Secondary Electron Emission Properties Molybdenum ... · The secondary electron emission ... secondary electron ... Comparison of XRD spectra for samples G and H. MoS2 films produced

Secondary Electron Emission Properties of Molybdenum Disulfide Thin Films

Pierre Richard Sébastien Cormier

A thesis submitted to the Department of Physics

in codormity with the requirements for

the degree of Master of Science

Queen's University

Kingston, Ontario, Canada

September, 1998

QPierre Richard Sébastien Cormier

National Library Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services senrices bibliographiques 395 Wellington Street 395, rue Wellington OttawaON KlAON4 Ottawa ON Kt A ON4 Canada Canada

The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distniiute or sel copies of this thesis in microform, paper or electronic formais.

The author retains ownership of the copyright in this thesis. N e i k the thesis nor substantial extracts fkom it may be printed or otherwise reproduced without the author's permission.

L'auteur a accordé me licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distriiuer ou vendre des copies de cette thèse sous la forme de microfiche//6iùn, de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou amernent reproduits sans son autorisation.

Abstract

As higher demands are being placed on materiais for new technology, the

e n g i n e e ~ g of secondary electron emission properties of materials is becoming more

important. Sputtered thin f i h of MoS, with acicdar surface rnorphology were found to

lower the secondary electron emission yield of aiuminum.

The sputtered thin films of MoS, were created with theu characteristic acicular

morphology and crystal structure by optimizing sputtenng parameters and submte

preparation. After characterizhg the films through x-ray difhction @RD) and scanning

electron rnicroscopy (SEM), secondary electron yield (6) properties as a function of

primary electron energy &) were measured using a modified JEOL840 SEM.

The secondary electron emission characteristics of thin films of MoS, were shown

to be a function of the acicular morphology's effect on the path of the secondary electrons

rather than of the crystal1ographic structure. The overall effect was to reduce the

secondary electron coefficient (6) of the aluminum abstrate by approximately 40%.

Acknowledgements

1 would fmt like to thank my supervisor Dr. Sayer for his guidance and honesty

when directing my work as well as his patience and understanding dining the dificult

times.

Many of my colleagues in the physics department deserve my gratitude: Sarah

LangstafT for her always uplifting fnendship and merciless editing, Brian Leclerc for his

handiwork and for showing me that I forgot to plug it in sometimes, Guofang Pang for his

incredible work with the sputtering machine, and Lichun Zou for showing me the

sputtering ropes. 1 would also like to thank Paul Nolan, Marc Lukacs, Katia Dyrda,

Liping Sun, Alyssa Markowitz, Moira Grunwell, Yan Chen, Tim Olding, Margaret

Moms, Teny Busse, and Douglas Morren.

1 would also like to th& Teer Inc. for theu invaluable 'hot tip' which saved Dr.

Sayer a large s u m of money and saved me a large amount of work. Thanks are also due

to Aican for donating the 'quirky' sput te~g machine 1 came to lovehate.

Special thanks to Iohan Sebastian Bach and Pearl Jam for providing me with six

cello suites and 'Ten' sweet songs, respectively, to soothe and caim me during those

final heiiish days.

Lip, for king Lip. Mes parents que j ' h e tant, je vous remercie pour qui je suis,

pour votre amour, et pour tout votre support sans fi.

Thank you Dreeni, for your love and for keeping my song playing. 1 love you.

Table of Contents

.................................................... 1 . ~ O D U C T I O N 1 ....................................................... References 4

2 . BACKGROrn ..................................................... 5 ..................................... 2.1 Secondary Electron Emission 5

2.1.1 Elastic and Inelastic ScatteMg ............................. 5 ......................................... 2.1.2 Escape Depth - 6

...................... 2.1.3 Escaping Electron Energy Distribution 7 ............................ 2.1.4 Electron Emission Coefficients - 8

.......................................... 2.2 Molybdenum Disulfide 9 ...................................... 2.2.1 Structure of MoS, 11

............................................... 2.3 Industriai Work 13 ...................................................... References 15

