Growth of Single-Walled Carbon Nanotubes on Surface with Controlled Structures
on Surface with Controlled Structures
Prof. Jin ZHANG
Center for nanochemistry
College of Chemistry and Molecular Engineering P ki U i it B iji 100871 P R CHINA Peking University, Beijing 100871, P. R. CHINA
Email: [email protected]
The University of Tokyo, 1st Nov, 2010, Tokyo
Introduction of Single-walled Carbon Nanotubes
C
Chh = n= naa11 + m+ maa22 (n,m)(n,m)
Zigzag Nanotube Armchair Nanotube
/ ) (
3a n2 m2 nm 1/2 dt CC
m n
m 2 tan 1 3
tor semiconduc
metal 1
3 3
p
m p n
Nanotube transistor Nanotube RAM Flat panel display Nanotube interconnect
N o ube s s o N o ube p e d sp y Na otube te co ect
A Scalable CVD Synthesis of High-Purity SWNTs with
P M O S t M t i l
Porous MgO as Support Materials
2000 2500
it)
SEM
TEM
1500
ty (arb. uni
500 1000
Intensit
400 800 1200 1600
0
Raman shift (cm-1)
Support:MgO;Catalysts:Fe;Carbon Source:CH4
J. Zhang et. al. , J Mater Chem, 12(4): 1179-1183, 2002; Carbon, 40(12): 2282-2284, 2002;Carbon, 40(14): 2693-2698, 2002
Surface Growth of SWNTs by CVD
Catalyst
Growth Process:
C SWNTs
Nanoparticley
C
C
C
SWNTs
Substrate Substrate Substrate
Q ti
Questions:
1. Growing SWNTs on Surface Directly with Controlled Density, Position and Orientation
and Orientation
2. Growing SWNTs on Surface with Controlled Diameter
3 Growing SWNTs on Surface with Controlled Metallic and Semi 3. Growing SWNTs on Surface with Controlled Metallic and Semi-
conducting Properties
4. Growing SWNTs on Surface with Controlled Chirality 4. Growing SWNTs on Surface with Controlled Chirality
Controlled CVD Growth of SWNTs on Surface
a
The first time The second time
2m
J Ph Ch B 2005 109 10946
J Ph Ch B 2004 108 12665 J. Phys. Chem. B, 2005, 109, 10946 J Ph Ch C 2008 112 8319 J. Phys. Chem. B. 2004, 108, 12665
84°
J. Phys. Chem. C, 2008, 112, 8319
J. Am. Chem. Soc. 2005, 127, 8268.
J. Phys. Chem. B, 109 (2005) 2657-2665 J. Am. Chem. Soc. 2005,127,17156
130°
J. Phys. Chem. B, 109 (2005) 2657 2665
Challenges for the Application of Carbon Nanotubes in Future Device
1)How to achieve a structure-controlled synthesis of nanotubes ?
Diameter
Lattice geometry (armchair zigzag chirality)
Lattice geometry (armchair, zigzag, chirality)
Semiconduting or Metallic Nanotubes 2)How to fabricate a desired device structure ?
Controlled surface growth M i l ti
Manipulation
3)What architecture should the nanotube device have ?
4)How to integrate trillions of individual nanotube devices ?
Our strategy towards SWNTs-based CMOS chips
A i l B d St t E i i
Photocatalytic cutting
• Direct CVD growth
—— Axial Band Structure Engineering
Substrate-induced local deformation Photocatalytic cutting
Direct CVD growth
• Transfer-printing
local deformation
Organic/inorganic
d l ti l Cl i h
modulating layer
Diameter-tuning growth
m/s sorting or chemical conversion
Cloning growth
conversion Chinese patent: ZL 2006 1 0113214.5
ZL 2006 1 0113212.6
Catalytic CVD Growth of SWNTs Arrays on Surface
a-plane
Lattice Directed Growth Mode
Carbon source : Methane, ethanol, etc Substrate: Sapphire, quartz, SiO2/Si, etc
Gas Flow Directed Growth Mode
etc
Lattice Directed Growth vs. Gas Flow Directed Growth
on SiO2/Si ♦ Tube diameter: 1-2 nm
♦ Length:> 5 cm
♦ Spacing: 1 50 m
rected wth
♦ Spacing: 1-50 m flow dir VD growGas CV
Diameter:ca. 1 nm on sapphire
nted th
Length:ca. 1 mm
Spacing:~0.2 m pp
ice orien VD growt
1.4mm
Latti CV
Outline
Controlling Morphology of SWNTs on SurfacesSurfaces
Direct Growth of Semiconducting SWNT ArraysC C
T i Di t f SWNT b CC
Tuning Diameters of SWNTs by Temperature
Cap-engineering for SWNTsGrowth with Controlled Chiralities
Controlling Morphology of SWNTs on
Surfaces by Combing the Two Growth Modes
Surfaces by Combing the Two Growth Modes
Growth of serpentine SWNTs
(a) (c)
Substrate
(b)
Quartz or Quartz or
sapphire Certain lattice direction
♦The spacing between two parallel sections is mainly determined by the
♦The spacing between two parallel sections is mainly determined by the landing rate of the ultralong SWNT.
