of

the

prepared

samples,

we

show

in
Fig.

2

SEM

images

of

representative

regions

in

the

pristine

and
doped

nanospheres

and

the

nanoparticles.

As

seen

in

Fig.

2(a),
the

Y-free

sample

is

indeed

characterized

as

nanospheres,

which
are

uniformly

distributed.

These

nanospheres

are

coarse

on

sur-
face

(Fig.

2(b)

and

(c))

and

hollow

inside,

as

clearly

verified

in

a
broken

nanosphere

(Fig.

2(d)).

Interestingly,

the

nanospheres

are
self-assembled

to

radially

aligned

nanorods

of

∼150

nm

in

length
from

their

cores

yet

to

self-wrapped

irregular

nanoparticles

at

the
cores.

Pores

turn

up

on

the

nanosphere

surfaces,

indicating

that

the
as-synthesized

pristine

samples

are

not

only

hollow

but

porous.
Such

a

hierarchical

morphology

are

further

corroborated

from

the
TEM

images

showing

a

difference

in

the

image

contrast

between

the
margin

and

center

of

nanospheres,

i.e.,

the

center

seems

brighter,

which

indicates

the

formation

of

the

well-defined

hollow

nano-
structures

(Fig.

3(a)).

Fig.

3(b)

and

(c)

gives

TEM

images

of

edge
regions

of

the

nanospheres,

which

show

unambiguously

the

pores
(Fig.

3(b)),

nanorods,

and

nanoparticles

(Fig.

3(c)).

Fig.

3(d)

presents
a

high-resolution

TEM

(HRTEM)

image

of

an

edge

of

a

nanosphere
(only

the

edge

is

likely

for

imaging

due

to

the

large

thickness

away

from

surface),

from

which

lattice

fringes

are

clearly

visible.
The

spacing

between

neighboring

lattice

planes

is

estimated

to

be
∼0.26

nm,

in

line

with

that

between

the

(0

0

0

1)

planes

of

a

hexag-
onal

ZnO

(inset

of

Fig.

3(d)),

suggesting

that

the

ZnO

nanorods

grow
in

the

[0

0

0

1]

direction.
Such

interesting

hollow

and

porous

nanospheres

are

not

dis-
turbed

significantly

by

Y

doping

(Fig.

2(e)–(g)),

although

their

size
becomes

larger

due

to

the

growth

during

post-annealing.

Fig.

2(h)
shows

the

morphology

of

nanoparticles

for

a

comparison,

which
are

accumulated

with

a

mean

size

of

∼50

nm.

Fig.

3(e)–(g)

presents
TEM

images

of

the

Y-doped

samples,

where

they

retain

the

porous
and

hollow

nature.

Like

what

was

seen

in

the

pristine

sample,

the
nanorods

also

grow

in

the

[0

0

0

1]

direction

even

when

the

Y

is
doped.

To

identify

chemically

the

samples,

we

perform

an

energy-
dispersive

X-ray

spectroscopy

(EDS)

analysis

of

a

representative
nanosphere

in

the

pristine

and

4%

Y-doped

sample,

as

shown

in
Fig.

3(h).

The

nanospheres

in

the

pristine

sample

are

composed

of
40.4

at%

O

and

59.6

at%

Zn,

while

those

in

the

4%

Y-doped

sample
34.21

at%

O,

63.78

at%

Zn

and

2.01

at%

Y,

demonstrating

that

the
embedded

additive

of

Y

is

really

present

in

the

ZnO

matrix.

Further
EDS

mapping

of

both

the

entire

sphere

and

the

edge

reveals

an

even
distribution

of

O

(Fig.

3(j))

and

Zn

(Fig.

3(k)),

providing

direct

evi-
dence

to

the

uniform

distribution

of

Y

in

the

doped

sample

(Fig.

3(l))
and

further

testifying

the

second

oxide

phase

is

present

in

the

4%
Y-doped

ZnO

matrix,

in

consistence

with

the

XRD

results.
3.2.

Formation

mechanism

of

hollow

and

porous

nanospheres
To

gain

insight

into

formation

mechanism

of

the

hierarchi-
cal

nanostructures

and

how

morphology

evolves

with

processing
conditions,

we

first

investigate

systemically

the

role

of

solvent

com-
position

on

the

structures

of

nanomaterials.

As

seen

in

Fig.

4(a)
and

(b),

the

ZnO

nanorods

are

clustered

when

the

MEA

is

not
introduced.

Once

the

MEA

is

added

(5

mL),

the

nanorods

are

bun-
dled

irregularly

and

loosely

with

a

mean

length

of

500

nm

(Fig.

4(c)
and

(d)).

Further

increase

in

the

MEA

concentration

(15

mL)

renders
these

bundles

self-assembled

to

fan-shaped

hemispheres

(Fig.

4(e)
and

(f)).

The

hollow

and

porous

nanospheres

are

formed

when

the
concentration

of

MEA

is

increased

further

to

30

mL

(Fig.

4(g)

and
(h)).

The

nanospheres

become

denser

with

fewer

holes

on

surfaces
as

the

MEA

concentration

is

increased

to

40

mL

(Fig.

4(i)

and

(j)).
However,

the

nanospheres

are

nonporous

and

solid

when

the

con-
centration

of

MEA

is

increased

to

50

mL

(Fig.

4(k)

and

(l)),

implying
that

the

precise

control

of

the

MEA

concentration

is

essential

to
producing

a

hierarchical

superstructure.
The

formation

of

nanorods

in

the

[0

0

0

1]

direction

without

MEA
is

understood

upon

the

structural

anisotropy

and

surface

polarity

of
ZnO.

The

(0

0

0

1)

polar

plane

is

the

most

energetically

unfavorable

and

bears

the

highest

growth

rate,

followed

by

(1

0

1

1),

(1

0

1

0),
(
1

0

1

1),

and

(0

0

0

1)

planes

(inset

of

Fig.

3(d))

[43,44].

Once

the
MEA

is

in

the

ethanol

solution,

the

coordinated

[Zn(MEA)
m
]
2+
ions
(where

m

is

an

integer)

are

generated,

restraining

the

formation
of

free

Zn
2+
ions

and

the

Zn(OH)
2
,

the

nuclei

of

ZnO

nanomaterial.
The

chemical

reactions

in

presence

of

MEA

during

hydrothermal
process

can

be

expressed

as:
Zn
2+
+

mMEA

↔

[Zn(MEA)
m
]
2+
,

(1)
Zn(OOCCH
3
)
2
·2H
2
O

+

2C
2
H
5
OH

→

Zn(OH)
2
+

2H
2
O
+

2CH
3
COOC
2
H
5
, (2)
Zn(OH)
2
↔

ZnO

↓

+

H
2
O. (3)
As

the

temperature

is

increased

in

the

autoclaves,

the

[Zn(MEA)
m
]
2+

ions

are

ready

to

be

decomposed

to

Zn
2+
ions

and

ethanolamine
molecules

(Eq.