3.EXPERIMENT ..................................................... 16 ......................................... 3.1 Overview of Sputtering 16

............................................ 3.2 Sample Preparation 19 ........................................... 3.3 Sputterhg Procedure 19

................................... 3.1.2 Sputtering Parameters 20 ...................... 3.1.3 Optimization of Sputtering Parameters 21

................................... 3.4 Characterization of Thin Films 21 .......................... 3.4.1 X-Ray Difiction Experiments -22

................................... 3.4.2 Talystep Profilorneter 22 ...................... 3.4.3 Scanning Electron Microscope (SEM) 23

3.5 Secondary Electron Emission Measurements Using a SEM ............. 23 ..................................... 3.5.1 Sample Preparation 23

......... 3 .5.2 Secondary Electron Emission Measurement O v e ~ e w 24 ...................................................... References 27

................................................ 4 . FILM PRODUCTION 28

............................................... 4.1 Deposition rate -28 ......................................... 4.2 Silicon ûxide Substrate 31 ........................................ 4.3 Stainless Steel Substrate 32

4.4 Aluminum Substrate ........................................... 35 4.4.1 Determining the Effect of Substrate Temperature ............. 38

................... 4.4.2 DetemÜning the Effect of Argon Pressure 40 ....................... 4.4.3 Detennining the Effect of RF Power 43

................. 4.4.4 Determining the Effect of Sputtering Time -46 ........................ 4.5 Mechanical A l i m e n t ofCoating Structure -49

.................................... 4.6 Summary of Film Production 52 ...................................................... References 52

Table of Contents

.................................... 5 . SECONDARY ELECTRON YIELD 53 ........................................... 5.1 Detennination of V , 53

. . . . . . . . . . . . . . . . . . . . . . . . 5.2 Calibration rneasurements and Stainless Steel 55 . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Caiibration measurements on Aluminum 57

............... 5.4 Secondary Electron Yield Measurements of MoS2 fihs 60 5.5 Secondary Electron Emission Yield Measurements of Bumished MoSZ films

......................................................... 64 ...................................................... References 65

.................................................... 6 . CONCLUSIONS 66

... .................................. APPENDIX A: Representative Data Set vui

VITA

List of Figures

Figure 2.1: Primary electrons (PE) elastically scatter to becorne refiected electrons (RE). Both primary and reflected electrons will produce secondary electrons (SE)

... which, if within the average escape depth (A), can escape firom the material .6 ........................... Figure 23: Energy Spectrum of Escaping Electrons .7

Figure 23: Typical graph for secondary electron yield (6) vs ptimci~y electron energy (Ep) ............................................................. 9

Figure 2.4: Planar hexagonal structure of magnetron sputtered MoS, crystallites. [3] . 10 Figure 2.5: Platelets of magnetron sputtered MoS2 will have its platelets organized in

planes in a combination of 2 types: type I planes at an angle to the substrate and ................................. type II planes p d e l to the substrate. 12

Figure 2.6: X-ray specûa of thin films of MoS, produced by magnetmn sputterhg. If the two peaks (type I at 14" and type II at 34") are of equal intensity, there will be approximately 100 times more type I than type II. Therefore (a) represents a

......... predominantly type 1 film and (b) a predominantly type II film.[7] -12 Figure 3.1: Representation of magnetron sputtering process whereby plasma ions strike

................... a target to fiee material for deposition on the substrate. 17 ............................ Figure 3.2: MRC Mode1 8667 Sputtering System .18

Figure 3.3: Secondary electron yield memement using an SEM such that current absorbed by the stage may be measured as a function of the bias voltage on the stage. .......................................................... 25

Figure 3.4: Schematic diagram representing the experimental set-up used for m e a s u ~ g ......................................... the total incorning cumnt -26

Figure 3.5: SEM micrograph of the hole in the aperture grid. The hole is black because ............................................... no electrons escape .26

Figure 4.1: Deteminhg the effect of argon pressure and RF power on deposition mte ............................................................. 30

