Surface Chemistry and Reaction Dynamics of Electron Beam Induced Deposition Processes

Transcription

1 Surface Chemistry and Reaction Dynamics of Electron Beam Induced Deposition Processes e -? 2 nd FEBIP Workshop Thun, Switzerland 2008 Howard Fairbrother Johns Hopkins University Baltimore, MD, USA

2 Outline Background / Motivation Experimental Approach Analytical techniques and electron source Surface Chemistry and Kinetics Electron Stimulated Dissociation Mechanism Summary & Outlook

3 Electron Beam Induced Deposition Electron Beam Organometallic precursor Low Pressure Substrate The ability to focus electron beams into small spots, control electron beam fluence and raster the beam makes EBID an ideal method for growing a wealth of different nanostructures

4 Motivation The fundamental surface processes that are responsible for electron beam induced deposition of nanostructures are not well understood Many questions about EBID process Chemical reactions at the surface? σ reaction (E)? If we can better understand the chemistry, we can: Choose precursors more selectively Improve deposition purity (carbon) Improve purification techniques Increase metallic characteristics

5 Outline Background / Motivation Experimental Approach Surface Chemistry and Kinetics (500eV) Electron Stimulated Dissociation Mechanism Summary

6 Our Approach To understand the EBID process using well established surface analytical techniques Adsorbing a nanometer scale film of EBID precursor to a substrate provides a clean environment for in situ observation Surface coverage can be controlled An UHV environment enables analysis of gas phase products A film, on the order of cm 2 in area, can be analyzed using common surface analytical techniques

7 Broad Beam Surface Irradiation Electron Beam Production of a film over a large surface area enables traditional surface analytical techniques to probe the EBID process Gold Substrate (~195K) Gold Substrate (~195K) Why gold?

8 Electron Source: Flood Gun Why use a flood gun? Uniform electron beam over a wide area (necessary for XPS and RAIRS) High target current Relatively broad range of Energies (40 500eV) Extractor Adjustable Filament Current e - Wehnelt Tungsten wire Electron Flood Gun can produce: (a) ev electrons (b) μa target currents

9 Sample Type K thermocouple Au Substrate (~195 K) Manipulator Cu leads provide for heating and cooling (100K-450K)

10 Instrumental Techniques X-ray Beam QMS Electron Beam Electron Analyzer FT-IR Beam hν hν neutrals e - e - hν IR Detector Gold Substrate (~195K) We have studied the electron stimulated reactions of the wellknown precursor, Trimethyl(methylcyclopentadienyl)- platinum(iv), adsorbed onto gold using the above techniques:

11 X-ray Photoelectron Spectroscopy (XPS) Ejected Photoelectron hν Vacuum Level 2p 2s 1s hν Parent KE BE Product BE = hν -KE (4f) Electron Irradiation Time (sec) Counts (a.u.) Au(4p) 600 (4p) Au(4d) 400 Binding Energy (ev) (4d) C(1s) Au(4f) (4f) 200 XPS enables quantitative determination of chemical composition and effective oxidation state 0 Counts (a.u.) Binding Energy (ev) Reduction of indicated by peak shift to lower BE

12 Reflection Absorption Infrared Spectroscopy (RAIRS) hν hν FT-IR Beam Ultra-High Vacuum IR Detector Reflective Au ν(c-h) Electron Irradiation Time (sec) IR Intensity (a.u.) ν(c-h) 0.01 Abs MeCpMe 3 vibrational structure IR Intensity (a.u.) Abs Wavenumber (cm -1 ) Wavenumber (cm -1 )

13 Mass Spectrometry (MS) MS was used to: Verify purity of organometallic precursor Observe gaseous products of EBID Counts (a.u.) H 2 HC CH 2 (g) Ensure cleanliness of UHV H 2 O CH 2 CH 2 H 2 C HC CH CH Mass (a.m.u.) (g) Electron Ionization (70eV) M + + 2e - M + Mass Selector Detector Electron bombardment within the MS creates ionized fragments representative of species present in the gas phase

14 Outline Background / Motivation Experimental Approach Surface Chemistry and Kinetics (fixed electron energy = 500eV) Electron Stimulated Dissociation Mechanism Summary

15 Adsorption of MeCp(IV)Me 3 onto Gold Substrate Controlling film thickness Film thickness, d, calculated from attenuation of Au(4f) signal 7 6 Influence of Dosing Time on Film Thickness P dosing = 1x10-6 Torr d I = λ*cos( θ)*ln( ) I d = adsorbate thickness λ = ~2nm for Au(4f) photoelectron θ = 54 photoelectron take-off angle I = Au(4f) area 0 Adsorbate Layers XPS and MS Experiments d ave =.96nm Conducted in this Regime Dosing Time (seconds) Au

16 MeCpMe3 Cx Production of Amorphous Platinum/Carbon Film (4f) XP spectra of sputter deposited platinum on gold Au Counts (a.u.) Substrate heated to room temperature leaving electron beam irradiation product Electron beam irradiation for 20 sec (20μA, 500eV) C x C x Au + Au MeCp(IV)Me 3 adsorbed onto gold substrate (~195K) Au Binding Energy (ev)

17 Influence of e - Beam Irradiation on Surface Composition of Adsorbate Layer The XP spectra of the C(1s), Au(4f), and (4f) regions shows: Irradiation Time (sec) C(1s) Au(4f) MeCpMe3 CX (4f) No change in film thickness as determined by Au(4f) attenuation Shift in platinum environment from precursor to product Counts (a.u.) 15 0 MeCp(IV)Me 3 is stable under x-ray irradiation for >2hrs Binding Energy (ev)

18 XPS Analysis of Deposition Kinetics ln [(4f) t ]/[(4f) t=0 ] [(4f) t ]/[(4f) t=0 ] Irradiation Time (sec) Cx MeCp(IV)Me3 Target Current 10 ma 50 ma 100 ma Quantification of the deconvoluted (4f) region fit to exponential decay shows first order kinetics Irradiation Time (sec) Decay profiles show that the observed rate constant increases with increasing target current.