(1)).

Simultaneously,

there

occurs

the

esterifica-
tion

of

zinc

acetate

with

ethanol

to

produce

Zn(OH)
2
(Eq.

(2)),
which

is

ultimately

decomposed

to

ZnO

nanomaterials

(Eq.

(3)).
The

ethanolamine

molecules,

which

are

adsorbed

on

the

surfaces

of
ZnO

nuclei,

can

serve

as

assembling

agents,

refraining

crystals

from
forming

nanorods

along

the

[0

0

0

1]

direction

[45].

The

metastable
nanoparticles

are

produced

instead

at

the

initial

nucleation

stage,
which

is

important

for

the

next-stage

Ostwald

ripening

procedure.
Two

factors

are

responsible

for

the

evolution

of

morphology

at
solvothermal

condition:

the

initial

nucleation

status

and

the

solu-
bility

of

precursors

in

solvent

under

saturation

vapor

pressure

[46].
Note

that

the

solvent

MEA

is

lower

than

ethanol

in

the

saturation
vapor

pressure

due

to

its

higher

boiling

point

(78.29
◦
C

for

ethanol

while

170
◦
C

for

MEA).

This

gives

rise

to

extensive

nucleation

of
metastable

nanoparticles,

which

are

aggregated

into

nanospheres
to

lower

their

surface

areas

and

energies

[47].

In

addition

to

form-
ing

the

spherical

configuration,

the

MEA

also

plays

a

pivotal

role
in

making

the

nanospheres

hollow

and

porous.

In

contrast

to

the
fast

migration

and

high

nucleation

rate

of

the

reactive

species

in
ethanol,

it

is

kinetically

slower

to

form

metastable

nanocrystals

in
MEA

solution

due

to

the

higher

boiling

point

and

viscosity

of

MEA.
This

allows

the

mixture

of

nanocrystals

with

varying

growth

orien-
tations

to

assemble

into

spherical

nanoparticles.

On

the

other

hand,
the

MEA

is

able

to

facilitate

the

formation

of

metastable

nanoparti-
cles,

the

interiors

of

which

are

susceptible

to

be

dissolved,

thereby
producing

the

hollow

nanospheres.
To

shed

more

light

on

the

formation

mechanism

of

hollow,

porous

nanospheres,

we

conduct

a

series

of

time-dependent

inves-
tigations,

as

shown

in

Fig.

5.

At

the

early

stage

(4

h),

solid

spherical
nanoparticles

(Fig.

5(a)

and

(b))

are

formed

(Fig.

5(c)).

As

the
reaction

time

is

increased

to

8

h,

hollowing

process

starts

at
the

nanosphere

cores

(Fig.

5(d)

and

(e)),

and

the

surfaces

of
nanospheres

turn

rough

(Fig.

5(f)),

indicating

that

a

portion

of
particles

on

surfaces

are

dissolved.

Further

extension

of

reaction
time

(16

h)

enhances

the

hollowing

effect

(Fig.

5(g)

and

(h)),

and
the

numerous

nanorods

with

pores

on

surfaces

are

assembled

to
nanospheres

(Fig.

5(h))

due

to

the

dissolution

and

recrystalliza-
tion

(Fig.

5(i)).

The

hollow

and

porous

nanospheres

are

formed

as
the

reaction

time

is

24

h.

However,

there

emerge

urchin-like

struc-
tures

comprising

a

large

amount

of

nanorods

with

a

small

number
of

nanoparticles

(Fig.

5(j))

as

the

reaction

time

is

30

h

(Fig.

5(k)

and
(l)).

Interestingly,

most

of

the

nanoparticles

are

dissolved

when

the
56 W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62
Fig.

2.

SEM

images

of

the

pristine

ZnO

nanospheres

taken

at

(a)

low

and

(b)

high

magnification.

(c)

and

(d)

Magnified

SEM

images

of

an

open

hollow

and

porous

ZnO
nanosphere.

SEM

images

of

the

Y-doped

nanospheres

taken

at

(e)

low

and

(f)

slightly

higher

magnification.

(g)

Magnified

SEM

image

of

an

open

nanosphere

in

the

Y-doped
sample.

(h)

SEM

image

of

the

nanoparticles.
reaction

time

is

increased

to

35

h,

leaving

behind

slim

nanorods
(Fig.

5(m)

and

(o)).

The

disappearance

of

the

nanospheres

sug-
gests

the

important

role

of

the

nanoparticles

in

the

stabilization
of

nanospheres

(Fig.

5(n)).
These

imply

such

a

mechanism:

the

Ostwald

ripening

[48]

cou-
pled

with

the

grain

rotation

induced

grain

coalescence

(GRIGC)
[49].

The

Ostwald

ripening

involves

the

aggregation

of

nano-
crystallites,

followed

by

an

outward

mass

transfer

to

form

hollow
structures.

The

GRIGC

process

occurs

when

particles

collide,

and
the

grain

rotation

takes

place

thereafter.

Such

grain

rotation

low-
ers

the

energy

of

system

and

eliminates

the

grain

boundaries,
producing

single-phase

nanocrystals

(i.e.,

coalescence

process).
Fig.

3.

(a)

TEM

image

of

the

Y-free

ZnO

sample,

highlighting

that

the

nanospheres

are

hollow.

(b)

and

(c)

Enlarged

TEM

images

of

the

Y-free

ZnO

nanospheres

on

edge.

(d)

HRTEM

image

of

a

Y-doped

ZnO

nanosphere.

(e)

TEM

image

of

the

Y-doped

ZnO.

(f)

and

(g)

Enlarged

TEM

images

of

the

doped

sample

on

edge.

(h)

EDS

for

the

pristine

(upper)
and

doped

(lower)

ZnO

nanomaterials.

The

horizontal

axis

denotes

the

energy

and

the

vertical

one

the

counts

(i.e.,

intensity).

(i)

Original

area

and

EDS

mapping

of

(j)

O,

(k)

Zn,

and

(l)

Y

elements

in

a

Y-doped

nanosphere.

The

insets

show

mapping

of

the

edge

region.
W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62 57
Fig.

4.

SEM

image

of

the

ZnO

nanospheres

prepared

under

different

concentrations

of

monoethanolamine

(MEA):

(a)

and

(b)

0

mL,

(c)

and

(d)

5

mL,

(e)

and

(f)

15

mL,

(g)

and
(h)

30

mL,

(i)

and

(j)

40

mL,

(k)

and

(l)

50

mL.

The

amount

of

added

ethanol

is

fixed

to

be

40

mL.
Fig.

