PROBING THE GAS PHASE CHEMISTRY INVOLVED IN DIAMOND CHEMICAL VAPOUR DEPOSITION (CVD)

1

PROBING THE GAS PHASE CHEMISTRY

INVOLVED IN DIAMOND CHEMICAL

VAPOUR DEPOSITION (CVD)

Mike Ashfold

School of Chemistry

University of Bristol

Bristol BS8 1TS

http://www.chm.bris.ac.uk/pt/laser/

2 Chemical vapour deposition of diamond films

Activation of gas mixture by• Hot filament (Tfil~2450 K)

• Microwave plasma• DC arc jet

Polycrystalline films grow on Si, Mo, W, … substrates; Tsub > 700°C.

Growth of single crystal diamond by CVD demonstrated (Isberg et al, Science, 297, 1670 (2002))

Properties:• High thermal conductivity• Optical transparency (UV mid IR)• Chemically inert• Electrical insulator – can be doped

3 SEM images of polycrystalline CVD diamond

4 Film characterisation by Raman spectroscopy

Raman Shift / cm-1

600 800 1000 1200 1400 1600 1800

Inte

ns

ity

/ arb

. un

its

1332 cm-1

5 Recent studies by the Bristol Diamond GroupHydrocarbon / H2 mixtures in a hot filament (HF) reactor • Molecular beam mass spectrometry• Laser probing H atoms, CH3 radicals and C2H2 molecules

• Modelling CH4 / H2 and C2H2 / H2 gas mixtures(Ashfold et al., Phys. Chem. Chem. Phys. (2001), 3, 3471)

• Probing and modelling CH4/NH3/H2 gas mixtures(Smith et al., J. Appl. Phys. (2002), 92, 672)

CO2 / CH4 mixtures in a microwave (MW) reactor • Molecular beam mass spectrometry• Modelling H/C/O gas phase chemistry• Effect of S additions (as H2S or CS2)(Petherbridge et al., J. Appl. Phys. (2001), 89, 1484, 5219)

CH4 / H2 / Ar mixtures in a DC arc jet reactor

• Cavity ring down measurements of C2H2 and of C2 and CH radicals

• Modelling plasma activated CH4 / H2 gas mixtures(Wills et al., J. Appl. Phys. (2002), 92, 4213Rennick et al., Chem. Phys. Lett. (2004), 383, 518Rennick et al., Diam. Rel. Mater. (2004), 14, 561)

6 Diamond film growth in a hot filament reactor

0.5-1% CH4 in H2

Total flow rate = 100 sccm

Pressure = 20 Torr

Film growth rate ~ 1m/hr

7 In situ molecular beam mass spectrometry, I

0

0.2

0.4

0.6

0.8

1800 2000 2200 2400 2600 2800

Filament Temperature / K

Hyd

roca

rbon

Mol

e F

ract

ion

(10

-2)

0.0

0.2

0.4

0.6

0.8

1.0

Tot

al C

arbo

n F

ract

ion

CH4

C2H2

C2H4

C Bal.

1%CH4 / H2

Gas mixture sampled through a skimmer by a differentially pumped quadrupole mass spectrometer, equipped with variable energy electron ioniser.

X [CH4] ~ X [C2H2]

8 Predictions from equilibrium thermodynamics

1.0E-06

1.0E-05

1.0E-04

1.0E-03

1.0E-02

1.0E-01

1000 1500 2000 2500

Temperature / K

Mo

le F

rac

tio

n

CH4

CH3

H

C2H2

C2H4

[CH4] ~ [C2H2] at Tgas ~ 1500 K

Gas temperature / K

9 In situ molecular beam mass spectrometry, II

0

0.1

0.2

0.3

0.4

0.5

0.6

1800 2000 2200 2400 2600 2800

Filament Temperature / K

Hyd

roca

rbo

n M

ole

Fra

ctio

n (

10

-2)

0.0

0.2

0.4

0.6

0.8

1.0

To

tal C

arb

on

Fra

ctio

n

CH4

C2H2

C2H4

C Bal.

0.5%C2H2/H2

Equilibrium thermodynamics predicts no C2H2 CH4

conversion!

X [CH4] ~ X [C2H2]

10 MBMS probing HF-CVD reactors: Conclusions

• Near HF (i.e. where Tgas > 1400 K) the hydrocarbon is present as a CH4C2H2 mixture, irrespective of original source gas used.