Figure 43: Sarnple A: XRD Spectra of MoS, film on stainless steel. The peak at 2 8 , e O is due to the substrate. The broad, low intensity peak centered at 28,.=34" is indicative of a weakly orgarkd type 1 MoS2 ............... .33

Figure 4.3: Sample B: XRD spectra of Type 1 MoSz on stainless steel. AI1 peaks are indicative of the substrate except for the broad peak at 28,,,=34" which is due to type 1 MoS2 Note that this is a stronger peak than that in Figure 4.2. All other peaks present are due to the stainless steel substrate. .................... .34

Figure 4.4: SEM micrographs Uustrating the substrate surface at ~10,000 magnincation: (a) ComDev duminum substrate and @) polished aiuminum substrate ...... .37

Figure 4.5: XRD spectra of sampies C and D. MoS, on aluminum whose sputtering conditions are identical (1 OOW, 1 OmTorr and 0.5hr) aside f h m substrate temperature. Note that the intensity of the Type 1 MoS, peak at 28,,=34" increases as the temperature increases. The XRD spectra for the 300°C sarnple has been offset by 15CPS for clarity. ................................. .39

List of Figures

Figure 4.6: XRD spectra of samples D, E and F. MoS2 on aluminum whose sputterhg conditions are identical (1 00 W, 03.. and 300°C) aside from argon pressure. Note that the intensity of the Type 1 MoS, peak at 20,u=340 decreases as the argon pressure decreases. For clarity, the XRD spectra for the 2OmTorr and

........... 30mTorr samples are offset by 20CPS and 40CPS, respectively. . 4 1 Figure 4.7: SEM micrographs illustrating the ef5ect argon pressure on the surface

morphology: (a) Sample D: IOmTorr, (b) Sample E: 2OmTon and (c) Sample F: 30mTorr. Al1 of the sarnples were sputtered for 0.5 hours at 1 OOW

...................................................... and300°C 42 Figure 4.8: Comparison of XRD spectra for samples E and H. MoS, films sputtered at

different RF power values. The sarnple sputtered at lOOW has a larger arnorphous background, with evidence of crystallization. The sample sputtered at 200W shows increased crystallization. For clarity the spectra for the sample sputtered at

.................................... 200W has been offset by SOCPS. .44 Figure 4.9: SEM micrographs illustrating the surface morphology of MoS2 films

sputtered at an argon pressure of 20mTorr and substrate temperature of 300°C for (a) Sarnple E: 0.5 hours at 100W and @) Sample H: 0.25 hours at 200W. .. .45

Figure 4.10: Comparison of XRD spectra for samples G and H. MoS2 films produced with same sputtering conditions (ZOOW, 20mTon and 300°C) except for different lengths of time (0.25 hours and 0.5 hours). The decrease in the Type 1 peak at 28,,=34O with sputtering time corresponds to a decrease in crystallization. The narrowing of the peak indicates an increasing in grain size with increasing sputtering t h e . The sample sputtered for 0.5 hours has been offset by 6OCPS for

........................................................ clarity. -47 Figure 4.1 1: SEM micrographs illustrahg the surface morphology of MoS, sputtered at

an RF power of 200W, an argon pressure of 20mTorr, a substrate temperature of 3 OO°C and a sputtering time of (a) Sample H: 0.5 hours and (b) Sample G: 0.25 hours. .......................................................... 48

Figure 4.12: Comparison of XRD spectra for the same sample prior to and after burnishing with a cloth (Kimwipe'%X-L). There is a very small decrease in the type 1 peak at 2Oc,=34O. The burnishing produces a peak at 2ec5 Mo which shows a small amount of Type II MoS2 has been produced. For clarity the burnished spectnim has ken offset by 6OCPS. ......................... -50

Figure 4.13: SEM micrographs illustrating the sdace morphology of a sample sputtered for one hour at an RF power of 1 SOW, an argon pressure of 2OmTorr and a substrate temperature of 300°C (a) prior to and (b) after bumishing. ... -51

Figure 5.1: Detennining the necessary bias voltage to be applied to the stage to ensure the capture of ail secondary electrons. ................................ .54