6

shows

schematically

formation

evolution

of

the

hollow,
porous

nanospheres.

At

the

initial

stage,

the

ZnO

nanocrystals

are
generated

randomly.

As

the

reaction

time

is

increased,

the

ZnO
colloids

are

aggregated

to

form

solid

nanospheres

through

the
Ostwald

ripening

effect,

which

is

driven

by

the

minimization

of

sur-
face

energy.

Since

crystallites

have

a

higher

surface

energy

in

the
interiors

than

on

the

surfaces,

they

are

more

readily

to

be

dissolved.
Once

being

heated,

the

nano-crystallites

are

easier

to

be

collided
Fig.

5.

SEM

image

illustrating

evolution

of

morphology

of

the

nanospheres

with

the

reaction

time:

(a–c)

4

h,

(d–f)

8

h,

(g–i)

16

h,

(j–l)

30

h,

and

(m–o)

35

h.

Three

images
with

different

magnification

are

provided

in

each

case.
58 W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62
Fig.

6.

Schematic

illustration

of

morphology

evolution

of

the

ZnO

nanospheres.
and

rotated,

giving

rise

to

coalescence

of

neighboring

grains

to
form

large

single-phase

grains.

Such

a

process

lowers

the

inter-

facial

energy

associated

with

large

interfacial

area.

For

the

polar
ZnO,

the

(0

0

0

1)

plane

is

most

likely

to

be

coalesced

due

to

its
highest

energy

of

all

planes,

which

explains

the

observation

that
ZnO

crystallites

are

grown

in

the

[0

0

0

1]

direction

to

produce

the
rod-like

ZnO

in

the

shell

of

hollow

nanospheres.

Meanwhile,

the
rotation

and

migration

of

particles

induce

pores,

which

results

in
the

hollow

and

porous

morphology.
3.3.

Resistance

as

a

function

of

temperature
To

gain

more

insights

into

surface

properties

of

the
nanospheres,

we

show

in

Fig.

7

nitrogen

adsorption–desorption
isotherm

and

size

distribution

of

the

pores

calculated

using

the
Barret–Joyner–Halenda

(BJH)

method.

Careful

analysis

of

the

plot
identifies

the

isotherm

as

a

type

IV

one,

indicating

the

formation
of

typical

porous

structure

(Fig.

7(a)).

Although

the

pore

spans
a

large

range

in

size,

the

majority

of

pores

have

a

diameter

of
5–15

nm,

in

accord

with

the

above

TEM

observations

(Fig.

7(b)).
The

specific

surface

area

of

the

nanospheres

reaches

89.5

m
2
g
−1

measured

using

the

BET,

confirming

the

porous

nature.

It

should
be

noted

that

the

specific

surface

area

of

the

nanoparticles

is

only
24.2

m
2
g
−1
,

which

indicates

that

the

morphology

affects

greatly
the

specific

surface

area.
Such

a

3D

hierarchical

porous

nanostructure

may

hold

sub-
stantial

promise

for

a

wide

range

of

applications,

especially

as
a

chemical

sensor

owing

to

the

large

specific

surface

area

that
can

greatly

enhance

gas

diffusion

and

mass

transport.

Extensive
effort

has

been

devoted

to

date

to

improving

gas-sensing

prop-
erties

of

ZnO,

including

the

fast

response

and

recovery

and

high
gas

response.

Among

them,

the

doping

of

rare-earth

elements,
e.g.,

Y,

has

been

demonstrated

as

one

of

effective

ways

to

activate
host

materials,

and

may

enable

fictionalization

of

the

hierarchical
nanospheres

as

well

for

advanced

functional

gas

sensors.
To

test

this

scenario,

we

first

present

in

Fig.

8(a)

the

resistance
(R)

as

a

function

of

temperature

(T)

for

the

sensors

fabricated

with
the

Y-free

and

Y-doped

nanospheres

in

air

together

with

the

sen-
sor

made

of

pristine

ZnO

nanoparticles

(Fig.

2(g)).

Overall

feature

is
different

between

samples

and

the

sample

doped

with

4%

Y

has

the
lowest

resistance.

The

resistance

decreases

with

increasing

amount
of

Y,

but

such

a

decrease

comes

to

a

halt

when

the

doping

concen-
tration

of

Y

is

beyond

4%.

Carre
˜
no

et

al.

reported

the

formation

of
a

second

phase,

Sn
2
Y
2
O
7
,

in

the

SnO
2
doped

with

a

small

amount
of

Y

[50].

Likewise,

a

similar

second

phase

of

ZnY
m
O
n
may

be

pro-
duced

in

our

samples

when

the

doping

concentration

is

lower

than
4%.

The

second

phase

has

a

low

resistance,

providing

conducting
channels

in

the

sample

and

hence

lowering

the

contact

resistance
of

the

ZnO

grains.

This

may

reduce

the

resistance

of

ZnO

sample.
However,

when

the

doping

concentration

is

above

4%,

the

doped
Y

in

ZnO

reaches

saturation,

forming

Y
2
O
3
precipitates

that

grow
along

the

ZnO

grain

boundary.

The

Y
2
O
3

has

a

higher

resistance
than

the

matrix,

thereby

increasing

the

contact

resistance.

This
consequently

suppresses

the

dropping

of

overall

resistance

signif-
icantly,

that

is,

the

resistance

is

increased.
Another

key

feature

in

Fig.

8(a)

is

that

resistance

drops

in

a

less
abrupt

manner

for

the

Y-doped

(Y/Zn

=

4%)

than

Y-free

ZnO

at

the
temperature

ranging

from

300

to

450
◦
C,

which

could

be

attributed
to

the

chemisorbed

O

on

surfaces.

However,

the

reversible

reactions
take

place

among

O

gas

(gas),

chemisorbed

O

(ads),

and

lattice

O
(lat)

with

the

rise

of

temperature

[51]:
O
2

(gas)

+

e
−
⇔

O
−
2
(ads),

(4)
1
2
O
2
+

e
−
⇔

O
−
ads
,

(5)

1
2
O
2
+

2e
−
⇔

O
2−
ads
,

(6)
O
2−
ads
⇔

O
2−
lat
, (7)
These

conclude

intuitively

that

electron

transfer

from

semiconduc-
tor

to

absorbed

O

is

responsible

for

the

increase

of

resistance.

It

has
been

reported

that

pure

ZnO

materials

exhibit

n-type

semiconduc-
tor

characteristics

due

to

the

existence

of

oxygen

vacancies

[52].
From

the

EDS

analysis,

we

find

that

the

O/Zn

ratio

decreased

from
67.7%

(ZnO)

to

53.6%

(4%

Y-doped

ZnO).