• High [hydrocarbon] increased growth rates, but poorer diamond ‘quality’ (as assessed by morphology, Raman etc.)

• High [H] reduced growth rates, but improved diamond ‘quality’.

• Large T gradients near HF preferential diffusion of heavier species (i.e. hydrocarbons) from hotter regions - Soret effect.

• CH3 is a key diamond growth species in weakly activated hydrocarbon / H2 gas mixtures.

• Narrow range of Tfil for optimal diamond growth: - Low Tfil: insufficient H2 H dissociation, - High Tfil: CH4C2H2 equilibrium shifts far to right, and [CH3] in the growth region falls.

11 ‘Ideal’ requirements of a diagnostic

• Species selectivity• High sensitivity• Spatial (and temporal) resolution• Minimal intrusion• Ease of implementation• Readily interpretable results

laser spectroscopy

Resonance enhanced multiphoton ionisation (REMPI) for H atoms and CH3 radicals in HF-CVD reactor – spatially resolved (relative) number densities.

Cavity ring down spectroscopy (CRDS) for CH(X), C2(a) and C2(X) radicals, and C2H2 molecules, in DC-arc jet reactor – spatially resolved (absolute) column densities

12 Hydrogen atoms, I

H atoms are crucial in most diamond CVD environments. They:• Initiate gas phase chemistry ( reactive C containing species)• Terminate the growing diamond surface, preventing reconstruction• Abstract surface H to create vacant sites for C-radical attachment• Etch non-diamond deposits such as graphite.

n = 1

n = 2

n = 3

IonisationContinuum

(a) (b)n = 1

n = 2

n = 3

IonisationContinuum

(a) (b)n = 1

n = 2

n = 3

ionisationcontinuum

• Probe by 2+1 REMPI in HF activated hydrocarbon / H2 gas mixtures

• Move filament and substrate relative to laser beam focus to obtain spatially resolved number density distributions

• Collect H+ ions with a biased Pt probe

13 Hydrogen atoms, II

Wavenumber / cm-1

82252 82256 82260 82264

= 1.70(5) cm-1

= 2.46(6) cm-1

(i)

(ii)

Wavenumber / cm-1

82252 82256 82260 82264

Wavenumber / cm-1

82252 82256 82260 82264

= 1.70(5) cm-1

= 2.46(6) cm-1

(i)

(ii)

• Measure H+ ion yield as a functionof laser excitation wavenumber.

• Laser bandwidth makes negligible contribution to measured 2s 1slineshape.

• Area of lineshape local numberdensity of H atoms, NH.

• Local gas temperature obtained from FWHM of Doppler broadened(Gaussian) lineshape.

D

c

k ln=

14 Hydrogen atoms, III

Filament distance, d / mm

0 2 4 6 8 10 12

Tem

pera

ture

/ K

600

800

1000

1200

1400

1600

1800

2000

Filament distance / mm

0 2 4 6 8 10 12

Rel

ativ

e H

ato

m n

umbe

r de

nsit

y0

1

2

3

(a)

(b)

Tgas vs d, fixed Tfil.

Tgas (from FWHM values) declines monotonically with distance, d, from the HF.

NH vs d, fixed Tfil.

NH falls with increasing d, but much more slowly than would be predicted by assuming that the H2 2H equilibrium was determined by the local Tgas.

15 Hydrogen atoms, IV

Measured NH dependences are consistent with H atom formation by dissociative chemisorption of H2 on HF surface, and subsequent diffusion throughout reactor.

Gas phase H atom recombination is very inefficient when H2 is the main ‘third body’.

Tga

s / K

1000

1250

1500

1750

Tfil / K

1900 2000 2100 2200 2300 2400 2500 2600 2700

H a

tom

num

ber

dens

ity /

arb.

uni

ts

0

5

10

15

20

25

30

(a)

(b)

Tga

s / K

1000

1250

1500

1750

Tfil / K

1900 2000 2100 2200 2300 2400 2500 2600 2700

H a

tom

num

ber

dens

ity /

arb.

uni

ts

0

5

10

15

20

25

30

(a)

(b)

p(H2) / Torr

0 10 20 30 40 50 60

H a

tom

num

ber

dens

ity

/ arb

. uni

ts

0.0

0.2

0.4

0.6

0.8

1.0

1.2

NH independent of p(H2) – zero order kinetics

NH increases rapidly with Tfil

16 Methyl radicals, I

• CH3 radicals can be detected by 2+1 REMPI via the v = 0 level of the 3pz,2A2” Rydberg state.