Figure 5.2: The secondary electron yield (6) as a function of primary electron energy (EJ for stainless steel. ............................................ .56

List of Figures

Figure 53: The secondary electron yield (6) as a fiinction of primary electron energy (E& for polished and ComDev duminum ............................. . 58

Figure 5.4: Schematic diagram iilustrating the path of secoiidary electrons in materials with either a smooth or rough surface. Secondary electrons created in materials with a rough surface will have more surface area through which to escape than a

................................................. smooth swface. .59 Figure 5.5: Graphs of secondary electron yield as a Wction of primary electron energy

for MoS2 films sputtered on ComDev alurninum substrates at (a) Sample C: lOOW, lOmT, 1/2hr, 300°C, (b) Sample E: 100W, 20mT, 1/2hr, 300°C, (c) Sample F: 100W,30mT, l/îhr,300°C, and (d) a cornparison of ail three res ul t s . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . .+ . . . . . . . . . . . . . 6L

Figure 5.6: SEM micrographs illustrating the substrate and dBerent MoS2 film surface morphologies created using the MRC sputte~g system at an RF power of 100W, a substrate temperature of 300°C and a sputtering t h e of % hr: (a) ComDev duminuni substrate, @) 1 OrnTorr argon pressure, (c) 20mTorr argon pressure and

....................................... (d) 30mTorr argon pressure. .62 Figure 5.7: Secondary electron yield (6) vs primary energy @$for MoS2 on Al for type I

and bumished type 1 (The thick curve represents 6 for the aluminurn substrate) ............................................................... 65

List of Tables

Table 2.1: 6- and E,- for various elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 Table 3.1: Sputtering Parameters Suggested nom the Literature . . . . . . . . . . . . . . . . - 2 1 Table 4.1: Deposition Rates Measured as a Function of Argon Pressure and RF Power29 Table 4.2: Surnmary of Results for Samples Prepared on SiOz Substrates. . . . . . . . . . 3 1 Table 43: Summary of Results for Samples Prepared on Stalliless Steel Substrates. . -32 Table 4.4: Summary of Results for Samples Prepared on Aluminum Substrates. . . . . .35 Table 5.1: Comparison of Experirnenttil and Published values of 6- and EPmax for

Stainless Steel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Table 5.2: Comparison of experimental and published values of 6- and Epm for

duminum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . 5 7

CL* MO& M O S T MRC mTorr 11111

PE RE RF SE SEE SEM

List of Symbols and Abbreviations

degrees Celsius XRD 20 assuming Cu & as the incident radiation broadening of the difiction iine (full width at half maximum) total electron yield secondary electron yield maximum secondary electron yield electron incident energy primary electron energy E, at which 6, occurs electron volts gallium arsenide primary current minus reflected and secondary electron current primary current minus reflected electron current total incoming current secondary electron current 1 O3 electron volts mean escape depth x-ray wavelength (Cu K,J 1 o6 s-' I O4 metres molybdenum disulfide MoS2/titanium coatings patented by Teer Coatings Ltd. Materials Research Corporation 105 T O ~

metres primary electrons reflected electrons radio fiequency secondary electron secondary electron emission scanning electron microscope reflected electron yield maximum escape depth grain sue angle Watts 103 Watts X-ray difiaction atomic number

vii

Appendix A: Representative Data Set

Sample data for measurement of 6 for polished aluminum surface:

1. INTRODUCTION

The secondary electron emission properties of matends have been manipulated by

scientists for several decades. The ability of a material to give up secondary electrons

depends on a variety of surface and bulk properties such as surface morphology,

con taminant layers, bullc density, conductivity and atornic number. When a material's

secondary electxon emission properties are undesirable, it is often dificult to find a

replacement material and alternative methods must be found. Quite ofien these methods

involve depositing thin films with appropriate secondary electron emission properties

over the material. This thesis investigates the secondary electron emission properties of

sputtered thin films of MoS, for the purpose of secondary electron emission reduction. It

originated in a need by ComDev Ltd (Cambridge, ON) to reduce the potential for spark

discharges in communications satellites due to the build up of surface charges on

materials located near high power rnicrowave transmitters.

