novel materials for organic and thin film electronics · novel materials for organic and thin film...
TRANSCRIPT
Novel Materials for Organic and Thin Film Novel Materials for Organic and Thin Film ElectronicsElectronics
A
B1 µµµµm
Ruud M. TrompIBM Research Division
T.J. Watson Research CenterPO Box 218
Yorktown Heights, NY [email protected]
The road beyond CMOS : The road beyond CMOS : NanoNano ??Nano: things smaller than 100 nm
Opportunities for revolutionary new materials,processes, and technologies
But: also many opportunities to improve, extend, andtransform present technologies, including computerhardware technologies
����
A
B1 µµµµm
������� ���
Silicon Logic Silicon Logic -- Already at the Already at the NanoscaleNanoscale
• Gate length = 6 nm• Power4 Chip
• 174 million transistors
B. Doris et al., IEDM , paper 10.6, 2002. J. Warnock et al., IBM J. R&D, p. 27, 2002
TSi=7nm Lgate=6nm
Source Drain
Gate
(That’s what got us into trouble!)
One alternative: Carbon One alternative: Carbon NanotubeNanotube FETFET
Semiconductor mobilities @ RT (cm2/V.s)InSb 77,000CdSe 650c-Si 1,500 – 100a-Si 1CNT 100,000Polymer 10-5 – 10-2
Pentacene 0.1 – 5Chalcogenides (spinc.) 1 -20
Carbon Carbon NanotubeNanotube InverterInverter
-2
-1
0
1
2
-4 -2 0 2 4
VIN (V)
VO
UT
(V)
Gain>1
Nanotube Nanotube TechnologyTechnology ??
CNTAu
100µm
How do you get from here to there?
����������������� ���������
� ����� �������������
Some of the issues• CNT synthesis and purification; control of diameter and chirality• CNT placement in integrated circuit hierarchy with nm precision• Site- and/or area-selective n- and p-type CNT doping on nm scale• Control over contacts to CNT, injection barrier to n-, p-, i-CNTs• Elimination of parasitics for high performance• Elimination of 1/f noise in CNFET devices• Theoretical modeling of CNT physics, chemistry, and devices• CNT and device physical characterization• Optoelectronic properties
1/f noise in carbon nanotubesPhilip G. Collins, M. S. Fuhrer, and A. ZettlAPL 76, 894 (2000)
3
2
1
0
Nor
mai
lized
Cou
nts
1.61.41.21.00.80.6
Diameter (nm)
Normalized diameter distribution Gaussian based curve fitting
Contacts
CVD CNT, Pd contacts, 8nm HfO2
CVD CNT, Pd contacts, L~2umCVD CNT, Pd contacts, L=50nm
Arc discharge CNT, Pd contactsArc discharge CNT, Ti contacts
0.8
0.6
0.4
0.2
0.0
Nor
malized
Cou
nts
10-4
10-3
10-2
10-1
100
101
Ion
(����A)
Pd Contacts Ti Contacts Al Contacts
Vds = -0.5V
Zhihong Chen, Joerg Appenzeller, Joachim Knoch,Yu-Ming Lin, Phaedon AvourisA26-5
CNT Optoelectronics
Vds is fixed, Vg is scanned50 µm
SEM
0.5 0.6 0.7 0.80
1
Inte
nsity
(a.u
.)
Energy [eV]
Emission spectrum
DRAIN
SOURCE
e-
h+
e-
h+h+
e- e-
0.0 0.2 0.4 0.6 0.8 1.01.0
0.8
0.6
0.4
0.2
0.0
12 13 14
0
10
20
30
40
50
gate voltage �g
spot
( �
m )
spot
( L
)
1.0
0.8
0.6
0.4
0.2
0.0
CNT optoelectronic properties are key in understanding basic transport properties, effects of defects, CNT electronic structure, excited states, etcetera.There is room for much experimental work, as well as more advanced theoretical understanding.
Exciton binding energy is CNTs (0.1-0.5 eV) much larger than in III-V’s (10’s of meV). CNT exciton bonding energy depends on environment.
Ph. Avouris, A26-4
Nanotube Nanotube TechnologyTechnology ??
CNTAu
100µm
How do you get from here to there?
����������������� ���������
� ����� �������������
Not ‘just engineering’
Lots of basic science !
A.J. Heinrich, C.P. Lutz, J.A. Gupta, D.M. EiglerScience 2002
A Molecular Computer A Molecular Computer –– Slow but it works (once)Slow but it works (once)
Molecular TransistorsMolecular TransistorsElectrostatics of Molecular TransistorsElectrostatics of Molecular Transistors
C. R. Kagan, A. Afzali, R. Martel, L. M. Gignac, P. M. Solomon, A. G. Schrott, B. Ek, Nano Letters, 3, 119 (2003).
gate
source drain
gate dielectric
L
tdielectric
Au Au
molecule Barrier lowered byEnergy level offset
Hybridizationand Charge transfer
between metal-molecule
Design molecules with L>2.5-3 nmTailor tunnel barriers through chemistry
OFF state: Limited by Tunneling1
Gate Modulation: Electrostatics2
gate
source drain
gate dielectric
tdielectric ���� L
to attain gate field not dominated by
drain fieldYet
dielectric not leaky TSi=7nm Lgate=6nm
Source Drain
Gate
Is Si really that much bigger?