• The two photon transition from the ground state is dominated by an intense Q branch.

• To convert measured REMPI intensities into CH3 radical number densities we must correct for the Tgas dependence of:

(i) the vibrational partition function(ii) the rotational band contour.

59600 60000 60400

Simulation

Experimental

Two-photon wavenumber / cm-1

17 Methyl radicals, II

[CH3] vs d from HF activated CH4/H2 and C2H2/H2 mixtures

1% CH4 / H2 0.5% C2H2 / H2

Corrected Raw Data

18 Methyl radicals, III

0.0 0.5 1.0 1.5 2.0

0.0

0.2

0.4

0.6

0.8

1.0

1.2

[CH

3] /

arb.

uni

ts

%C

2100 2200 2300 2400 2500 2600

0.0

0.2

0.4

0.6

0.8

1.0

[CH

3] /

arb.

uni

ts

Tf / K

• [CH3] vs %C and vs Tfil for 1% CH4/H2 ( ) and 0.5% C2H2/H2 (o), probed at d = 4 mm, plotted to match at 0.5% C and at Tfil = 2575 K.

• All measured [CH3] dependences are explicable in terms of gas phase chemistry.

19 Gas phase reaction mechanism - qualitative

CH4/H2 chemistry:

Radical formation: H + CH4 CH3 + H2

(‘H-shifting’ reactions) H + CH3 CH2 + H2, etc.

C1 C2 conversion: CH3 + CH3 + M C2H6 + M

C2H6 + H C2H5 + H2

…… C2H2 + H2

C2H2/H2 chemistry:

Analogous H + C2H2 C2H + H2 radical initiation step is endothermic.

Earlier models often invoked a role for heterogeneous chemistry (onsurface of HF, or on the growing diamond film), but spatially resolved [CH3] profiles led us to suggest that gas phase addition processes like C2H2 + H + M C2H3 + M

may occur in cooler regions of the reactor.

20 Gas phase reaction mechanism - quantitative

3-D modelling of the Bristol HF-CVD reactor. (Mankelevich and Suetin, Moscow State University)

Model consists of 3 blocks that describe:• activation (gas heating, H2 dissociation on filament)• gas phase processes (heat and mass transfer, reaction kinetics)• gas-surface processes at the substrate

Gas phase reaction kinetics and thermochemistry from GRIMECH 3.0 detailed reaction mechanism for C/H/(O/N) mixtures.

Conservation equations for mass, momentum, energy and number densities are integrated numerically until steady-state is achieved.

Model outputs include spatial distributions of gas temperature, flow field and species number densities.

21 Summary of elementary reaction rates

Reaction Rate / cm-3 s-1

Reaction 730 K 1200 K 1750 K 2000 K

H + C2H2 + M C2H3 + M 1.85E +16 4.00E +15 4.07E +14 1.64E +14

C2H3 + M H + C2H2 + M 3.72E +12 1.23E +15 1.93E +16 3.23E +16

H + C2H3 H2 + C2H2 1.05E +16 3.34E +15 5.24E +15 6.07E +15

H2 + C2H2 H + C2H3 7.15E +1 2.54E +9 8.58E +12 6.97E +13

Total: C2H2 C2H3 8.00E +15 5.73E +14 2.41E +16 3.82E +16

Total: C2H3 C2H4 7.87E +15 1.10E +14 2.42E +16 4.25E +16

Total: C2H4 C2H5 4.97E +15 3.45E +14 1.47E +15 1.53E +15

Total: C2H6/C2H5 CH3 4.97E +15 4.51E +15 3.75E +15 5.00E +15

Total: C2H6 C2H5 4.23E +14 4.45E +15 4.44E +15 3.48E +15Entries in red confirm the suggestion that, at low Tgas, there is net C2 C1 conversion via C2H2 C2H3 C2H4 C2H5 CH3.