The manipulation of secondary electron emission necessarily takes one of two

paths. The fkst path is increasing the number of electrons produced. One such example,

is the use o f magnesiurn oxide (Mgû). Due to its high secondary electron emission yield,

thin films of Mg0 have been used as a dynode in electron multiplies [l] and also to

protect electrodes in plasma display panels [2]. Another example of increasing secondary

electron emission is utilised in scanning electron microscopy (SEM). The doping of

gallium arsenide (GaAs) by silicon increases the secondary electron emission properties

optimizing the observation of GaAs in the SEM [3]. The second path is the attempt to

1

1. INTRODUCTION

lower secondary electron emission Much work bas been done to suppress secondary

electron emission in wdls of plasma containers because the production of electrons cools

down plasmas. Recentiy, an obliquely incident magnetic field was used to mppress

electrons fiorn the container surface [4]. Carbon coatings have also been also used to

reduce the secondary electron emission yield [SI.

nie most promising and practical method of secondasr emission control is placing

a thin film on the surface. The small average escape depth of electrons, 30-70nm in

metals and 1 0- 1 0 0 ~ 1 in insulators [6], makes thki films an effective secondary electron

emission control mechanism. Thin films play an important role in both science and

industry. Their thickness, which is of the order of a micron, makes it possible to change

d a c e properties while maintahkg the bulk properties of the material. Simple

examples of these changes include changing the colour or the rougbness of a d a c e . It is

also possible to protect a bulk matenal from heat or radiation with an appropriate

insulating coatîng. Thin films of molybdenum disuifide (MoS3 have traditionally been

used as a dry lubricant for various inert environrnents such as vacuum or space

applications [7][8]. The excellent lubricant properties of MoS, in particuiar, are

atûibuted to its layered crystal structure. Under ideal conditions, the structure is

cornposed of 'platelets' (Figure 2.4) which can easily slide over each other. The film is

strongly bonded within the plane and has only weak van der Waals bonding between the

planes. It is this structure that makes the investigation of secondary electron emission

1. INTRODUCTION

properties of MoS, films potentially interesting. Secondary electrons created within the

MoS, must interact with several individual layers to escape. Aligning the platelets

parallel or perpendicular to the substrate wouid coustitute a method for studying the

possible effects of this layered structure on the secondary electron emission properties.

The goal of this project was to determine how the structure of a thin film of MoS,

would affect the secondary electron emission (SEE) properties of a buik material. It was

hypothesized that a layered structure of platelets parallel to the surface would posses a

lower secondary electron yield than films with platelets perpendicular to the surface. It

was necessary to ascertain if the material could be deposited in a strongly textured form.

This study involved the creation and characterization of thin films of MoS,, the

measurement of the secondary electron emission properties, and the cornparison of the

results of layers of parallel and perpendicular MoS,. In the end, it was hoped that the

secondary electron emission properties of a matenal could be tailored by varying the

sputtering parameters. While the initial objective was to create strong parallel

orientation, the magnetron sputtering system used proved more capable of creating the

perpendicular orientation. This study has therefon emphasized this orientation.

The thesis is structured in the foiiowing manner. The theory of secondary electron

emission and its measurement are discussed in Chapter 2. Earlier work on the sputtering

of MoS, is also presented in this chapter. Chapter 3 describes both the experirnental

procedure for film deposition by RF magnetron sputtering and the scanning electron

1. INTRODUCTION

microscope based experimental setup for the measurement of secondary electron

coefficients. Film production results are discwed in Chapter 4 and secondary electron

emission measurements are discussed in Chapter 5. Finally, conclusions about the

secondary electron emission properties of MoS, are summarised in Chapter 6.

References

N.R. Whetten and A.B. Laponsky, Physical Review, lO7(1957) p 1521.