Directed Assembly of Molecular DevicesDirected Assembly of Molecular Devices
Wavelength (nm)
200 400 600 800
Abs
orba
nce
0.00
0.02
0.04
0.06
0.08
0.10Average bilayer coverage = 0.9x10-10 mol/cm2
(X-ray single crystal structure: 1.0x10-10 mol/cm2)
Wavelength (nm)200 300 400 500 600
Abs
orba
nce
0.00
0.01
0.02
0.03
0.04
Number of Bilayers0 2 4 6 8 10
Incr
ease
of A
bsor
banc
e
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Rh 4 loo
p at
258
nm
Dipyridine linker at 298 nm
LayerLayer--byby--Layer AssemblyLayer Assembly
n
Rh
RhO
N
N
N
N
O
OO
Rh
RhO
N
N
N
N
OOO
N N
N N
N N
S S
Au
surfacetemplate1
inorganic2
organic3
Manipulate chemistry of template, organic, and
inorganic
n+ Si
SiO2
M1 M1
Device AssemblyDevice Assembly
Manipulate device geometry and molecular assembly
V (V)-10 0 10 20 30 40 50
I (A
)
0
1e-6
2e-6
3e-6
4e-6
5e-6
Rh
RhO
N
N
N
N
O
OO
Rh
RhO
N
N
N
N
OOO
NNNN
N
N
ZnNNNN
N
N
Zn
V (V)
-60 -40 -20 0 20 40 60
I (A
)
-2e-6
-1e-6
0
1e-6
2e-6
3e-6
4e-612345
V (V)0 20 40 60 80
I (A
)
-2.0e-7
0.02.0e-74.0e-7
6.0e-78.0e-71.0e-61.2e-6
1.4e-61.6e-61.8e-6
Rh
RhO
N
N
N
N
O
OO
Rh
RhO
N
N
N
N
OOO
N N
N N
V (V)-60 -40 -20 0 20 40 60
I (A
)
-3e-7
-2e-7
-1e-7
0
1e-7
2e-7
3e-7
4e-7
5e-71234
Device CharacteristicsDevice Characteristics
Study low and high bias characteristics
C. Lin, C. R. Kagan JACS, 125, 336 (2003)
PentacenePentacene: the world: the world’’s best organic semiconductors best organic semiconductor
GateSubstrate
Insulator
����������������������������������������
DrainSourceSemiconductor
Pentacene on Au contact
Pentacene onSiO2 Insulator
����� ������������������
����������������
1 ����m
V D (V )-4-3-2-10
I D( µµ µµ
A)
-5
-4
-3
-2
-1
0
V G =-4 V
-3 V
-2 V
-1 V0 V
Mobilities up to 5 cm2/Volt.secreported (a-Si: 1 cm2/Volt.sec)
Vacuum deposition
Key Question: How do Molecules interact Key Question: How do Molecules interact with Contacts?with Contacts?
������� ���
������ ��
������× ��������� �������
��������
������� ���
����
�������
�������
Pn stands up on non-metallicsurfaces, such as oxides, semiconductors or semimetals,but lies down flat on metals.
Fractal growth and scaling is consistent with classical growth modelsdeveloped for inorganic materials
J. Sadowski, G. Thayer, F. Meyer zu Heringdorf, R. Tromp
Growth on clean Si
Field of view 65 µmRoom temperature1 image/minute
Meyer zu Heringdorf, F.-J.; Reuter, M.C.; Tromp, R.M. Nature 2001, 412, 517-520
Deposition time (minutes)
Integral coverage (ML)0 0.5 1 1.5 2 2.5
Lay
er c
over
age
(%)
0
50
100
0 30 60 90 120
Aromatic and Pi-conjugated systems:
Benzene/Si(001) also undergoes transition from sp2 to sp3 hybridizationReversibly adsorbs, desorbs
Multiple binding configurations
Benzene liquid
Benzene/Si(001)
Wavenumber (cm-1)
2800 2900 3000 3100
K.P. Weidkamp, C.A. Hacker, M.P. Schwartz, X. Cao, R.M. Tromp and R.J. Hamers,
Abs
orba
nce
310030002900Frequency (cm
-1)
22002100
(a) Bulk pentacene
÷÷÷÷200
(b) Monolayer pentacene
2x10-4
(c) Multilayer pentacene
(d) Multilayer pentaceneafter heating
Pentacene on Si(001): Infrared Spectra
• Pentacene (bulk) shows no sp3
hybridized C-H stretching vibrations
• Monolayer coverage – peaks at 2091, 2870, 2922 cm-1 indicate Si-H and sp3
hybridized C-H bonds, also thermally stable
• Multilayer coverage – small Si-H peak, peaks above 3000 cm-1 are much larger than any below 3000 cm-1
• Heating multilayer – broader Si-H peak, peaks almost identical to those of monolayer(x4)
Interfacial layer involves transition of some C atoms to sp3 hybridization and some dissocation. Layer appears to be thermally stable and irreversibly bound.
