ICP NickelICP Nickel Plasma for Deposition of Deposition of NanodotsNanodots using using NanopantographyNanopantography

Nader Sadeghi,University Joseph Fourier-Grenoble & CNRS,

FranceLin Xu, Vincent M. Donnelly and Demetre J. Economou

University of Houston, USA

IMP - Plasma Dynamics, 2006 Tehran

OutlineNanopantographyEtching of 10 nm holes by using a monoenergeticAr+ ion beamMaximizing the Ni flux for deposition of nanodotsDeposition of ~10 nm nickel nanodots by using a monoenergetic Ni+ ion beamSummaryFuture workAcknowledgement

Broad area collimated ion beamVb

conducting film or substrate, Vs

metal, Vmdielectric

Side View: TopView:

Essential components:Collimated monoenergetic ion beams and electrostatic lenses built on the wafer

Nanopantography can do:Sub-10nm patterning simultaneously;Self-alignment, immune to vibration and thermal expansion;Etching (Ar+), deposition (M+), oxidation (O+).

What is nanopantography?

Essence of nanopantographyWhatever is “written”once on the imaginary plane here

is reproduced at the bottoms of all the lenses here.

Side view

Writing of any desired patterns simultaneously.

Top view

Etching of ~10 nm holes by a monoenergetic Ar+ ion beam

Size reduction factor of ~95XThe depth of etched holes in the center is about 110 nm

L. Xu, S. Vemula, M. Jain, S. K. Nam, V. M. Donnelly, D. J. Economou and P. Ruchhoeft, Nano Letters, 5, 2563 (2005).a)

SEM

950 nm

b)

20 nm

c)

10 nm 50 nm

HAADF-STEM, dots

~110 nm

TEM

Ni nanodots deposited by nanopantography

Application 10-nm Ni nanodots:large-area fabrication of uniform, location-specific, and sub-10 nm Ni catalyst dotsto grow single-wall carbon nanotubes (SWCNTs)

Ni Ni nanodotsnanodots patterned by patterned by nanopantographynanopantography Growth of Growth of SWCNTsSWCNTs

The keys to the success:A monoenergetic Ni ion beam and high Ni+ ion flux

Lens array voltage

Sample with lens array

Deposition Chamber

Collimated, monoenergeticAr+ and Ni+ ion beamDrift tube

Pumping

Pumping

Setup of nanopantography for Ni deposition

Ar pulsed plasma

Matching Network

Ni ion source Ni targetNi coil

To monochromator and detectorOES, OAS

Pulsed Ni lamp

The extraction of a monoenergetic Ni ion beam

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4(b)

102.0 eV71.5 eV51.0 eV30.5 eV

dc bias: 30 V ; 50 V 70 V; 100 V

Nor

mal

ized

IED

Ion energy (eV)

Lin Xu, et al. Appl. Phys. Lett. 87, 041502 (2005).

The energy spread was 3.4 eV (FWHM) for a peak ion beam energy of 102.0 eV, FWHM measured by a high-resolution ion energy analyzer was 1.2 eV for

a 100 eV ion beam.

Ni+ ion beams should have similar IED as Ar+ ion beam ;By setting the ion landing energy below the etching threshold (e.g 20 eV), net deposition takes place.

Approaches to determine Ni atom and ion densities:

Ni atom density was measured by optical absorption spectroscopy (OAS) and compared with model; Ni+ density was calculated from global model using the measured Ni atom density (optical absorption signal of Ni+ was too weak!)

Characterisation of the plasma source to maximize the Ni and Ni+ densities

Knowing other plasma parameters such as Tg, ne, and Te allows more accurate determination of Ni atom or Ni+ ion

densities.

Gas temperature (Tg) measurement

Tg was determined by adding a small quantity of N2 to the gas and fitting the population distribution in rotational levels of the N2(C3Пu; v) state, obtained from the N2 second positive (C3Пu- B3Пg) plasma-induced emission spectra, with a Boltzmann distribution

[ ] )/exp()12( rJJ kTFJN −+∝

However, due to the very efficient energy transfer reaction:Ar*(3P2) + N2 (X) → Ar + N2 *(C3Пu; v'= 0, 1, 2, high J)that populates high J rotational levels of v’=0, the population distribution of v’=0 must be fitted with a two exponential law:

[ ] ⎟⎟⎠

⎞⎜⎜⎝

⎛−+−+∝ )/exp()/exp()12( H

rJHr

LrL

rJJ kTFTT

kTFJN ϕ

Fit of 0-0 4-4 band of N2(C3Пu- B3Пg) transition

3320 3330 3340 3350 3360 3370 3380

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

3320 3330 3340 3350 3360 3370 3380

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

Y Ax

is T

itle

0-0, 20 mTorr, 40 W

Y Ax

is T

itle

Wavelength (A)