19 Signal Resolution vs. Signal Intensity Irradiation Time (sec) C(1s) Au(4f) MeCpMe3 (4f) CX 1200 High resolution / low intensity emphasizes elemental environment Low resolution / high intensity emphasizes stoichiometry Counts (a.u.) Binding Energy (ev)

20 Evidence for Carbon Loss During Irradiation 11 MeCpMe 3 Film C x Film 10 C: (XPS Ratio) 9 8 C 9 C 8 Electron Beam Irradiation C: (XPS Ratio) 9 8 C 9 C 8 7 C 7 7 C Film Thickness (nm) The C: ratio of 10 XP spectra of the MeCp(IV)Me 3 prior to e - beam irradiation is representative of the initial stoichiometric ratio of 9 carbon atoms to 1 platinum atom The stoichiometric loss of 1 The C: carbon ratio after atom e as - beam a result of C 9 irradiation irradiation decreases is to independent a ratio of 8:1, of 500 ev e indicating film the thickness loss of 1 carbon atom - (ads) per molecule as a result of irradiation Au (~195 K) C 8 Au

21 Gas Phase Products H 2 HC CH 2 (g) Electron beam irradiation results in the production of hydrogen and methane (ads) Au (~195 K) 500 ev e - CH 4 (g) / H 2 (g) Counts (a.u.) H 2 O CH 4 CH 2 CH 2 H 2 C HC CH CH 2 MS capable of measuring individual masses as a function of time Au e - Methane Reference Mass (a.m.u.)

22 Kinetic Analysis of Methane Production MS signal (arb. units) Electron irradiation m/z = Irradiation time (sec) Tracking methane production during electron beam irradiation fit to exponential decay m/z = 15 is a unique mass representative of methane ln[(ch 4 ) t /(CH 4 ) t=0 )] Target Current μ 9.88 A μ A A μ Methane loss fit to first order kinetics indicates an increase in the observed rate constant with increasing target current Irradiation Time (sec)

23 Complementary Techniques k obs (s 1 ) 6.0e-2 5.0e-2 4.0e-2 3.0e-2 2.0e-2 1.0e-2 XPS Gradient = k obs / I target k I obs σ = XPS MS (m/z=15) targ et A σ (cm 2 ) 1.40E E e-2 MS k obs (s -1 ) 2.5e-2 2.0e-2 1.5e-2 1.0e-2 5.0e Target Current (μa) σ ave, 500eV = 2.28E-16 cm 2 XPS, RAIRS, and MS provide similar σ values though they measure different processes μ

24 How does Film Thickness Influence the Process? 3e-16 Methane Area σ (cm 2 ) 2e-16 1e Adsorbate Layers Adsorbate Layers Methane area independent of film thickness σ independent of film thickness

25 Surface Chemistry CH 4 (g)/h 2 (g) containing amorphous C films (ads) Au (~195 K) 500 ev e - One electron r.d.s C 8 Au Electron beam irradiation of surface adsorbed MeCp(IV)Me 3 results in the formation of platinum atoms embedded in an amorphous carbon film via an electron impact process in which bond cleavage releases hydrogen and methane. Each precursor molecule that undergoes electron stimulated decomposition losses exactly one carbon atom.

26 Outline Background / Motivation Experimental Approach Surface Chemistry and Kinetics (500eV) Electron Stimulated Dissociation Mechanism Summary

27 Dependence on Incident Electron Energy Observed Using XPS Electron Beam Energy 1.6e e eV 1.2e e-16 Counts (a.u.) 750eV 200eV σ (cm 2 ) 8.0e e e eV 40eV Binding Energy (ev) Above XP spectra: 25μA, t=60 sec 2.0e Electron Energy (ev) Optimal σ between eV suggests electron impact ionization process

28 Dependence on Incident Electron Energy Observed Using MS 2.5e ev 200 ev 2000 ev 2.0e-16 ln ( CH 4 / CH 4(t=0) ) σ (cm 2 ) 1.5e e e Electron Energy (ev) Irradiation Time (sec) σ 60eV < σ 200eV > σ 2000eV MS supports XPS results

29 Dissociative Ionization Process + Decomposition products

30 Substrate Dependence? 4e-16 3e-16 /C Substrate Graphite Substrate Gold Substrate Silicon Substrate σ (cm 2 ) 2e-16 1e Adsorbate Layers NO All experiments conducted at ~195K with 20μA target current and 200eV electron energy

31 Summary one electron r.d.s process + CH 4 (g)/h 2 (g) atoms embedded in an amorphous C film C 8 Au A UHV surface science approach can provided valuable information on reaction rates and fundamental chemical processes involved in EBID

32 Exptl. Requirements Substrate - Chemically unreactive towards precursor Film - Thin!!! (1-3 Monolayer regime so that each precursor molecule experience the same electron flux) Electron Beam Broad and defocused (uniform irradiation)

33 Methods? Technique Pros Cons XPS Reasonable quantification Central atom must change oxidation state AES Destructive (Not applicable) Mass Spectrometry (a) TPD Good Quantification Duty Cycle Slow Decomposition may compete with desorption (b) Analysis of Simple Indirect monitor of desorbing species reasonable quantification surface processes IR (reflection) Parent can easily be monitored Poor Sensitivity Modest Quantification

34 Acknowledgements Joshua Wnuk, Johns Hopkins University Justin Gorham, Johns Hopkins University Willem van Dorp, Rutgers University Ted Madey, Rutgers University Kees Hagen, Delft University of Technology