22 CH4 C2H2 interconversion

-15

-10

-5

0

5

10

15

0 2 4 6 8 10 12 14 16 18 20

Filament distance, d / mm

Pro

du

ctio

n -

Lo

ss R

ate

/ 10

16 c

m-3

s-1

500

800

1100

1400

1700

2000

Tg

as

/ K

HCH3

C2H2

C2H6

C2H4

CH4

TGAS

-15

-10

-5

0

5

10

15

0 2 4 6 8 10 12 14 16 18 20

Filament distance, d / mm

Pro

du

ctio

n -

Lo

ss R

ate

/ 10

16 c

m-3

s-1

500

800

1100

1400

1700

2000

Tg

as

/ K

HCH3

C2H2

C2H6

C2H4

CH4

TGAS

lower Tgas; net C2H2 CH4

high Tgas; net CH4C2H2

23 Experiment and theory compared

1% CH4 / H2

0.5% C2H2 / H2

Tfil = 2475 K

modelexpt.

24 Conclusions from studies of HF-activated CH4/H2 and C2H2/H2 gas mixtures

• With CH4/H2 input mixtures, gas phase H abstraction reactions initiate the overall CH4 C2H2 conversion in regions of high Tgas near the HF.

• Diffusion rates are much faster than gas replacement rates in these reactors (typical gas flow rate ~100 sccm), so much of the C2H2 formed by CH4 C2H2 conversion near the HF will diffuse to cooler regions. (confirmed by cw CRDS monitoring of C2H2 rotational temperature and the time constants for its build up and decay – Wills et al., Diamond Rel. Mater. (2003), 12, 1346)

• H addition reactions in cooler regions of the reactor drive the reverse C2H2 CH4 conversion, offering a means of regenerating methane and eventual CH4 C2H2 equilibration.

• Purely gas phase processes can account for the observed C2 C1

interconversion in CH4/H2 and C2H2/H2 input gas mixtures.

25 Recent gas phase diagnostics studies in Bristol

Hydrocarbon / H2 mixtures in a hot filament (HF) reactor • Molecular beam mass spectrometry of stable species• REMPI laser probing H atoms and CH3 radicals

• Modelling CH4 / H2 and C2H2 / H2 gas mixtures (Ashfold et al., Phys. Chem. Chem. Phys. (2001), 3, 3471)

• Probing and modelling CH4/NH3/H2 gas mixtures (Smith et al., J. Appl. Phys. (2002), 92, 672)

CO2 / CH4 mixtures in a microwave (MW) reactor • Molecular beam mass spectrometry • Modelling H/C/O gas phase chemistry

CH4 / H2 / Ar mixtures in a DC arc jet reactor

• Cavity ring down measurements of C2H2 and of C2 and CH radicals

• Modelling plasma activated CH4 / H2 gas mixtures (Wills et al., J. Appl. Phys. (2002), 92, 4213 Rennick et al., Chem. Phys. Lett. (2004), 383, 518 Rennick et al., Diam. Rel. Mater. (2004), 14, 561)

26 Diamond film growth in a DC arc jet

How to probe gas phase chemistry and composition?

Optical emission spectroscopy (OES) and cavity ring down spectroscopy (CRDS).

10 kW DC arc jet

1%CH4 in Ar/H2 at 50 Torr

Growth rates ~100 m hr-1

Aggressive activation: much higher gas temperatures and flow rates than in HF or MW reactors.

27 DC arc jet in operation

diamond film growing on Mo substrate

plasma jet

CH4 injection ring

28 Diamond films grown with DC arc jet

SEM images of polycrystalline diamond films grown in the DC arc jet

29 Optical emission from the arc jet plume

What are primary growth species in this highly activated environment?

C atoms? C2 radicals? Latter show strongly in optical emission.

350 400 450 500 550 600 650 700

0

5000000

10000000

15000000

20000000

25000000

Em

issi

on in

tens

ity /

arb

. un

its

Wavelength / nm

H

C2 Swan system (d3g a3u)Spatially resolved C2(d-a) emission

30 Proposed mechanism for diamond growth by C2

C2 addition to H-terminated and to bare diamond (110) surfaces has been calculated to be barrierless and exothermic. (D.A. Horner et al. Chem. Phys. Lett. 233 (1995) 243)

31 In situ diagnosis of the arc jet plume, I

Optical emission spectroscopy (OES)• Only fluorescent species can be observed. • Provides information about the (minor) electronically excited components in plume – how to relate to ground state concentrations, properties, etc? • Spatially resolved measurements difficult.

Resonance enhanced multiphoton ionisation (REMPI)• Used successfully to probe ground state H atoms and CH3 radicals in HF reactor, but ion probe will not survive harsh plasma environment and background ion/electron signal would be a problem.

