07_Martin

COST Workshop Barcelona June 2013

Presenter NameFunction in the Action (MC Member/Substitute, WG Member, Participant, External Expert) / Email addressAffiliation / CoPhil MartinNon-COST Partner

[email protected]

CSIRO Australia

CARBON BASED NANOMATERIALSFOR GAS SENSING

CSIRO MATERIAL SCIENCE AND ENGINEERING

Phil Martin, Lakshman Randeniya, Avi Bendavid

Lindfield, Sydney NSW

Australia

CarbonCarbon--Based Nanomaterials for Gas SensingBased Nanomaterials for Gas Sensing

Workshop Barcelona 20 June 2013Workshop Barcelona 20 June 2013


Vertically aligned MWCNTsVertically aligned MWCNTs

1. MWCNT yarn sensorsMWCNT yarn sensors

3. GrapheneGraphene

2. SWCNT sensorsSWCNT sensors


Carbon nanotube yarns for environmental gas

sensing applications

Vertically aligned

MWCNTs

Carbon nanotube yarns


Advantages of carbon nano tubes

Gas adsorption capability, Large specific surface area, High sensitivity,Low operating temperature

Advantages of carbon nanotube yarns

Functional surface availableMechanically robust structuresEasy to functionalise and handleCan be used as a simple chemiresistor

Growth of Vertically Aligned MWCNTs


95% Helium &

5% Acetylene

@580 sccm

Fe catalyst

Al2O3 layer

SiO2 buffer layer

Si substrate

Aligned CNT

Forest

Structure for ethylene pre-cursor

Quartz reactor tube

Silicon wafer with 5 nm Fe

catalyst layer

Furnace @~ 680 centigrade


Schematic of web formation

~20 m

(undensified)

sensor

CNT Yarn Properties


Data for CNT YarnsElectrical Conductivity (S/cm):

Singles yarn ~300

Density/(g/cm3):

Singles yarn ~0.8

Modulus/GPa: (Singles 24 twist)

~100

Strength/GPa: (Singles 24 twist)

~0.92

Toughness/(J/g):

Singles Yarn ~14

Twofold Yarn ~20

No creep:

>20 h at 6% strain (~50% breaking strain)

Knots do not degrade tensile strength:

Retain flexibility/strength :

after heating in air at ~450C

when immersed in liquid N2

Comments and comparisons Electrical Conductivity/(S/cm):

Graphite CF: ~167 3333

Density/(g/cm3):

Graphite CF ~1.8

Modulus/GPa:

Graphite CF ~300

Strength/GPa:

Graphite CF ~3

Toughness/(J/g)

Graphite CF ~12

Solution-spun SWNT/PVAyarns ~600

Knots degrade tensile strengthsof most textile fibres


Keithley Picoammeter

Mass Flow

Controllers

Gas Cell (mini-16 CF, 60 cm3)

Dry N2

Test Gas (H2, NH3, CH4, NO2)

CNTY sensor

Gas exhaust

Gas Sensing

Cell

Gas inlet

Functionalization of the Yarns


P-type semiconductor (500 600 Scm-1)

Surface treatments (create active sites)

Plasma and acid methods

Introduce various functional groups

(Carboxylic, N-, S-), and defects (oxygenated

vacancies, pentagon-heptagon pairs etc.)

Incorporation of metal nano clusters

(improves specificity)

Sputtering techniques

Electrochemical techniques (self-

fuelled electrodeposition, SFED)

Pulsed-DC PECVD system used for

surface functionalisation


Decoration with Au enhances the sensitivity (factor of 5 8)Mechanism is unclear; could be related to the modulation of the Schottky barrier

Ammonia Sensing with Au-decorated CNT yarns

CNTY Surface Treatment

12 | COST Workshop Barcelona June 2013

NH3 : Stability and recovery dependent on surface treatment

0 10000 20000 30000

240

270

300

0 3000 6000 9000 12000222

224

226

228

230

232

Acid-treated

Ar/H2-plasma-treated

NH3 off

10 min

R

e

s

i

s

t

a

n

c

e

(

o

h

m

)

Time (s)

5 min

NH3 on

Plasma

treated

NH3

500 ppm

50 ppm



0 500 1000 1500 2000 2500

540

550

560

570

580

500 ppm

100 ppm

50 ppm

on

R

e

s

i

s

t

a

n

c

e

(

o

h

m

)

Time (s)

10 ppmoff

0 6000 12000 18000380

384

388

392

396

400

On500 ppb

1 ppm5 ppm

50 ppm

R

e

s

i

s

t

a

n

c

e

(

o

h

m

)

Time (s)

500 ppm

Off

Plasma-treated sample

Quick response

Good recovery

Low concentrations

Reproducible

Acid-treated sample

Quick response

Long or no recovery

Low concentrations

Non-reproducible

NH3



Particle sizes 10 20 nm

Sparsely distributed

Acid-treated sample Plasma-treated sample

Dual populations

Sparse 10-20 nm

Dense and uniform 2 -3 nmDiffusion out of smaller particles easier ??