The

fewer

amounts

of

oxy-
gen

and

zinc

in

the

Y-doped

ZnO

reveal

the

increase

of

defects

with
the

introduction

of

Y

in

the

ZnO

nanospheres.

Meanwhile

the

asso-
ciated

increase

in

lattice

constant

gives

rise

to

increased

intrinsic
defects,

e.g.,

V
•
O
,

V
••
O
,

and

O
//
i
[53].

During

the

hydrothermal

process,

defects

can

be

produced

and

further

enhanced

by

the

doping

of

Y
in

the

ZnO

nanospheres.

It

is

worth

noting

that

ZnO

has

a

hexago-
nal

close-packed

lattice

with

a

relatively

open

structure

in

which
Zn

atoms

occupy

half

of

the

tetrahedral

sites

and

all

the

octahe-

dral

sites

are

empty.

In

general,

the

oxygen

vacancy

(V
••
O
)

has

lower
formation

energy

than

the

zinc

interstitials

(Zn
••
i
),

resulting

in

Zn-
rich

compositions

in

the

real

wurtzite

ZnO

[52].

In

this

sense,

the
intrinsic

defects

and

extrinsic

dopants

can

be

introduced

during
the

fabrication.

Xu

also

pointed

out

that

the

O
2
molecules

interact
strongly

with

oxygen

vacancies

on

the

surface

of

ZnO

[54].

These
imply

that

the

Y

doping

can

increase

the

concentration

of

O

vacancy
and

hence

absorb

more

oxygen

on

the

ZnO

surface,

which

as

a

result
increases

the

concentration

of

O
−
.
3.4.

Gas-sensing

performance
To

gain

insight

into

gas-sensing

properties

of

the

ZnO

nano-
structures,

we

present

in

Fig.

8(b)

gas

response

to

formaldehyde
(HCHO)

gas

as

a

function

of

temperature

at

50

ppm.

The

Y-free
nanospheres

show

a

higher

gas

response

(maximum

value

of

47.4
at

350
◦
C)

than

the

undoped

nanoparticles,

indicating

that

mor-
phology

is

critical

to

the

enhancement

of

gas-sensing

functionality.
Evidently,

response

of

ZnO

nanospheres

is

improved

with

the

addi-
tion

of

Y.

However,

gas

response

is

saturated

to

a

maximum

value

of
65.7

when

the

Y

concentration

reaches

4%.

Further

increase

in

the
Y-doping

concentration

results

in

an

adverse

effect,

i.e.,

lowers

the
gas

response.

The

Y-doped

nanospheres

have

a

lower

optimal

tem-
perature

(300
◦
C)

than

the

Y-free

ones

(350
◦
C).

The

enhancement
W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62 59
Fig.

7.

(a)

Nitrogen

adsorption–desorption

isotherm

and

(b)

corresponding

pore

size

distribution

of

the

ZnO

hollow

and

porous

nanospheres.
of

gas-sensing

properties

by

Y

doping

can

be

understood

as

fol-
lows.

In

the

pristine

case,

the

O

molecules

are

adsorbed

on

surfaces
and

capture

e

from

the

ZnO

semiconductor,

forming

chemisorbed
O

species

(Eqs.

(7)–(10)).

Such

process

gives

rise

to

surface

deple-
tion

layers,

which

eventually

increases

resistance

of

the

samples.
When

being

exposed

to

the

HCHO,

the

HCHO

molecules

react

with
the

adsorbed

O

on

surfaces

in

the

following

manner:
HCHO

(gas)

+

2O
−
ads
⇒

CO
2
+

H
2
O

+

2e
−
.

(11)
This

process

releases

the

trapped

electrons

back

to

conduction

band
of

ZnO,

increasing

thereby

concentration

of

carriers

in

the

semi-
conductor

[55,56].

The

introduction

of

Y

induces

oxygen

defects
in

ZnO,

increases

concentration

of

O
−
ads
and

hence

improves

gas
response.

On

the

other

hand,

the

low

optimal

operating

tem-
perature

after

the

Y

doping

can

be

ascribed

to

the

formation

of
weakly

bonded

complexes

ZnY
m

O
n
.

The

chemisorption

of

oxygen
species

depends

strongly

on

the

temperature

and

nature

of

mate-

rial.

The

O
2
is

chemisorbed

at

low

temperature

while

O
−
and

O
2−
are

chemisorbed

at

high

temperature.

Since

ZnO

is

a

semiconduc-
tor,

oxygen

absorption

and

electron

transfer

are

rather

difficult

to
occur

at

room

temperature.

The

thermal

activation

of

the

semicon-
ductor

is

required

to

observe

gas

adsorption

on

surface.

This

is

why
change

in

resistance

is

not

observed

when

the

ZnO

nanospheres
are

exposed

in

the

reduced

gases.

However,

the

low-temperature
gas

adsorption

becomes

possible

by

the

Y

doping

due

to

the

pres-
ence

of

the

weakly

bonded

complexes

ZnY
m
O
n
on

the

ZnO

grain
surface.

The

absorption

of

oxygen

ions

can

occur

on

the

ZnO

sur-
face

at

room

temperature

due

to

the

high

conducting

nature

of

the
ZnY
m
O
n
.

In

this

respect,

the

Y

activates

reactions

of

HCHO

to

the
adsorbed

O

due

to

the

spillover

effect

[57–59],

resulting

in

a

lower
optimal

operating

temperature.
Fig.

9

shows

response–recovery

characteristics

for

the

three

sen-
sors

fabricated

with

the

pristine

nanoparticles,

Y-free

and

Y-doped
nanospheres

at

different

operating

temperatures.

Six

representa-
tive

species

of

volatile

organic

compound

(VOC)

gases

are

chosen
purposely,

including

CH
4
,

NH
3
,

HCHO,

CH
3
OH,

CO,

and

(CH
3
)
2
CO.
The

gas

concentration

is

fixed

to

50

ppm.

The

voltage

is

increased
sharply

when

the

test

gas

is

in,

yet

returns

to

its

original

state
when

gas

is

out.

The

key

difference

among

the

three

samples

is
that

the

voltage

is

increased

in

the

most

strikingly

manner

for

the
Y-doped

sample,

verifying

again

the

gas-response

enhancement

by
morphology

and

Y

doping.

Moreover,

the

response

and

recovery
transient

of

these

sensors

is

superior

to

HCHO

than

to

the

rest

of
the

tested

VOC

gases,

especially

in

the

Y-doped

case

(Fig.

9(a)–(c)).
Upon

closer

inspection,

we

find

that

the

response

and

recovery

time
is

∼14

and

17

s

for

the

pristine

nanoparticles,

while

∼10

and

12

s

for

the

Y-free

nanospheres.

They

are

further

shortened

to

∼4

and
6

s

for

the

Y-doped

nanospheres

(Fig.