Laser induced fluorescence (LIF)• Species of interest must have fluorescent excited state.• Need to quantify excited state quenching characteristics in order to relate measured LIF signal intensities to ground state populations of interest. • Detector likely to be overwhelmed by intense spontaneous emission from plume.

32 In situ diagnosis of the arc jet plume, II

Absorption spectroscopy

Beer-Lambert behaviour

I = I0 exp{-[X] L}

Advantages:• Straightforward• General• Quantitative

Disadvantages:• Insensitive• Non-selective

Fractional absorption per pass

I = (I0 – I)/I0 10-4

33 In situ diagnosis of the arc jet plume, III

Intra-cavity absorption spectroscopy

Build cavity around sample•Multipass a light pulse•Detect rate of loss of light•Cavity ring-down spectroscopy

I(t) = I0 exp{-k0t - ct} ; = [X] ; I min ~ 10-8

Change in ring-down rate as a function of excitation wavelength gives the absorption spectrum

34 Cavity Ring Down Spectroscopy in the DC arc jet

Variables include: - CH4 flow rate - power into plasma- distance from substrate

35 C2(a) radical detection

Portion of C2 d3ga3u (0,0) band

pA)0v,a(Cgg

c8d 2

a

d2

line00

integrated absorptioncoefficient of measured line

A00 = Einstein A coefficient for vibronic transition of interest.

p = fraction of total oscillator strength within probed rovibrational transition (T dependent).

C2(a, v = 0) column density.

C2 (a) number density IF we know Tgas (and thus qvib) and the absorbing column length, L (from OES).

[C2(a3u)] ~ 1.1 1013 cm-3 for 3.3%CH4/H2 gas mixture, 6

kW input power, assuming Tgas = 3300 K and L = 1 cm.(

36 C2(a) radical detection – gas temperature determination

Boltzmann plots of C2(a) rotational state population distribution measured in the plume (2 < z < 25 mm) give Trot = 3300 200 K.

‘Doppler’ linewidth analyses give similar Tgas for z > 5mm, but overestimate Tgas close to the substrate – a consequence of plume flaring in the boundary layer.

probe

37 C2(X) radical detection

Tvib = 3000 500 K

[C2(X1g )] = (3.00.9) 1012 cm-3 again assuming L = 1 cm.

= 0.270.08

c.f. 0.23 if the a and X states of C2 were in thermal equilibrium at 3300 K – implies intersystem crossing is faster than reaction (with e.g. H2) under operational conditions.

Portion of C2 (D1uX1g) spectrum recorded in free plume at ~ 235 nm

(a)]2

[C

(X)]2

[C

38 CH(X) radical detection

Portion of CH A2X2 (0,0) band ~ 427 nm

Non-zero absorbance between peaksprobably attributable to C3 radicals.

[CH(X )] = (7.0 1.3) 1012 cm-3 in the free plume under normal operating conditions.(again assuming L = 1 cm).

39 C2(a) and CH(X) radical column densities as fn(z)

3% CH4/H2 , 6 kW input power, range of probe transitions

C2(a) CH(X)

40 C2(a) and CH(X) radical column densities as fn[CH4]

x sccm CH4 / 1.8 slm H2 / 12.2 slm ArArc jet power 6 kW, range of probe transitions

C2(a) CH(X)

41

ECDL AOM

photodiode

wavem eter

isolator

piezo mount fibre optic

fibre optic

trigger

V

V

CH / H / (Ar)4 2

CVD reactor

cw CRDS probing of C2H2 in the DC arc jet reactor

ECDL: Littman configuration extended cavity diode laser AOM: acousto-optic modulator

42 cw CRDS probing of C2H2 in the DC arc jet reactor

R(22) line of 1 + 3 combination band of C2H2

= 0.022 0.003 cm-1

(650 90 MHz).

pressure broadening: ~200 MHz at 50 Torrlaser bandwidth: ~4 MHz

Tgas = 550 150 K

C2H2 present along whole viewing column?