Hydrogen detection using Pd and Pd-Pt coated yarns

(low concentrations)

COST Workshop Barcelona June 201315 |

Addition of Pt layer enables the detection of lower concentrations

0 2000 4000 6000 8000

0.0

0.2

0.4

0.6

0.8

0 2 4 6 8 10 120.0

0.2

0.4

0.6

0.8

1.0

R

/

R

0

*

1

0

0

Time (s)

20 ppm

40 ppm 100 ppm

[H2 ]1/2

(ppm)1/2

R

/

R

0

*

1

0

0

0 2000 4000 6000 8000 10000

0.0

0.2

0.4

0.6

0.8

0 2 4 6 8 10 120.0

0.3

0.6

R

/

R

o

*

1

0

0

100 ppm9 ppm

Time (s)

ON

OFF

5 ppm

R

/

R

0

*

1

0

0

[H2]1/2

(ppm1/2)

Pd Only Pd-Pt layered structureH2

CNTY: Pd and Pd-Pt nanostructures


Pd: Sparse Pd populations Pd-Pt: Dense Pd populations

CNTY:Pd and Pt distributions on yarn


PtPt

PdPd

Pt is dispersed particulates on larger Pd grainsPt is dispersed particulates on larger Pd grains

Hydrogen detection using Pd-Pt-coated

(higher concentrations in air)


Excellent response times and recovery timesExcellent response times and recovery times

0 500 1000 1500 2000 2500

0

1

2

3

4

1

0

0

*

R

/

R

0

Time (s)

.04%

1%

2%3% 4%

0.5%

0.1%

CNTY

Hydrogen detection using Pd and Pd-Pt-coated

CNT yarns


Work function of Pd and lattice parameters are sensitive to hydrogen

Room-temperature detection

Concentrations from 20 ppm to 2 % (20,000 ppm) are detectable with Pd-

MWCNT-yarn chemiresistor

Concentrations from 5 ppm to 2 % (20,000 ppm) are detectable with Pt-Pd-

MWCNT-yarn chemiresistor

Excellent response times and recovery times are obtained

Flexible, robust and simple chemiresistor systems

CNTY: Summary


CNT yarn has potential to be used as gas sensors

Robustness

Easy handling

Easy to functionalise

Room-temperature operation

SFED method is a quick and reliable method for incorporating nanocrystaline metals (onto semi-conducting and conducting substrates)

Particle size has a significant impact on the recovery of NH3 and hydrogen sensors

Addition of Pt on Pd increases sensitivity at low concentrations

SWCNT


SEM diagram showing thinly-spread bundles of

CNT on a mesoporous alumina membrane. The

resistance ( ~ 3 mm x 5 mm) ~ 1.6 M under

ambient conditions.

Raman shift for original (without ultrasonic

treatment) and ultrasonic-treated

samples. The increase in D peak intensity

with respect to that of G peak is attributed

to the increase in defects caused by

ultrasonic treatment.


Keithley Picoammeter

Mass Flow

Controllers

Gas Cell (mini-16 CF, 60 cm3)

Dry N2

Test Gas (H2, NH3, CH4, NO2)

Graphene/SWCNT

sensor

Gas exhaust

Gas Sensing

Cell

Gas inlet

Water

bubbler

SWCNT


Two modes of response of SWCNT/Al2O3 chemiresistors to water vapour.

(a) response change with time when a chemiresistor equilibrated under ambient conditions is introduced into the chamber under dry air flow. The

resistance first drops due to drop in humidity (mode 2 response) and then rises slowly as the water molecules come off (mode 1 response).

(b) Mode 2 responses of the chemiresistor to injections of water (by diverting a fraction of buffer gas flow through an enclosed bubbler) which

temporarily raise the humidity in the chamber to the levels marked.

Mode Mode --22--FastFast

Water molecules donate Water molecules donate

electrons to pelectrons to p--type type

conductor. When removed conductor. When removed

conductivity increasesconductivity increases

Mode Mode --11--SlowSlow

More tightly bound water molecules donate holes More tightly bound water molecules donate holes

When slowly removed conductivity decreasesWhen slowly removed conductivity decreases

SWCNT


Scheme used for NO2 measurements using water-assisted recovery

NO2

Humidity raised (50-75%)

at this point

Vapour off and system

returns to the baseline

c

o

n

d

u

c

t

a

n

c

e

SWCNT: substrate effect


recovery time

tR (= t1 + t2)

~ 10 min

AlAl22OO33

Free standing Free standing

filmfilm

NO2

55 mins 5 ppm

Dry air only

No residual

effect on

recovery

15 mins

SWCNT

on


(a) repeated detection of 5 ppm NO2 using SWCNT deposited on Si wafer with natural oxide layer

(SWCNT/SiO2 chemiresistor). Chemiresistor is exposed to NO2 for 5 min Water-enriched airflow is

maintained for 5 min in each case before reverting to dry airflow. (b) Measurement of different

concentrations of NO2 with 3-min exposures in each case followed by water-assisted recovery. A

SWCNT/Al2O3 chemiresistor was used. -Inset to (b) shows a response versus log concentration

curve.