9(d)).
To

shed

more

light

on

the

Y-doped

sample,

we

further

measure
the

gas-sensing

properties

at

the

optimal

operating

temperature

of
300
◦
C,

as

shown

in

Fig.

10.

The

gas

response

is

increased

drasti-
cally

as

the

gas

concentration

is

increased

up

to

250

ppm,

yet

in

a
more

gentle

fashion

as

the

concentration

is

increased

further.

The
response

is

saturated

at

∼800

ppm.

Interestingly,

the

gas

response
is

increased

almost

linearly

when

the

gas

concentration

ranges
from

10

to

100

ppm

(inset

of

Fig.

10(a)),

implying

that

the

Y-doped
ZnO

nanomaterial

works

even

at

low

gas

concentration.

Fig.

10(b)
shows

gas

response

of

the

Y-doped

sensor

to

the

six

types

of

VOC
gases

at

50

ppm.

The

result

made

clear

is

that

the

response

to

the
HCHO

reaches

a

maximum

value

of

65.7

but

is

no

larger

than

16
to

other

gases.

This

implies

that

the

Y-functionalized

nanosphere
can

act

as

an

efficient

gas-sensing

material

for

on-site

selective
detection

of

formaldehyde.

Since

the

formaldehyde

has

a

single
aldehyde

and

high

reducibility

in

detecting

gases,

the

unsaturated
Y

ions

tend

to

absorb

HCHO

molecules,

forming

complex

species
of

Y–HCHO

[59].

Simultaneously,

the

absorb

oxygen

on

the

surface
oxidizes

the

HCHO

into

H
2

O

and

CO
2
,

resulting

in

a

good

selectiv-
ity

to

HCHO

for

the

sensor.

Fig.

10(c)

shows

a

single-cycle

response
for

the

Y-doped

sensor

at

different

HCHO

concentrations

at

300
◦

C.
The

voltage

signal

(i)

is

enlarged

with

the

rise

of

HCHO

concentra-
tion,

(ii)

is

stabilized

in

4

s

when

the

sensor

is

exposed

in

the

HCHO
atmosphere,

and

(iii)

returns

to

original

state

in

6

s

once

the

sen-
sor

is

exposed

in

air.

Fig.

10(d)

presents

representative

reversible
cycles

of

the

gas

response

in

HCHO

(50

ppm),

where

one

can

see
Fig.

8.

(a)

Sensor

resistance

of

Y-free

nanoparticles,

Y-free

and

Y-doped

nanospheres

as

a

function

of

temperature

in

air.

(b)

Response

of

the

sensors

fabricated

with

Y-free
nanoparticles,

Y-free

and

Y-doped

nanospheres

with

various

concentrations

to

HCHO

of

50

ppm

measured

at

temperatures

from

200

◦
C

to

500
◦
C.

Grain

sizes

of

various

ZnO
samples

are

listed

in

Table

1.
60 W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62
Fig.

9.

Response–recovery

characteristic

for

the

sensors

fabricated

with

(a)

Y-free

ZnO

nanoparticles,

(b)

Y-free

and

(c)

Y-doped

ZnO

nanospheres.

Six

types

of

VOC

gases,
CH
4
,

NH
3
,

HCHO,

CH
3
OH,

CO,

and

(CH
3
)

2
OH,

are

chosen

purposely.

The

operation

temperature

for

the

sensors

fabricated

with

the

Y-free

nanomaterials

is

350
◦
C

and

that

for
the

sensor

fabricated

with

Y-doped

nanomaterial

300
◦
C.

The

concentration

of

the

tested

gas

is

fixed

to

be

50

ppm.

(d)

Single-cycle

response

and

recovery

transients

of

the
three

sensors

to

the

HCHO

gas

at

50

ppm.
Fig.

10.

(a)

Gas

response

of

the

sensor

made

of

Y-doped

ZnO

as

a

function

of

HCHO

concentration

operated

at

300
◦
C.

The

inset

highlights

sensing

characteristic

at

low

gas
concentration.

(b)

Gas

response

of

the

sensor

made

of

Y-doped

ZnO

to

the

six

types

of

gases

of

choice.

The

concentration

of

each

gas

is

fixed

to

be

50

ppm

and

the

operating
temperature

is

300
◦
C.

(c)

Single-cycle

response

of

the

Y-doped

ZnO

to

the

HCHO

gas

at

different

concentrations

at

300
◦
C.

(d)

Typical

response

and

recovery

characteristic

of
the

sensor

fabricated

with

the

Y-doped

ZnO

to

the

HCHO

gas

of

50

ppm

at

300
◦
C.

A

few

representative

cycles

are

shown

only,

demonstrating

stability

of

the

Y-doped

sensor.
that

the

response

and

recovery

characteristics

are

reproduced

well
with

no

remarkable

attenuation.

These

imply

that

the

Y-doped

ZnO
nanospheres

can

improve

significantly

gas-sensing

performances.
Such

improvement

cannot

be

realized

for

the

nanoparticles

and
hence

the

Y-doped

nanospheres

shall

hold

substantial

promise

for
the

development

of

a

practical

sensor

device

for

the

on-site

detec-
tion

of

the

harmful

HCHO

gas.
4.

Conclusions
We

have

fabricated

successfully

novel

hollow

and

porous

ZnO
nanospheres

via

the

simple

template-free

hydrothermal

technique
in

organic

solution,

and

investigated

the

gas-sensing

functions.

We
demonstrate

that

the

ratio

of

MEA

in

solution

is

critical

to

mor-
phology

because

it

facilitates

formation

of

metastable

nanoparticle,
restrains

the

growth

of

nanorods,

and

serves

as

fundamen-

tal

building

blocks

for

nanospheres.

Systematic

microstructural
studies

reveal

a

coupling

of

the

Ostwald

ripening

with

the
grain-rotation-induced

grain

coalescence

growth

mechanism
which

is

responsible

for

the

formation

of

the

hollow

and

porous
nanospheres.

Such

hierarchical

nanospheres

possess

a

large

spe-
cific

surface

area

and

can

be

functionalized

with

Y

for

advanced
chemical

gas-sensing

application.

Gas-sensing

performance

to

the
HCHO

is

found

to

be

enhanced

in

the

doped

sample

with

the

Y

con-
centration

of

4%.

This

work

indicates

that

the

Y-doped

hierarchical
structures

represent

an

important

step

forward

to

exploring

the
novel

gas

sensors

for

future

on-site

detection

of

harmful

gases.
Acknowledgements
This

work

was

supported

in

part

by

the

National

Natural

Science
of

China

(51202302)

and

China

Postdoctoral

Science

Foundation
(No.

2012M511904).

Z.W.

appreciates

financial

supports

from

the
Grant-in-Aid

for

Young

Scientists

(A)

(grant

no.