K.P. Weidkamp, C.A. Hacker, M.P. Schwartz, X. Cao, R.M. Tromp and R.J. Hamers,
Cyclo-addition reaction on Si
Organic surface termination renders Si surface inert, providing anideal substrate for subsequent pentacene growth
cyclopentene cyclohexene cyclo-octadiene
SiSiSiSiSiSi
From cyclical molecules to chains
Chainlike molecules render Si surface inert, but affectdiffusion and aggregation during pentacene growth
hexene nonene dodecene
SiSi SiSiSiSi
Cyclooctadiene Hexene
Dodecene off axisDodecene on axis
Crystals at 90 degreesincrements
Crystals at 45 degreesincrements
Selection of azimuthal angles: Role of molecular species
LEED 8/20 dodecene Si(001) on axis
LEED 6/11 COD Si(001) on axis Epitaxial growth, rectangular unit cell
Pn
Pentacene:a = 0.627 nmb = 0.777 nmγ = 84.7o
Si(001):dimer-dimer: 0.384row-row: 0.768
Mismatch ~ 1%
Epitaxial growth of pentacene on Si:diffraction analysis
tilted illumination
on axis 4 degrees off axis
PENTACENE /Si(001) EPITAXY
0.768 nm
Pentacene MD on IBM BlueGeneclassicaldiffusion
non - classical diffusion pentacene diffusion on dodecene
Soluble Soluble pentacenepentacene precursorsprecursors
hνpolymeric precursor
SNO
O
anneal
Photosensitive version for patterningSoluble pentacene for spincoated TFTs
µµµµlin = 0.4µµµµsat = 0.8on/off > 106
Highest performancefrom solution processed organic TFT
ChalcogenidesChalcogenides: a new look at some old materials: a new look at some old materials Eg µ µ µ µn µ µ µ µp
(eV) (cm2/V-sec) (cm2/V-sec)
SnS2 2.6 18
SnSe2 1.6 27
ZnSe 2.7 600
ZnTe 2.25 100
CdS 2.42 250
CdSe 1.73 650
Organic-derivatized CdSe nanocrystals -- n-type 1.0 cm2/V-s[Ridley et. al., Science 286, 746 (1999)]
Chemical Bath Technique CdSe -- n-type 15 cm2/V-s[Gan et. al., IEEE Trans. Electr. Devices 49, 15 (2002)]
Assembled Nanorod / Nanoribbon -- n-, p-type <300 cm2/V-s[Duan et. al., Nature 425, 274 (2003)]
Semiconductor mobilities @ RT (cm2/V.s)InSb 77,000CdSe 650c-Si 1,500 – 100a-Si 1CNT 100,000Polymer 10-5 – 10-2
Pentacene 0.1 – 5Chalcogenides (spinc.) 1 -20
Solution Solution ProcessableProcessable ChalcogenidesChalcogenides
��� ������������� �� ���
�� �����������
���������� �
��� ������
ChalcogenidesChalcogenides –– a new lowa new low--T spinT spin--on on semiconductor with high mobilitysemiconductor with high mobility
A B
Sn
S SeS
Sn
10 15 20 25 30
170 175 180 185 190 195 200
10 15 20 25 30
170 175 180 185 190 195 200
2 (degree)θ
Inte
nsity
(arb
. uni
ts)
Energy (keV)
2 (degree)θ
Inte
nsity
(arb
. uni
ts)
Energy (keV)
1 mµ 1 mµ
RMS Roughness: 6 Å (A) and 14 Å (B)Film thickness: only 7- 9 unit cells thick!Composition: SnS1.8 and SnS1.4Se0.5
µµµµsat = 12.0 cm2/V-sµµµµlin = 2.4 cm2/V-sIon/Ioff > 106
L = 14 µµµµm; W = 250 µµµµmHighest spin-coated channel mobility by ~ 10X
Developed new chemistriesto spin on thin chalcogenidefilms with low processingtemperatures < 300 C
Nanoscience may give rise to future revolutionarytechnologies relevant to Information Technology:Nanotube or molecular logic; Novel, dense crosspointmemories; New high performance spin-coatable semi-conductors; etcetera.
Also, Si technology itself is a rapidly changing revolutionary technology that provides a ‘ready’ platformfor nanotechnology integration:Materials; Processes; Devices; Lithography; etcetera.
Development of new technologies to replace an existing technology takes decades, not months – but we can utilize nanotechnology on shorter timescales by insertion in existing technologies, and in niche application.