SiO2

Al2O3

NO25 ppm

SWCNT

on


Detection of ammonia using a SWCNT/Al2O3- chemiresistor; the response is calculated as a percent increase in conductance. (a) Repeated

detection of 50 ppm with water-assisted recovery time (tR) of about 15 minutes when humidity is raised to ~ 75 %; (b) partial and very

slow recovery in dry air; (c) detection of 5 ppm and 500 ppb using the same method. In all cases 5-minute exposure to ammonia used

except for the detection of 500 ppb where a longer exposure time was used.

NH3

Al2O3

SWCNT: humidity effect on different substrates


Lower humidity to achieve recovery.

50% of the buffer gas was diverted

through water bubbler. Relative

humidity in the chamber rose to 52 %

Detection of 5 ppm NO2 using

SWCNT on amorphous TiO2substrate

TiO2Al2O3

NO2

Graphene


SEM images of sections

of chemiresistors used

for gas detection

(a), (c) preparations in

pure water

(b), (d) preparations in

nitric acid

Whatman Anopore

membranes (~ 50 nm

diameter pores) were

used as substrate.

H2OHNO3

Graphene


TEM images of (a) graphene flakes obtained from pure water dispersion (b) graphene

nanomesh obtained from nitric acid dispersion. Schematics of NO2 adsorption on

graphene flakes and graphene nanomesh flakes are shown below the corresponding

TEM images.

water HNO3 nn-- typetypepp-- typetype

Graphene


TEM and HRTEM images of

graphene obtained from HNO3acid dispersions and schematic

interpretation of structures. (a) A

section of regular TEM and (b) to

(d) are HRTEM images of areas

marked as 1, 2 and 3. (e) - (g) are

schematic representations of the

corresponding HRTEM images.

Graphene sensor: enhanced recovery using water vapour


0 1000 2000 3000 4000 5000 6000 7000-10

-5

0

5

10

15

20

0 1000 2000 3000 4000 5000 6000 7000

0

5

10

Time (s)

H2O off

1

0

0

*

G

/

G

0

NO2 on

NO2 off, H2O onb

NO2 off

NO2 on

a 5 mins NO2 5 ppmVery slow

recovery

5 mins NO2 x 5

+ H2O purging

Recovery

dramatically

reduced

NO2

58% RH

1% RH

pp--type responsetype response

Graphene: HNO3-treated


0 1000 2000 3000 4000 5000 6000-40

-30

-20

-10

0

10

20

NO2 on1

0

0

*

G

/

G

0

Time (s)

NO2 on

NO2 off, H2O on

H2O off

NO2 off

baseline

The response of HNO3-treated graphene chemiresistor to 5 ppm of NO2 (5 min exposure). The conductance decreases in

response to NO2 exposure confirming that acid-treated graphene behaves as an n-type semiconductor. There is an upward drift

in baseline (marked by a dashed line) after each run. Four consecutive measurements are shown.

5 ppm NO2

nn--type responsetype response

Graphene: HNO3-treated


0 1000 2000 3000 4000

-30

-20

-10

0

10

0 2000 4000 6000 8000 10000-12

-8

-4

0

4

8

Time (s)

EtOH off

NH3 off,EtOH on

1

0

0

*

G

/

G

0

NH3 on

b

NH3 off

NH3 on

aNH3 off, H2O on

H2O off

NH3 on

Effect of H2O

Effect of EtOH

NH3

1 ppm 10 min

5 ppm

Dry air only

Graphene


A schematic of a plausible mechanism of NO2removal from graphene flakes influenced by

water vapour. The strong adsorption of NO2on graphene is due to overlap of molecular

states (M) and substrate defects states (S)

near the Fermi level (EF ) with the conduction

band (CB) of graphene. When water

molecules are introduced, the electrostatic

forces causes a movement of substrate

defects states and the strong overlap

between NO2 orbitals with that of graphene

is lost.

Wehling et al Appl. Phys. Lett. (2008) 93, 202110

Graphene


1000 1500 2000

5000

10000

15000 HNO3-treated graphene layer film

I

n

t

e

n

s

i

t

y

(

a

r

b

.

u

n

i

t

s

)

Raman Shift (cm-1)

15851342

1579

1346

Non-treated graphene layer film

Raman spectra for graphene. The increase in ID/IG ratio in the nitric-acid-

treated samples is further evidence for increased defects in the structure

Summary


MW-Carbon nanotube yarns: simple robust sensor

with enhanced response when acid-etched and

coated with nanoparticles

SWCNT recovery can be dramatically accelerated

by using water vapour purging

Graphene response is enhanced by acid treatment

to form nanomesh structures. Water vapour

accelerates recovery

CSIRO MATERIAL SCIENCE AND ENGINEERING

Dr Phil Martin

Telephone +61 2 9413 7126email [email protected]

Thank You

Documents

07_Martin