Simulation330 K + 20% 1700 K

3320 3330 3340 3350 3360 3370 3380

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

0.60

0.65

0.70

3320 3330 3340 3350 3360 3370 3380

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

0.60

0.65

0.70

Y Ax

is T

itle

0-0, 20 mTorr, 200 W

Y Ax

is T

itle

Wavelength (A)

Simulation450 K + 17% 1500 K

Fit of 3-3 and 4-4 bands of N2(C3Пu- B3Пg) transition

3250 3255 3260 3265 3270 3275 3280 3285 3290

0.000

0.005

0.010

0.015

3250 3255 3260 3265 3270 3275 3280 3285 3290

0.000

0.005

0.010

0.015

Y A

xis

Title

3-3, 8 mTorr, 40 W

Y A

xis

Title

Wavelength (A)

Simulation3-3 300 K + 4-4 350 K

3250 3255 3260 3265 3270 3275 3280 3285 3290

0.000

0.005

0.010

0.015

0.020

0.025

3250 3255 3260 3265 3270 3275 3280 3285 3290

0.000

0.005

0.010

0.015

0.020

0.025

Y A

xis

Title

3-3, 8 mTorr, 200 W

Y A

xis

Title

Wavelength (A)

Simulation3-3 350 K + 4-4 470 K

Estimate of electron temperature (Te)

effeB

Areizgeiz

dTunTknTk 1

)()()( *

*

=+

The electron temperature (Te) was estimated from the global model (see M. A. Lieberman, and A. J. Lichtenberg, Principles of Plasma Discharges and Materials Processing , 2nd edition (Wiley, New York, 2005) including ionization of Ar metastables. Thus, by rearranging the Ar+ mass balance one obtains:

12eff

L R

RLdRh Lh

=+

with the effective plasma length deff given by

*

*]))()([()( 2

***Ar

Areeizeqgeeex n

DnTkTknnTk

Λ++=

and the density of argon metastables nAr* deduced from density balance:

Fitting and calculation results for Tg and Te are:

Plasma condition Tr (K) Tg (K) Te(eV)

Power (W) Pressure(mTorr) 0-0 3-3 4-4

40 20 350 350 370 360±20 2.19

200 20 450 420 480 450±30 2.30

40 8 360 300 350 355±20 2.53

200 8 440 350 470 455±20 2.62

0-0 and 4-4 bands allways provided similar rotational temperature;

Tgs at intermediate powers were obtained by extrapolation.

The absolute electron density in the plasma was estimated by an energy balance equation:

, ,g x g x g eI ak n n=

One can also estimate the relative variation of ne with plasma parameters from the variation of argon 419.8 nm line intensity, which is exclusively excited from the ground state of argon:

Estimate of electron density (ne)

0abs

B eff T

Pneu A ε

= 2 ( )eff L RA R Rh Lhπ= +

and εT is the total energy loss per electron-ion pair formed

where

I419.8∝ ak419.8nArne

Comparison of electron density variation from two methods

0 50 100 150 200 2500.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

8 mTorr

12 mTorr

15 mTorr 8 mTorr 12 mTorr 15 mTorr

I g,x/k

g,xn

g(Arb

. Uni

ts),

n 0(1011

cm-3)

Power to the coil (W)

n0 variation derived from OES (symbols) agrees well with the model prediction (lines).Linear dependence with ICP power.

Pumping

Ar pulsed plasma

Matching Network

Ni ion source Ni targetNi coil

To monochromator and detectorOES, OAS

Pulsed Ni lamp

The fractional absorption, , where I0 and I are the transmitted line intensities without and with plasma, respectively, is given by

Measurement of Ni atoms density by resonance line absorption methode

0

0

III

AL−

=

02

1

( )! 1

m

Lm

k lAm mα

∞

=

−= −

+∑ ><

Δ=

−

][1025.8 13

0 Nifxk AσWhere α2 = (TE/TA)

D. Ohebsian, N. Sadeghi,… Opt. Commun. 32, 81 (1980)

Nickel Hollow cathode

Anode

Cathode -200 V

The Ni atom density measured by OAS

Ni absorption line

Transition Energy level Oscillator strength

232.0 nm a3F4 -- y3G05 0 cm-1-- 43 089.578 cm-1 0.68

The Ni density measured here is only for Ni ground state not total Ni density. Ni has several metastable states with energies close to that of ground state.

