Sangil Kim1,2, Francesco Fornasiero1, Michael Stadermann1, Alexander Chernov1, Hyung Gyu Park1, Jung Bin In3, Ji Zang5, David Sholl5, Michael Colvin4, Aleksandr Noy1,4, Olgica Bakajin,1,2 and Costas P. Grigoropoulos3

1 Physical and Life Sciences, LLNL; 2 NSF Center for Biophotonics, UC Davis; 3Mechanical Engineering, UC Berkeley; 4School of Natural Sciences, UC Merced, 5Chemical and Biochemical Engineering, Georgia Tech

Gated Transport through Carbon Nanotube MembranesNIRT CBET-0709090

CA zzm nA C

CARBON NANOTUBE MEMBRANE:A NANOFLUIDIC PLATFORM

Unique surface properties of carbon nanotubes enable very rapid and very efficient transport of gases and liquids

We need to understand: Fundamental physics of transport through these

nanoscale channelsMembrane selectivity and rejection propertiesFabrication issues associated with making CNT

membranes with desired geometry and propertiesControl of transport through CNT membranes:

Are artificial ion channels possible?

ION EXCLUSION

Si

DWCNT / Si3N4

Free standing membrane Highly aligned DWCNTs Inner diameter

~ 1.6 nm LPCVD Si3N4 pinhole-free

matrix

MULTI-COMPONENT GAS PERMEATION SYSTEM

BINARY GAS PERMEATION

30 40 50 60 70 800.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

Ideal selectivity

298 K

263 KKnudsen Separation

Se

lectivity

(CH 4/N

2)

CH4 in Feed (%)

20 30 40 50 60 70 800.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Ideal Selectivity

298 K

263 K

Knudsen Separation

CO2 in Feed (%)

Se

lectivity

(CO 2/N

2)

CH4/N2 and CO2/N2

At 263 K, the separation factor increased because of increased gas solubility at lower temperature.Comparison with atomistic simulations (CH4/N2)

CONCLUSIONS

Part of the work at LLNL was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

PUBLICATIONS

• Holt et. al., Science, 312, 1034 (2006)• Noy et. al., Nano Today, 2, 22 (2007) • Fornasiero et. al. Proc. Natl. Acad. Sci

USA, 105, 17217 (2008)• Stadermann et. al., Nano Letters, in

revision (2008)

GROWTH OF ALIGNED NANOTUBE ARRAYS

• Selectivity ≡ A/B= [ yA/(yB) ]/[ xA/(xB) ]=[ yA/(1-yA) ]/[ xA/(1-xA) ]

where x : the mole fractions of gas species at the feed side y : the mole fractions of gas species at the permeate side

0.3 0.4 0.5 0.6 0.7 0.81

2

3

4

5

6

7

Pf=1.5 Bar, P

p=1 Bar

Se

lect

ivity

(C

H4/N

2)

CH4 In Feed (%)

(40,40), 298K

(40,40), 263K

(20,20), 298K

(20,20), 263K

(10,10), 298K

(10,10), 263K

0.0 0.2 0.4 0.6 0.8 1.02.0

2.2

2.4

2.6

2.8

3.0

3.2

3.4

3.6

3.8

4.0

4.2

Se

lect

ivity

(C

H4/N

2)

# ratio of (10,10) SWNT

298K 263K

Nanotube membrane made of (10,10) and (40,40) SWNT

Pf=1.5 Bar, P

p=1 Bar

Smaller tube has higher separation factor for CH4/N2. Polydisperse of tube size in CNT membrane affects the separation factor.

SINGLE GAS PERMEATION

Strongly absorbing gas species (CO2, CH4, and C2H4) deviated from the scaled Knudsen permeance Weakly absorbing gas species (He, N2, Ar, and SF6) did not show the deviation.

0.0 0.1 0.2 0.3 0.4 0.5 0.60

1

2

3

4

5

6

7

(b)

Ar

CH4

C2H

4

SF6

N2

CO2

He

Pe

rme

an

ce

(x1

0-5,

mo

l.m2 .s

ec-1

.Pa-1

)

M-1/2(mol1/2g-1/2)

P263K P293K

-20%

0%

20%

40%

60%

80%

100%

3.0 2.0 1.0 1.0 0.5 0.5

Z-/Z+

% R

ejec

tion

CationAnionDonnan

K3Fe(CN)6

K2SO4

CaSO4 KCl

CaCl2Ru(bipy)3Cl2

Electrostatic interactions dominate the ion rejection mechanism The largest ion in this series, Ru(bipy)3Cl2, permeates freely through

the membrane suggesting that size effects are less important

Rejection declines at larger salt solution concentrations

Rejection ~ constant when the Debye length is >> CNT diameter

Debye length dependence

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18 20

lD [nm]

Donnan 1:3

Donnan 1:1

Concentration dependence

0

20

40

60

80

100

0.1 1 10 100

concentration [mM]

% R

eje

cti

on

co

eff

icie

nt

Anion

Cation

XY (Scatter)3XY (Scatter)4

K3Fe(CN)6KCl

K3Fe(CN)6

KCl

CNT membrane

Pressure

Feed (salt solution)

Permeate

0

20

40

60

80

100

Re

jec

tio

n [

%]

pH=7.2

Cation

Anion

0

20

40

60

80

100

Re

jec

tio

n [

%]

pH=3.8

Cation

Anion 6.7 A

8.1

A

1permeate

feed

cR

c

Ion rejection coefficient:

• CNT growth rates exhibit a non-monotonic dependence on total pressure and humidity. Optimal process pressure and water concentration produce growth rate of ~30m/min.

• Nanotube growth rate remains essentially constant until growth reaches an abrupt and irreversible termination.

• We developed a model that predicts termination kinetics

Iijima’s modelPoisoning model

•VA-CNT arrays grow from catalytic decomposition of carbon precursor, C2H4, over nanoscale Fe catalyst

KINETICS OF CARBON NANOTUBE ARRAY GROWTH

K+

CNT Aquaporin K+ channelGas transport in CNTs and other nanoporous materials

CNT MEMBRANE

• Carbon nanotube membranes support high flux transport of liquids and gases

• Nanotube growth kinetics studies allowed high-yield, high-quality growth of aligned nanotube arrays

• CNT membranes show good ion rejection characteristics• Ion rejection mechanism is based on electrostatic repulsion and

follows Donnan model predictions• Strongly absorbing gas species deviated from Knudsen

permeance due to preferential interactions with CNTs side walls. • At low temperature gas separation factor increased because of

increased gas solubility; overall gas separation factors are still lower than necessary for practical gas separation