♦The landing rate can be controlled by cooling rate.
Therefore, the folding density of serpentine SWNTs can be controlled by cooling rate
cooling rate.
Serpentine SWNTs grown at different cooling rates
r
10 min 20 min 30 min
r u u
r u
25 25
25 25 m 25 m
25 m
40 min 50 min
r r
40 min 50 min
r u
r
25 m u 25 m r
u
T from 975oC to 775 oC.
The slower the cooling speed, the higher the tube folding density.
High Performance SWNT-FET with Identical Chirality
10 m 19 CNT
10 m
-1 0x10-4 0.0
(A)
4 CNT
Pd electrode, 800nm SiO2
gate dielectric -2.0x10
-4
-1.0x10
Vg= -20V
I ds(
19 CNT
2.5 m
4 CNT -2 Vds-1 (V) 0
3.0x10-4 Vds= 0.5V
10-6 10-4
s(A)
19 CNT 4 CNT
2.5 m
1.0x10-4 2.0x10-4
Ids(A)
-20 0 20
10-8
Vds= 0.1V
V (V)
I ds 4 CNT
0 6 12 18
0.0
Number of tubes Vg (V)
J. Zhang, et. al., Adv. Mater. 2009, 21, 4158-4162. (Inter Front Cover Paper).
The highest density we have achieved
(a) (b)
High density High density of SWNTs make it 1×1μm2
(spacing of ~50 nm)
possible to observed
i k
105
LF 2RBM
G RBM=290
a.u.) G RBM
various weak second-order Raman from
104
NC1 NC2
iTOLA LF
**
*
**
* D
2RBM
IMF
M
G'(2D)
sity (a G+RBM Raman from
SWNTs.
100 600 1100 1600 2100 2600 3100 103
* ** * NC3 2G
Intens
100 600 1100 1600 2100 2600 3100
Raman shift (cm-1)
Grow SWNT cross-bars in one batch
aa- plane
Lattice Directed Growth Mechanism
Gas Flow Directed Growth Mechanism
Interaction between Cu, Fe Catlysts and Quartz Surface
a) and b) illustrate the interaction between Cu, Fe nanoparticles and surface of quartz, the red balls represent oxygen atoms. c) and d) High-magnification SEM image of the lattice assisted SWNTs catalyzed by Cu and Fe. e) and f) Results of gas flow directed growth of carbon nanotubes where Cu and Fe are used as catalysts
ti l respectively.
SWNT Cross-bars and Its Potential Application
Low
temperature favors for lattice oriented
growth mode and high T for gas flow directed
growth mode.
Wi h With a moderate 930-950oC, crossbar can crossbar can be grown.
J. Zhang et. al., J. Phys. Chem. C. 2009, 113, 5341-5344 (cover)
Direct Growth of Semiconducting SWNT
Arrays by UV Irradiation Assistance CVD
Arrays by UV Irradiation Assistance CVD
Separation of s-SWNTS and m-SWNTs after Growth
Ion-exchange chromatography density gradient ultracentrifugation
Functionalization
A Hirsch Angew Chem Int Ed 2002 M S Arnold Nature Nanotech 2006 M Zheng Nature Mater 2003
A. Hirsch, Angew. Chem. Int. Ed. , 2002 M. S. Arnold, Nature Nanotech. , 2006 M. Zheng, Nature Mater., 2003
alternating current dielectrophoresis gas-phase plasma hydrocarbonation Laser Irradiation
R. Krupke. Science, 2003 H. J. Dai. Science, 2006 H. J. Huang, J. Phys. Chem. B, 2006 R. Krupke. Science, 2003 H. J. Dai. Science, 2006 H. J. Huang, J. Phys. Chem. B, 2006
Why s-SWNTS and m-SWNTs can be Separated
When a reactant comes near the SWNT, the electron can transfer from the metallic SWNT, but cannot transfer from the semiconducting SWNT. After that, the metallic SWNT can be damaged and the semiconducting SWN still survived
still survived .