The line we chose:

The partition factor of Ni ground state

Configuration Term J Energy level(eV)

Statistical weight (2J+1)

Population in ref. 1

Population in ref. 2

3d54s2 a3F 432

00.0250.275

110.780.56

10.90.6

10.2

0.183d9(2D)4s a3D 3

21

0.0250.1090.212

0.780.560.33

5.77.40.3

1.61.1

0.25Other metastable

statesb1D 2 0.422 0.075 0.3

P(partition factor)

0.25 (1/4)

0.063(1/16)

0.21(1/4.65)

Ni population distribution after sputtering from Ni target by keV Ar+

1. C. He et al. Phys. Rev. Lett. 75, 3950 (1995).2. E. Vandeweert et al. Phys. Rev. Lett. 78, 138 (1997).

Will sputtered Ni atoms in metastable states quenchto ground state after traversing the plasma?

Most probable Kinetic energy of sputtered Ni: 2.2 eV [1];

Ni metastable states Ni ground state[2]

a3D (3d94s1, metastable) a3F (3d84s2, ground state)[3]

[1] M. W. Thompson, Philos. Mag. 18, 377 (1968).[2] H. Matsui, H. Toyoda, and H. Sugai, J. Vac. Sci. Technol. A 23, 671 (2005). [3] P. Baltayan, F. Hartmann, I. Hikmet, and N. Sadeghi, J. Chem. Phys. 97, 5417 (1992).

Since sputtered N population inversion and partial thermalization,we assumed:

P=1/4 and thus, NiTotal= 4*NGround

inefficient quenching

Ineffective quenching by rare-gas collisions@ 8~20 mTorr

Computation of Ni atom density

*,

/

ArPeNiizloss

NiNiBSNi nknkk

VAYunn++

=+

Mass Balance

Global model(0-D):

wDloss kkk /1/1/1 +=Combined with:

2,

Λ= effNi

D

Dk V

Avk ArNi

w,

)2(2 γγ−

=

KnNiArNieffNi DDD ,,,

111+=

More details can be found in the following references:M. A. Lieberman, and A. J. Lichtenberg, Principles of Plasma Discharges and Materials Processing , 2nd edition (Wiley, New York, 2005).C. Lee, M. A. Lieberman, J. Vac. Sci. Technol. A 13, 368 (1995).P. J. Chantry, J. Appl. Phys. 62, 1141 (1987).

Where,

Ni atom density from simulation and experiments (Without target bias)

0 40 80 120 160 200 2400

4

8

12

16

2020 mTorr

15 mTorr12 mTorr

8 mTorr

8 mTorr 12 mTorr 15 mTorr 20 mTorr

Tota

l Ni d

ensi

ty (1

09 cm-3)

Power on the coil (W)

Pressure Diffusion loss nNiPower Sputtering of coil nNiSaturated nNi after 90 W results from gas heating and Ni

ionization;Mismatch of model and measurement at lower pressure results

from assumption of thermal equilibrium between Ni atom with Ar gas in OAS.

Calculated Ni+ ions density

20 40 60 80 100 120 140 160 180 200 2200

4

8

12

16 20 mTorr 15 mTorr 12 mTorr 8 mTorr

Tota

l Ni+ d

ensi

ty (1

07 cm-3)

Power on the coil (W)

To obtain higher Ni+ density:plasma power, pressure and target power

“Hero Experiment”- a proof-of-principle experiment

Highly doped Si wafer

Plastic spacer150 V190 V180 V

200 eV Ni+Mesh with 150 um hole array

SEM image:

Assembly:

Ni dots

Element identification: Ni peaks in XPS and magnetic force in MFM

Ni dots

Nanodots deposited by nanopantography

SEM image (single lens, top view)

~700 nm

~10 nm

Size of nanodots: ~10 nmDemagnification factor: X70

AFM image ( lens array)

Height of nanodots: 5~10 nm

Ni nanodots

Conclusions

1. We have demonstrated the principle of nanopantography in etching of ~10 nm holes by nanopantography;

2. OAS combined with a model has been used to determine theNi atom densities at varied plasma condition, which is importantfor maximizing the deposition rate in nanopantography;

3. ~10 nm Ni nanodots have been deposited by nanopantography in a massively parallel fashion.

~

⊙⊙

Future work: Next-generation nanopantography system

Cryo pump

Turbo pump

Turbo pump

Load Lock

Cl atom beam

20”

2X

5X

2X

Substrate

2X

Total Flux: 2x5x2x2=40 !DreamDream

Writing of arbitrary patterns