[C2H2] = 1.2 0.2 1014 cm-3 for 0.83% CH4/H2 feed (i.e. only 25% of our ‘standard’ CH4 flow rate) and assuming L = 100 cm

43 CRDS in DC arc jet: summary of experimental findings

Probing in the free plume region of the arc jet, with a CH4 flow of 60 sccm:

• [C2(a)] ~ 1.1 x 1013 cm-3,

• [C2(X)] ~ 3 x 1012 cm-3,

• [CH(X)] ~ 7 x 1012 cm-3 (all assuming L = 1 cm )• Tgas = 3300 200 K

• [C2H2] ~ 1.2 x 1014 cm-3 (using a reduced

(15 sccm) CH4 flow, assuming L =100 cm)

• Tgas ~ 550 K

• There is a boundary region close to the substrate, where C2 and CH column densities increase – due to plume flaring and the longer L?• Increased linewidths at small z mainly due to plume flaring. Internal quantum state population distributions of radical species suggest Tgas relatively insensitive to z. • [C2H2], and Tgas value (average over all L?) is insensitive to z in range 2 – 25 mm.

44 Modelling of the DC arc jet plume (Mankelevich)

• 2-D (r,z) model, comprising of three blocks, describing: (i) activation of the reactive mixture (i.e. gas heating, ionisation, H2 dissociation

in arc jet and intermediate chamber, H atom loss and H2 production on

nozzle exit walls), (ii) gas-phase processes (heat and mass transfer, chemical kinetics), (iii) gas-surface processes at the substrate.

• Thermochemical data and the reduced chemical reaction mechanism builds on Yu.A. Mankelevich et al., Diam. Rel. Mater. (1996), 5, 888.

• Chemical kinetics scheme involves 23 species (H, H2, Ar, C, CH, 3CH2, 1CH2, CH3, CH4, C2(X), C2(a), C2Hx (x = 1-6), C3Hx (x = 0-2), C4Hx (x = 0-2))

and 76 reversible reactions.

• Set of conservation equations for mass, momentum, energy and species concentrations, with appropriate initial and boundary conditions, thermal and caloric equations of state, are integrated numerically in cylindrical (r,z) coordinate space until attaining steady state conditions.

• Model output includes spatial distributions of Tgas, the flow field, and the

various species number densities.

45 Modelling of the DC arc jet plume: Tgas

• Gas temperature distribution, Tgas

substrate

methane injection ring

H2/Ar plasma enters here

46 Modelling of the DC arc jet plume: H

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.

47 Modelling of the DC arc jet plume: CH4

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring

48 Modelling of the DC arc jet plume: C2H2

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2

49 Modelling of the DC arc jet plume: C4H2

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2)

50 Modelling of the DC arc jet plume: C3

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2) and C3 radicals

51 Modelling of the DC arc jet plume: C2H

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2) and C3 radicals• larger CxHy species break down as [H] and Tgas increase in the vicinity of the plume C2H

52 Modelling of the DC arc jet plume: C2

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2) and C3 radicals• larger CxHy species break down as [H] and Tgas increase in the vicinity of the plume C2H, C2,

53 Modelling of the DC arc jet plume: CH

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2) and C3 radicals• larger CxHy species break down as [H] and Tgas increase in the vicinity of the plume C2H, C2, CH radicals

54 Modelling of the DC arc jet plume: C

• Tgas

• H: H2 >90% dissociated; high [H] at substrate.• CH4 injected through ring• rapidly converted to C2H2 • and to larger CxHy compounds (e.g. C4H2) and C3 radicals• larger CxHy species break down as [H] and Tgas increase in the vicinity of the plume C2H, C2, CH radicals and C atoms• C1Hy formation on axis requires high [H] and Tgas, and sufficient time for diffusion into core of plume

55 Summary of results from modelling arc jet plume

• Gas temperature and flow velocity distributions show a cylindrical hot plume with Tgas~3000-4000 K, in good accord with optical emission studies.

• Highly activated gas mixture. [H]/[H2] ratio just above the substrate is

~ 0.25 (cf ~0.01 in typical low power HF or MW PECVD reactors). Surface chemistry is dominated by H abstraction and addition reactions.

• Gas pressure is not uniform throughout the chamber - encouraging the recirculation needed to transfer hydrocarbon from injection ring into the hot plume.

• Numerous chemical transformations occur during this transport. Predicted number densities of C, CH, C2, C2H, C2H2 and C3 incident on the

growing diamond surface are all >1012 cm-3.

• Most, if not all, of these species must contribute to film growth given the high (~3%) utility of carbon source gas deduced experimentally by comparing observed film growth rates with the metered CH4 input.

56 Comparison with experiment: CH and C2(a)

• Model confirms that CH and C2 species are localised in the hot plume.