J. Zhang, et. al., J. Phys. Chem. C 112, 3849, (2008)
Setup of Our Home-Made CVD System
Li h Light
Source UV Light Filter
Furnace Furnace
Quartz Tube UV Light
Substrate
Furnace
SEM images of SWNTs arrays under different irradiation time 80mW/cm2
15min 80mW/cm2
10min
No Irradiation
100 m 100
80mW/cm2 20 i
80mW/cm2 30 i
270mW/cm2 30 i
100um 100um
100um
20min 30min 30min
100um 100um
40um
When UV beam acted on the substrate, the density of the SWNT array decreased obviously. From above, the shorter the irradiation time, the longer and denser the SWNTs were. If we continued increasing the irradiation time or the irradiation intensity SWNTs would become shorter and shorter and disappeared eventually
irradiation intensity, SWNTs would become shorter and shorter, and disappeared eventually
UV Beam
(A) (B) (C)
Furnace Furnace
Shield
( ) ( )
100um 100um
Furnace
Part B Part A
M S S
514 5 nm
S
632 8 nm
(D) (E) (F)
514.5 nm 514.5 nm 632.8 nm
Raman Shift (cm-1) Raman Shift (cm-1) Raman Shift (cm-1)
(A) Sketch map of the comparison experiment for SWNT growth with and without UV irradiation. (B)/(C) SEM image of the growth result without/with UV irradiation. (D) Raman spectrum for part B with 514.5 nm excitation. (E)/(F) Raman spectrum for part A with 514.5/632.8 nm excitation. The metallic SWNTs signals were collected in the yellow rectangle while the semiconducting SWNTs signals were collected in the blue rectangle separately for all the three spectra.
25 C A
100um
10 15 20 A
B
Vds = 100 mV, L = 5.6 um, dSiO2 = 800 nm
0 5 10 B
10um -6 -4 -2 0 2 4 6
Vg (V) 0
Raman spectra demonstrated an amazing result that almost 100%g SWNTs were semiconducting.
Electrical measurement data showed that 21 out of 22 SWNTs (~95%) were semiconducting.
J. Zhang et al., J. Am. Chem. Soc., 2009, 131, 14642-14643
C C C C
Tuning Diameters of SWNTs by
Temperature-oscillation CVD growth
Temperature oscillation CVD growth
Controlling Diameters of SWNTs
By Catalyst Particle By Carbon Feeding
J. Phys. Chem. B. 2002,106, 2429-2433 J. Phys. Chem. B. 2006,110,20254-20257
Our Approach: Tune the Diameter of SWNTs by T
Temperature
T t P
Temperature Program
940 ature /o C
0 40 80 120 160
900 920
Tempera
Time /min Time /min
C C
C
Micro-Resonance Raman Spectrum of Individual SWNTs
Elaserlaser
1μm SWNTSWNT
Micro
Micro--Raman spectroscopyRaman spectroscopy ResonanceResonance condition: Econdition: Elaserlaser==Eii
4 0 0 0 4 5 0 0 5 0 0 0
5 5 0 0 G-mode
Radial Breathing Mode (RBM)
632.8nm
1 0 0 0 1 5 0 0 2 0 0 0 2 5 0 0 3 0 0 0 3 5 0 0
Intensity Disorder-
induced Line (D-
ω=(dt-β)/ α
0 2 0 0 4 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0 1 8 0 0 - 5 0 0
0 5 0 0 1 0 0 0
R a m a n s h i f t ( c m - 1 )
induced Line (D
line) (m,n)
A power tool for both the atomic and electronic structure of SWNT !