24686069)

and

the
Challenging

Exploratory

Research

(grant

no.

24656376).
W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62 61
References
[1] J.H.

Lee,

Gas

sensors

using

hierarchical

and

hollow

oxide

nanostructures:
overview,

Sensors

and

Actuators

B

140

(2009)

319–336.
[2] H.

Zhang,

Q.

Zhu,

Y.

Zhang,

Y.

Wang,

L.

Zhao,

B.

Yu,

One-pot

synthesis

and

hier-
archical

assembly

of

hollow

Cu
2
O

microspheres

with

nanocrystals-composed
porous

multishell

and

their

gas-sensing

properties,

Advanced

Functional

Mate-
rials

17

(2007)

2766–2771.
[3]

W.

Guo,

T.

Liu,

H.

Zhang,

R.

Sun,

Y.

Chen,

W.

Zeng,

Z.

Wang,

Gas-sensing

per-
formance

enhancement

in

ZnO

nanostructures

by

hierarchical

morphology,
Sensors

and

Actuators

B

166–167

(2012)

492–499.
[4] H.

Liang,

H.

Zhang,

J.

Hu,

Y.

Guo,

L.

Wan,

C.

Bai,

Pt

hollow

nanospheres:

facile
synthesis

and

enhanced

electrocatalysts,

Angewandte

Chemie

International
Edition

43

(2004)

1540–1543.
[5] S.W.

Kim,

M.

Kim,

W.Y.

Lee,

T.

Hyeon,

Fabrication

of

hollow

palladium

spheres

and

their

successful

application

to

the

recyclable

heterogeneous

catalyst

for
suzuki

coupling

reactions,

Journal

of

the

American

Chemical

Society

124

(2002)
7642–7643.
[6]

Y.

Zhu,

T.

Ikoma,

N.

Hanagata,

S.

Kaskel,

Rattle-type

Fe
3
O
4
@SiO
2
hollow

meso-
porous

spheres

as

carriers

for

drug

delivery,

Small

6

(2010)

471–478.
[7]

W.

Wei,

G.

Ma,

G.

Hu,

D.

Yu,

T.

McLeish,

Z.

Su,

Z.

Shen,

Preparation

of

hierarchical
hollow

CaCO3

particles

and

the

application

as

anticancer

drug

carrier,

Journal
of

the

American

Chemical

Society

130

(2008)

15808–15810.
[8]

D.

Grosso,

C.

Boissiere,

C.

Sanchez,

Ultralow-dielectric-constant

optical

thin
films

built

from

magnesium

oxyfluoride

vesicle-like

hollow

nanoparticles,
Nature

Materials

6

(2007)

572–575.
[9]

S.C.

Yang,

D.J.

Yang,

J.

Kim,

J.M.

Hong,

H.G.

Kim,

I.D.

Kim,

H.

Lee,

hollow

TiO
2
hemispheres

obtained

by

colloidal

templating

for

application

in

dye-sensitized
solar

cells,

Advanced

Materials

20

(2008)

1059–1064.
[10] A.

Cao,

J.

Hu,

H.

Liang,

L.

Wan,

Self-assembled

vanadium

pentoxide

(V
2
O
5
)

hol-
low

microspheres

from

nanorods

and

their

application

in

lithium-ion

batteries,
Angewandte

Chemie

International

Edition

44

(2005)

4391–4395.

[11]

X.

Lou,

L.

Archer,

Z.

Yang,

Hollow

micro-nanostructures:

synthesis

and

appli-
cations,

Advanced

Materials

20

(2008)

3987–4019.
[12]

P.

Sun,

W.

Zhao,

Y.

Cao,

Y.

Guan,

Y.

Sun,

G.

Lu,

Porous

SnO
2
hierarchical
nanosheets:

hydrothermal

preparation,

growth

mechanism,

and

gas

sensing
properties,

CrystEngComm

13

(2011)

3718–3724.
[13] B.

Liu,

H.

Zeng,

Hollow

ZnO

microspheres

with

complex

nanobuilding

units,
Chemistry

of

Materials

19

(2007)

5824–5826.
[14]

W.

Guo,

T.

Liu,

W.

Zeng,

D.

Liu,

Y.

Chen,

Z.

Wang,

Gas-sensing

property

improve-
ment

of

ZnO

by

hierarchical

flower-like

architecture,

Materials

Letters

65
(2011)

3384–3387.
[15] Z.

Wang,

J.

Song,

Piezoelectric

nanogenerators

based

on

zinc

oxide

nanowire
arrays,

Science

312

(2006)

242–246.
[16]

Y.

Xia,

P.

Yang,

Y.

Sun,

Y.

Wu,

B.

Mayers,

B.

Gates,

Y.

Yin,

F.

Kim,

Y.

Yan,
One-dimensional

nanostructures:

synthesis

characterization,

and

applications,
Advanced

Materials

15

(2003)

353–389.
[17]

J.

Huang,

A.

Tao,

S.

Connor,

R.

He,

P.

Yang,

A

general

method

for

assembling
single

colloidal

particle

lines,

Nano

Letters

6

(2006)

524–529.
[18] V.

Shinde,

T.

Gujar,

C.

Lokhande,

Enhanced

response

of

porous

ZnO

nanobeads
towards

LPG:

effect

of

Pd

sensitization,

Sensors

and

Actuators

B

123

(2007)
701–706.
[19]

X.

Liu,

J.

Zhang,

L.

Wang,

T.

Yang,

X.

Guo,

S.

Wu,

S.

Wang,

3D

hierarchically
porous

ZnO

structures

and

their

functionalization

by

Au

nanoparticles

for

gas
sensors,

Journal

of

Materials

Chemistry

21

(2011)

349–356.
[20]

Y.

Zeng,

Z.

Lou,

L.

Wang,

B.

Zou,

T.

Zhang,

W.

Zheng,

G.

Zou,

Enhanced

ammonia
sensing

performances

of

Pd-sensitized

flowerlike

ZnO

nanostructure,

Sensors
and

Actuators

B

156

(2011)

395–400.
[21]

X.

Xue,

Z.

Chen,

L.

Xing,

C.

Ma,

Y.

Chen,

T.

Wang,

One-Step

Synthesis,

Gas-Sensing
Characteristics

of

Uniformly

Loaded

Pt@SnO
2
Nanorods,

The

Journal

of

Physical
Chemistry

C

114

(2010)

18607–18611.
[22] L.

Peng,

P.

Qin,

Q.

Zeng,

H.

Song,

M.

Lei,

J.N.

Mwangi,

D.

Wang,

T.

Xie,
Improvement

of

formaldehyde

sensitivity

of

ZnO

nanorods

by

modifying

with
Ru(dcbpy)
2
(NCS),

Sensors

and

Actuators

B

160

(2011)

39–45.
[23]

N.