T. Urade, T. Iemori, M. Osawa, N. Nakayama and 1. Monta, IEEE Trans. EElectrorz Devices, 23(lW6) p3 1 3.

F. Iwase, Y. Nakmura, Applied Physics Letters, 7l(l S), 13 Oct. 1997, p 2 142.44

S. Takamura, S. Mizoshita and N. Ohno, Physics of Plasmas, 3(12), Dec 1996, p 4310-12

D. Ruzic, Journal of Vacuum Science and Technologv, 20(4), April 1982, p 13 13- 1315

H. Seiler, 1 983, Journal ofApplied Physics, 54, p 1 - 1 8 .

T. Spalvins, Journal of Materials Engineering and Performance, 1(3), June 1992, p 348-51

E.W. Roberts, B.J. Williams, and J.A. Ogilvy, Jomal of Physics D: Appiied Physics, 25(1992) p A65-A70

2. BACKGROUND

2.1 Secondary EIectron Emission

When electrons impinge on a surface, various types of collisions take place which

change their paths and energies as welt as create new free electrons within the material.

2.1.1 Elastic and Inelastic Scattering

Consider 'primary' electrom impinging on a surface with incident energy ET

Once inside the material, the electrons wiU scatter either elastically with nuclei or

inelastically with other electrons. Elastic scattering results in the path of the primary

electrons changing while leaving the energy relatively constant at E,. Some of these

electrons will scatter back towards the surface, escaping from the material and are thus

tenned 'reflected' or 'backscattered' electrons.

in inelastic scattering, the primary electron imparts energy to electrons within the

material. This acceptauce of energy creates an excited state which quickly retums to the

ground state by emitting an electron or photon. The excitation of a valence electron

results in an emitted 'secondary electron' which is usuaIIy in the energy range of 0-50 eV.

When a core electron is excited, which has a much lower pr~bability~ it is cailed an

'Auger electron' . Both the primary and reflected electrons can produce secondaq and

Auger electrons.

2. BACKGROUND

Figure 2.1 schematicaily depicts the basics of production of reflected and

secondary electrons nom the primary electrons. A large number of secondary electrons

are produced as the primary and reflected electrons move within the material.

Figure 2.1: Prirnary electrons (PE) elastically scatter to become reflected electrons (RE). Both primary and reflected electrons will produce secondary electrons (SE) whicb, if within the average escape depth (A), can escape nom the material.

2.1.2 Escape Depth

Only those secondary electrons created near the d a c e have a substantial

probability of escape. For metals, the mean escape depth (A) has been shown to be on the

order of L = 30-70nm; for insulators Iç = 10-100 nm [Il. Metals are conductive and

readily emit secondary electrons whereas insulators absorb the primary electrons to

accumulate a negative charge that wiU eventually repel the primary k m .

2. BACKGROUND

2.13 Escaping Electroa Energy Distribution

Figure 2.2 depicts a typical energy spectnim of escaping electrons. The lowest

energy peak is due to the secondary electrons which escape the material with a wide

range of energies normdy nom 0-50 eV [l]. The highest energy peak, centred around

the primary electron energy E, cornes fkom the reflected elecüons. Reflected electrons

that were inelasticdly scattered fonn most of the background intensity. The small peaks

in the middle represent i) Auger electrons and ii) reflected electrons which are

backscattered with an energy Ioss by volume, surface plasmon excitation or ionization of

d a c e atoms [2]. Although used in Auger electron spectroscopy, the intensity of Auger

electrons and the extra reflected electrons is low enough to have a negligible effect on the

engineering control of secondary emission.