R a m a n s h i f t ( c m 1 )
Constant-temperature CVD
a c
1 m
) b m) 2 d
100 m
138
shift (cm-1
b
Height (nm
0
0 3000 6000
132
Distance along the tube (m)
Raman s H
Distance (µm)
0 1
SWNTs Grown by one Time Temperature O ill t d CVD (F 900oC t 950oC)
T2
T
Oscillated CVD (From 900oC to 950oC)
T1
nsity
T2
T2
nsity
-100 0 100 200
Distance alone the tube (m) Inten Inten
-1 t (cm) 150 (T
1 < T2)
1560 1600 120 160
T1 T1
man shift 140
T2 T1
145
1560 1600 120 160
Raman shift (cm-1)
S S
-100 0 100 200
Ram
Distance alone the tube (m)
135
S1 S2
Semiconducting-
Semiconducting Nanojunction Semiconducting Nanojunction
SWNTs Grown by one Time Temperature O ill t d CVD (F 950oC t 900oC)
Oscillated CVD (From 950oC to 900oC)
ty
T2
tyT1 T2
Intensit
T
Intensit
T2
-100 0 100
Distance alone the tube (m)
150 200 250
T1 T1
1500 1600 240
(cm-1 )
(T1 > T2)
Raman shift (cm-1)
man shift T2T1
210
-100 0 100
Ram 180
Distance alone the tube (m)
M1 M2
Metallic-Metallic Nanojunction
Distance alone the tube (m) j
Controlled Thinning of SWNTs via Temperature St U
Step-Up
(c) (d) (e) (f)
2 04±0 06 nm
1 m
1 86±0 05 nm
1 m
1 76±0 03 nm
1 m
1 32±0 04 nm
1 m
2.04±0.06 nm 1.86±0.05 nm 1.76±0.03 nm 1.32±0.04 nm
Controlled Thickening of SWNTs via Temperature Step Down
Step-Down
(c) (d) (e) (f)
1 34±0 04 nm 1 45±0 03 nm 1 69±0 04 nm 1 98±0 05 nm
1 m 1 m 1 m 1 m
1.34±0.04 nm 1.45±0.03 nm 1.69±0.04 nm 1.98±0.05 nm
Tube Diameter from Wide to Narrow Distribution
Narrowing the di t ib ti f t b distribution of tube
diameters by sequential temperature step-up or step-down
25
bes 20
25
bes
c) 4.0 mm site
30 35 40
f tube
b) 2.0 mm site a) 1.0 mm site
10 15 20
mber of tub
10 15 20
mber of tub
10 15 20 25 30
umber of
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0
5
di t ( )
num
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0
num 5
di t ( )
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0
5
n 10
di t ( )
diameter (nm) diameter (nm) diameter (nm)
J. Zhang et. al., J. Phys. Chem. C, 2009, 113(30),13051-13059
Multiple Intratube Nanojunctions by Repeating Temp. Oscillation
Six intramolecular junctions were induced by three temperature oscillations between 950oC and 880oC during CVD. (A) shows the scheme of temperature oscillation with time; (B) is an SEM image of several parallel ultralong SWNTs grown during the temperature oscillation; (C) shows Raman RBM peak
positions along a SWNT each peak corresponds to a time period in (A) positions along a SWNT, each peak corresponds to a time period in (A).
J. Zhang et. al., Nature Materials, 2007, 6, 283
Cap-engineering for Growing SWNT with Controlled Chiralities
Growth Mechanism of SWNT
1. Cap formation on catalysty
2. Carbon atoms bound to the cap
Tip growth
bound to the cap and form a
SWNT
www.adm.hb.se/~KIB/nanotubes.htm
Base growth
The cap structure determines the structure of the formed SWNT.
The cap structure determines the structure of the formed SWNT.
However, it is hard to control the cap structure by controlling the
t t f t l t ti l
structure of catalyst nanoparticles.
It follows the Vapor – Liquid – Solid (VLS) Mechanismp q ( )
Our Strategies --- Cap-engineering
Based on the concept of SWNTs Cloning, using a open end SWNT or opened Cp 6060 as seed/cap to grow SWNT. The structure of the open end p g p
SWNT or opened C60 determines the structure of the formed SWNT.