Han,

X.

Wu,

D.

Zhang,

G.

Shen,

H.

Liu,

Y.

Chen,

CdO

activated

Sn-doped

ZnO
for

highly

sensitive

selective

and

stable

formaldehyde,

Sensors

and

Actuators
B

152

(2011)

324–329.
[24]

S.

Kim,

M.

Kim,

W.

Lee,

T.

Hyeon,

Fabrication

of

hollow

palladium

spheres
and

their

successful

application

as

the

recyclable

heterogeneous

catalyst

for
suzuki

coupling

reactions,

Journal

of

the

American

Chemical

Society

124

(2002)
7642–7643.
[25]

F.

Caruso,

R.

Caruso,

H.

Mohwald,

Nanoengineering

of

inorganic

and

hybrid
hollow

spheres

by

colloidal

templating,

Science

282

(1998)

1111–1114.
[26]

M.

Yang,

J.

Ma,

C.

Zhang,

Z.

Yang,

Y.

Lu,

Janus

colloids

derived

by

bi-phasic

graft-

ing

at

pickering

emulsion

interface,

Angewandte

Chemie

International

Edition
44

(2005)

6727–6730.
[27]

J.

Gao,

B.

Zhang,

X.

Zhang,

B.

Xu,

Magnetic-dipolar-interaction-induced

self-
assembly

affords

wires

of

hollow

nanocrystals

of

cobalt

selenide,

Angewandte
Chemie

International

Edition

45

(2006)

1220–1223.
[28]

C.I.

Zoldesi,

A.

Imhof,

Synthesis

of

monodisperse

colloidal

spheres

capsules,
and

microballoons

by

emulsion

templating,

Advanced

Materials

17

(2005)
924–928.
[29] Q.

Peng,

Y.

Dong,

Y.

Li,

Cu
2
O

hollow

spheres

with

a

single-crystalline

shell

wall,
Angewandte

Chemie

International

Edition

42

(2003)

3027–3030.
[30]

R.

Penn,

J.

Banfield,

Imperfect

oriented

attachment:

dislocation

generation

in
defect-free

nanocrystal,

Science

281

(1998)

969–971.
[31]

D.

Yu,

X.

Sun,

J.

Zou,

Z.

Wang,

F.

Wang,

K.

Tang,

Oriented

assembly

of

Fe
3
O
4
nanoparticles

into

monodisperse

hollow

single-crystal

microspheres,

The

Jour-
nal

of

Physical

Chemistry

B

110

(2006)

21667–21671.
[32]

Y.

Chang,

J.

Teo,

H.

Zeng,

Formation

of

colloidal

CuO

nanocrystallites

and
their

spherical

aggregation

and

reductive

transformation

to

hollow

Cu
2
O
nanospheres,

Langmuir

21

(2005)

1074–1079.
[33]

B.

Liu,

H.

Zeng,

Symmetric

and

asymmetric

Ostwald

ripening

in

the
fabrication

of

homogeneous

core–shell

semiconductors,

Small

1

(2005)
566–571.
[34] B.P.

Jia,

L.

Gao,

Morphological

transformation

of

Fe
3

O
4
spherical

aggregates
from

solid

to

hollow

and

their

self-assembly

under

an

external

magnetic

field,
The

Journal

of

Physical

Chemistry

C

112

(2008)

666–671.
[35]

X.

Liang,

X.

Wang,

Y.

Zhuang,

B.

Xu,

S.

Kuang,

Y.

Li,

Formation

of

CeO
2
-

ZrO
2
solid
solution

nanocages

with

controllable

structures

via

Kirkendall

effect,

Journal
of

the

American

Chemical

Society

130

(2008)

2736–2737.
[36] Y.

Yin,

R.

Rioux,

C.

Erdonmez,

S.

Hughes,

G.

Somorjai,

A.

Alivisatos,

Formation
of

hollow

nanocrystals

through

the

nanoscale

Kirkendall

effect,

Science

304
(2004)

711–714.
[37]

Y.

Xiong,

B.

Wiley,

J.

Chen,

Z.

Li,

Y.

Yin,

Y.

Xia,

Corrosion-based

synthesis

of
single-crystal

Pd

nanoboxes

and

nanocages

and

their

surface

plasmon

proper-
ties,

Angewandte

Chemie

International

Edition

44

(2005)

7913–7917.
[38]

X.

Lou,

Y.

Wang,

C.

Yuan,

J.

Lee,

L.

Archer,

Template-free

synthesis

of

SnO
2
hollow

nanostructures

with

high

lithium

storage

capacity,

Advanced

Materials
18

(2006)

2325–2329.
[39]

F.

Li,

Y.

Ding,

P.

Gao,

X.

Xin,

Z.

Wang,

Single-crystal

hexagonal

disks

and

rings
of

ZnO:

low-temperature

large-scale

synthesis

and

growth

mechanism,

Ange-
wandte

Chemie

International

Edition

43

(2004)

5238–5242.
[40]

Z.

Miao,

Y.

Wu,

X.

Zhang,

Z.

Liu,

B.

Han,

K.

Ding,

G.

An,

Large-scale

production

of
self-assembled

SnO
2
nanospheres

and

their

application

in

high-performance
chemiluminescence

sensors

for

hydrogen

sulfide

gas,

Journal

of

Materials
Chemistry

17

(2007)

1791–1796.
[41] S.

Ho,

A.

Wong,

G.

Ho,

Controllable

porosity

of

monodispersed

tin

oxide
nanospheres

via

an

additive-free

chemical

route,

Crystal

Growth

and

Design
9

(2009)

732–736.
[42]

G.

Korotcenkov,

V.

Brinzari,

I.

Boris,

(Cu,

Fe,

Co

or

Ni)-doped

tin

dioxide

films
deposited

by

spray

pyrolysis:

doping

influence

on

film

morphology,

Journal

of

Materials

Science

43

(2008)

2761–2770.
[43]

X.

Sun,

X.

Qiu,

L.

Li,

G.

Li,

ZnO

twin-cones:

synthesis

photoluminescence,

and
catalytic

decomposition

of

ammonium

perchlorate,

Inorganic

Chemistry

47
(2008)

4146–4152.
[44]

W.

Li,

E.

Shi,

W.

Zhong,

Z.

Yin,

Growth

mechanism

and

growth

habit

of

oxide
crystals,

Crystal

Growth

and

Design

203

(1999)

186–196.
[45]

X.

Wang,

Q.

Zhang,

Q.

Wan,

G.

Dai,

C.

Zhou,

B.

Zou,

Controllable

ZnO
architectures

by

ethanolamine-assisted

hydrothermal

reaction

for

enhanced
photocatalytic

activity,

The

Journal

of

Physical

Chemistry

C

115

(2011)
2769–2775.
[46]

L.