• Quantitative agreement between observed and modelled column densities and rotational temperatures (~3300 K). Larger Doppler width seen at small z due to flaring of plume along observation axis.

• C2H2 predicted to be present throughout reactor – consistent with observed number densities and low (~550 K) associated average ‘temperature’.

57 Present and Future Work

• CRDS monitoring of CH(X) radicals in HF activated CH4/H2 and C2H2/H2 gas mixtures, and comparison with model predictions

• Probing B atoms in HF activated B2H6/CH4/H2 gas mixtures (by 2+1 REMPI at ~346 nm and by CRDS at ~249 nm) with a view to unravelling gas phase chemistry involved in B doped CVD diamond growth.

• Ab initio calculation of relevant elementary B/C/H gas phase and gas-surface reactions (with Harvey, Bristol), and modelling of full gas phase reaction mechanism (with Mankelevich).

• Spatially resolved CRDS measurements in arc-jet reactor to explore radial profiles of radical species in the plume.

• Extension of CRDS studies to new high pressure microwave activated CVD reactor designed for optical diagnosis.

58 Acknowledgements

Andrew Orr-EwingPaul MayColin WesternKeith RosserJames Smith

Steve Redman Roland TsangJames PetherbridgeMark Wallace Jon Wills

Andrew CheesmanChris Rennick William Boxford Alistair Smith Dane Comerford

Yuri MankelevichNikolay Suetin (Moscow State Univ.)

59

Reaction Rate / cm-3 s-1

Reaction 730 K 1200 K 1750 K 2000 K

H + C2H2+ M C2H3 + M 1.85E +16 4.00E +15 4.07E +14 1.64E +14

C2H3 + M H + C2H2 + M 3.72E +12 1.23E +15 1.93E +16 3.23E +16

H + C2H3 H2 + C2H2 1.05E +16 3.34E +15 5.24E +15 6.07E +15

H2 + C2H2 H + C2H3 7.15E +1 2.54E +9 8.58E +12 6.97E +13

Total: C2H2 C2H3 8.00E+15 5.73E+14 2.41E+16 3.82E+16

 H + C2H4 H2 + C2H3 7.73E +13 7.99E +15 5.78E +16 9.71E +16

C2H3 + H2 H + C2H4 7.64E +15 7.86E +15 3.35E +16 5.47E +16

H + C2H3 + M C2H4 + M 3.12E +14 2.21E +13 8.58E +12 5.93E +12

C2H4 + M H + C2H3 + M 0 5.25E +6 1.15E +12 2.38E +13

Total: C2H3 C2H4 7.87E+15 1.10E+14 2.42E+16 4.25E+16

 C2H4 + M H2 + C2H2 + M4.14E +1 8.76E +10 5.94E +14 4.40E +15

H + C2H4 + M C2H5 + M 5.02E +15 6.14E +14 3.55E +13 1.20E +13

C2H5 + M H + C2H4 + M 3.15E +11 2.29E +14 1.47E +15 1.52E +15

H + C2H5 H2 + C2H4 4.32E +13 4.01E +13 3.13E +13 2.37E +13

H2 + C2H4 H + C2H5 0 2.50E +7 5.88E +10 4.25E +11

Total: C2H4 C2H5 4.97E+15 3.45E+14 1.47E+15 1.53E+15

 CH3 + CH3 + M C2H6 + M 4.22E +14 2.77E +14 3.30E +13 1.13E +13

C2H6 + M CH3 + CH3 + M 6.74E +2 5.09E +11 8.01E +14 3.02E +15

CH3 + CH3 H + C2H5 3.73E +11 3.25E +13 1.47E +14 1.80E +14

H + C2H5 CH3 + CH3 5.39E +15 4.82E +15 3.13E +15 2.18E +15

Total: C2H6/C2H5 CH3 4.97E+15 4.51E+15 3.75E+15 5.00E+15

 H + C2H6 H2 + C2H5 5.94E +14 4.55E +15 4.76E +15 3.86E +15

C2H5 + H2 H + C2H6 7.34E +12 8.57E +13 3.24E +14 3.97E +14

H + C2H5 + M C2H6 + M 1.63E +14 1.70E +13 1.50E +12 4.98E +11

C2H6 + M H + C2H5 + M 0 2.10E +8 1.71E +12 1.10E +13

Total: C2H6 C2H5 4.23E+14 4.45E+15 4.44E+15 3.48E+15

60

61

62

63