Open-end SWNT as seed/cap Opened C60 as seed/cap
Vapor-Solid Growth Mechanism / Open-End Growth Mechanismp p
Our experimental scheme
A B C
Original “seed” catalyst EBL+O2 plasma Second growth
Cloning SWNTs on Quartz Surface Tube length
15 2m 16 2m g 15 8m 16 6m
increase:
0 m – 1m
15.2m
5 m
16.2m
5 m
15.8m 16.6m
0 m 1m
15 0 10 0 6.8 15.6m 6.8
5m
5m 5 m5 m
15.0m 10.0m 10.3m
5 m 5 m
6.6m Rate of success 7.5m
for cloning:
5 m 5 m
g
24/59 = 40.7%
Before growth After growth
6.3m
192.8 cm-1
(i) quartz
6.3m
7.8m
nsity
﹡ ﹡
quartz
quartz
8.0m
Inten
5 m
100 200
6.3m
7.5m
8.7m Raman Shift (cm-1) 5
h d Tube diameter & chirality
unchanged
shortened
increased
y keep unchanged after
cloning growthg g
Cloning SWNTs on SiO2/Si Surface
Tube length
increase: increased
increase:
0 m – 4.6 m
increased
Rate of success increased
Rate of success for cloning:
56/600 = 9 3%
increased
56/600 = 9.3%
• CH4/C2H4/H2=100/5/300
• Growth T=975oC for 15min
Before growth After growth
f
2 2
2 site 1
1 2
3 4
ght (nm) 0 1
0
0 1
0
1 m 1 m Heig
0 2
0 1
4
0 1
0
2 3
Distance (m)
0 1
0 1
Before growth After growth
303 SiO /Si
﹡
5
sity 6
SiO2/Si
252.7 cm-1
2m
66
2m
Intens 252.7 cm
2 m
55
2 m
250 300
Raman Shift (cm-1)
AFM and Raman evidence for the diameter & chirality maintenance AFM and Raman evidence for the diameter & chirality maintenance
J. Zhang et al., Nano Lett., 2009, 9, 1673
Open-End Growth Mechanism
“VS” growth model ?
Using C60 cap to grow SWNTs
Experiments
Thermal oxidation to open the cage
Fullerendione Cap with amorphous carbon Fullerendione Cap with amorphous carbon
Cap with hydroxy groups
Active cap
SWNT
Growing SWNTs Using Opened C60 as Seeds:
Growth Result
Evidence of SWNTs Grown from Opened C60
Morphological a SWNTgrown from opened C60 and after heat treatment
SEM image of a SWNT grown from opened C60 and XPS of the SWNTs sample
Water treatment at 900oC
Without water
treatment at 900oC
Water treatment is very important for SWNTs growth!
Water treatment is very important for SWNTs growth!
Different thermal oxidation temperature
400 ºC c
a b
100μm
f
100μm
100μm 10μm
d e
100μm
10μm 100μm
(a)-(e) SWNTs grown from fullerendione using thermal oxidation temperatures of 300, 350, 400, 450 and 500 oC respectively. (f) SWNTs grown from pristine C using an oxidation temperature of 500 oC
grown from pristine C60 using an oxidation temperature of 500 oC
Diameter Distribution of SWNTs grown by using Opened C caps
Opened C60 caps
Diameter distribution of SWNTs grown from fullerendione under different oxidation conditions (measured from AFM)
different oxidation conditions (measured from AFM).
J. Zhang et al., Nano Lett., 2010, 10, 3343
Possible Growth Mechanism
Thermal
oxidation Open-end
growth growth
Thermal oxidation
Annealing Open-end
growth
The caps from C60 has a discreet diameter distribution.
The formed SWNTs are expected to have a discreet diameter distribution.
The formed SWNTs are expected to have a discreet diameter distribution.
Summaryy
Single-Walled Carbon Nanotubes
Application the SWNTs in Future Devicespp
Growing SWNTs on Surface Growing SWNTs on Surface
with Controlled Structures
Although it is still difficult to make a precise control of the diameter, chirality and local band structure of single-y g walled carbon nanotubes, there exists a big space for further efforts.
Acknowledgement g
Fund Support:
Collaborates:
pp
NSFC (20725307, 50521201);
MOST(2006CB932701)
Prof. Zhongfan Liu PKU
Prof. Jing Kong MIT
MOST(2006CB932701) Prof. Mildred Dresselhaus MIT