Xu,

Y.

Hu,

C.

Pelligra,

C.

Chen,

L.

Jin,

H.

Huang,

S.

Sithambaram,

M.

Aindow,

R.
Joesten,

S.

Suib,

ZnO

with

different

morphologies

synthesized

by

solvothermal
methods

for

enhanced

photocatalytic

activity,

Chemistry

of

Materials

21

(2009)
2875–2885.
[47]

P.

Hu,

X.

Zhang,

N.

Han,

W.

Xiang,

Y.

Cao,

F.

Yuan,

Solution-controlled

self-
assembly

of

ZnO

nanorods

into

hollow

microspheres,

Crystal

Growth

and
Design

11

(2011)

1520–1526.
[48]

M.

Niederberger,

H.

Cölfen,

Oriented

attachment

and

mesocrystals:

non-
classical

crystallization

mechanisms

based

on

nanoparticle

assembly,

Physical
Chemistry

Chemical

Physics

8

(2006)

3271–3287.
[49]

E.R.

Leite,

T.R.

Giraldi,

F.M.

Pontes,

E.

Longo,

A.

Beltrán,

J.

Andrés,

Crystal

growth
in

colloidal

tin

oxide

nanocrystals

induced

by

coalescence

at

room

temperature,
Applied

Physics

Letters

83

(2003)

1566–1568.
[50]

N.

Carre
˜
no,

A.

Maciel,

E.

Leite,

P.

Lisboa-Filho,

E.

Longo,

A.

Valentini,

L.

Probst,

C.
Paiva-Santos,

W.

Schreiner,

The

influence

of

cation

segregation

on

the

methanol
decomposition

on

nanostructured

SnO
2
,

Sensors

and

Actuators

B

86

(2002)
185–192.
[51] W.

Zeng,

T.

Liu,

Z.

Wang,

Enhanced

gas

sensing

properties

by

SnO
2
nanosphere
functionalized

TiO
2
nanobelts,

Journal

of

Materials

Chemistry

22

(2012)
3544–3548.
[52]

C.

Kung,

S.

Young,

H.

Chen,

M.

Kao,

L.

Horng,

Y.

Shih,

C.

Lin,

T.

Lin,

C.

Ou,
Influence

of

Y-doped

induced

defects

on

the

optical

and

magnetic

properties
of

ZnO

nanorod

arrays

prepared

by

low-temperature

hydrothermal

process,
Nanoscale

Research

Letters

7

(2012)

372–377.
[53]

A.

Janotti,

C.G.

Van

de

Walle,

Fundamentals

of

zinc

oxide

as

a

semiconductor,
Reports

on

Progress

in

Physics

72

(2009)

126501–126529.
[54]

A.

Gurlo,

Interplay

between

O
2
and

SnO
2

:

oxygen

ionosorption

and

spectro-
scopic

evidence

for

adsorbed

oxygen,

ChemPhysChem

7

(2006)

2041–2052.
[55]

M.

Batzill,

U.

Diebold,

Surface

studies

of

gas

sensing

metal

oxides,

Physical
Chemistry

Chemical

Physics

9

(2007)

2307–2318.
[56] M.

Kung,

R.

Davis,

H.

Kung,

Understanding

Au-catalyzed

low-temperature
CO

oxidation,

The

Journal

of

Physical

Chemistry

C

111

(2007)
11767–11775.
[57]

R.

Joshi,

Q.

Hu,

F.

Am,

N.

Joshi,

A.

Kumar,

Au

decorated

zinc

oxide

nanowires

for
CO

sensing,

The

Journal

of

Physical

Chemistry

C

113

(2009)

16199–16202.
[58]

Q.

Xiang,

G.

Meng,

H.

Zhao,

Y.

Zhang,

H.

Li,

W.

Ma,

J.

Xu,

Au

nanoparticle

modi-
fied

WO
3
nanorods

with

their

enhanced

properties

for

photocatalysis

and

gas
sensing,

The

Journal

of

Physical

Chemistry

C

114

(2010)

2049–2055.
[59] S.

Morrison,

Selectivity

in

semiconductor

gas

sensors,

Sensors

and

Actuators

B
12

(1987)

425–440.
62 W.

Guo

et

al.

/

Sensors

and

Actuators

B

178 (2013) 53–

62
Biographies
Weiwei

Guo

is

currently

a

PhD

candidate

at

the

College

of

Materials

Science

and
Engineering,

Chongqing

University

in

China.

He

is

now

engaged

in

the

synthesis
and

characterization

of

the

semiconducting

materials

and

in

the

investigation

of
their

gas

sensing

properties.
Tianmo

Liu

is

a

professor

of

College

of

Materials

Science

and

Engineering

at
Chongqing

University

in

China

since

2001.

He

received

Dr.

Eng.

from

Department
of

Solid

Mechanics,

Chongqing

University

in

1999.

His

current

research

interest
involves

functional

materials

for

gas

sensors

and

magnesium

alloys.

He

is

now

also
holding

a

group

leader

position

at

the

National

Engineering

Research

Center

for
Magnesium

Alloys

at

Chongqing

University.

Rong

Sun

is

currently

a

PhD

candidate

in

the

Institute

of

Engineering

Innovation,
The

University

of

Tokyo

in

Japan.

Her

research

interest

involves

the

characterization
of

materials

using

advanced

transmission

electron

microscopy.
Yong

Chen

is

currently

a

PhD

candidate

at

the

College

of

Materials

Science

and
Engineering,

Chongqing

University

in

China,

and

also

an

exchange

student

at

Tohoku
University

in

Japan

since

2011.

He

is

now

engaged

in

fabricating

nanomaterials

and
in

characterization

using

advanced

transmission

electron

microscopy.
Wen

Zeng

received

his

PhD

degree

in

material

Science

from

Chongqing

University

in
China.

He

is

currently

a

lecture

at

the

College

of

Materials

Science

and

Engineering,
Chongqing

University.

He

is

focusing

on

synthesis

of

low-dimensional

functional
materials,

on

fabrication

of

semiconducting

sensors

and

on

first-principles

calcula-
tions.

Zhongchang

Wang

is

currently

an

assistant

professor

at

the

WPI

Research

Center,
Advanced

Institute

for

Materials

Research,

Tohoku

University

in

Japan.

He

received
his

master

degree

in

2004

from

Chongqing

University

in

China

and

PhD

in

2008
from

the

University

of

Tokyo

in

Japan.

He

is

now

mainly

focusing

on

gas-sensing
materials,

interfaces,

grain

boundaries,

dislocations

in

oxides,

and

quantum

electron
transport

by

combining

the

state-of-the-art

transmission

electron

microscopy

with
the

first-principles

calculations.