Figure 23: Energy Spectrum of Escaping Electrons

2. BACKGROUND

2.1.4 Electron Emission Coefficients

In order to investigate the electron emission properties of a material, it is usefil to

d e h e the following tems. The total number of electrons which escape the surface as a

hct ion of the primary b a r n is the total electron yield (d) which is given by:

- Secondary and Reflected Electrons .- . . d = -

Cd. 1) Primary Electrons

To characterize the behaviour of reflected electrons, one would define the reflected

electron yield (a) as: e

Reflected Electrons CY =

Primary Electrons (2.2)

For secondary electron emission, the secondary electron yield (6) is defined as:

Secondary Electrons 6 =

Primary Electrons (2-3)

Figure 2.3 shows schematically the typical behaviour of 6 as a function of the

primary energy (EJ. The yield initially inmeases and reaches a maximum (8"") at a

given energy (EPm=0.2 to 2 keV) , and then decreases with increasing Ep This

behaviour is common to all types of materids, conducting or insulating. Table 2.1 lists

6- and E" for some elements[2]. Secondary electrons have energies much higher than

those within the substrate. The degree to which the crystal structure of the film

influences 6 is therefore of interest.

2. BACKGROUND

Figure 2.3: Typical graph for secondary electron yield (6) vs prirnary electron energy (Ep)

Table 2.1: 6- and EpW for various elements

2.2 Molybdenum Disuffide

Molybdenum disuifide (MoS3 in the form of powders of fine crystallites is used

9

2. BACKGROUND

as a solid lubricant having the cornmon name of Molyslip. The crystalline unit ce11 of

sputtered MoS, contains S-Mo-S layers (See Figure 2.4, [3]) with intra-plane chemical

bonds much stronger than the inter-plane van der Waals bonds. Due to the weak van der

Waals bonds, the planes are able to siide over each other easily giving MoS2 its

characteristic low coefficient of fiction. This layered structure would seem of

signincance for secondary emission control of MoS, films. Thin films of MoSz can be

produced through magnetron sputte~g [4]. The details of this technique will be

presented in Chapter 3.

viii & wulr

Figure 2.4: Planar hexagonal structure of magnetron sputtered MoS, crystallites. [3]

10

2. BACKGROUND

2.2.1 Structure of MoS,

The crystd structure of sputtered MoS2 is organized in platelets. Films are

usually a combination of the two following manners by which platelets organize

themselves on the substrate [4]: Type 1 and Type II.

Type 1 thin films have the platelets growing at an angle to the substrate (see

Figure 2.5) and have a black sooty appearance. Type II films grow the platelets in an

orientation parallel to the substrate (see Figure 2.5) and have a silvery matte appearance.

Obviously, Type II films are superior in the tribological sense and may be of importance

to secondary emission.

Types 1 and II MoS, can be distinguished by x-ray diffhction ('RD) patterns.

The XRD spectra of oriented MoSz films ciiffer considerably fiom those seen with the

bulk [SI. As observed by a number of authors [4][6][7l, and shown earlier at Queen's 181,

oriented films have relatively broad peaks located at 28cu=140 and î8cu=340. The fust

peak corresponds to platelets having an [O021 orientation (type II) and the second peak is

due to [100] orientation (Type I). Sample XRD spectra are show in Figure 2.6 for

sputtered films on stainless steel prepared using an RF magnetron sputtering unit.

2. BACKGROUND

'I)rpe 1: Edge Planes Type II: Basal Planes

Figure 2.5: Platelets of magnetron sputtered MoS, will have its platelets organized in planes in a combination o f 2 types: type 1 planes at an angle to the substrate and type II planes parallel to the substrate.

Figure 2.6: X-ray spectra of thin nIms of MoS, produced by magnetron sputtering. If the two peaks (type I at 14" and type II at 20) are of qua1 &ternity, thete will be approximately times more type 1 than type II. Therefore (a) represents a predominantly type 1 film and (b) a predorninantly type II

2. BACKGROUND

Information regarding the crystallographic nature of a coating on a Bat substrate is

assessed by several means. The absolute intensity of a given peak or set of peaks is a

measure of how well the coating is crysbllized. A fdly crystallized coating generally has

large intensity peaks, while an amorphous or non-crystalline film has one or two low

intensity, broad and ill-defmed shoulders extending over a 20 range of 20-30'. Compared

to a polycrystalline powder, the relative intensity of a given peak for a crystallized coating

may be larger or smaller than expected. This cm be evidence for texture or a preferential

crystdographic orientation within the coating. This may be induced by substrate factors

or by the incident direction of the incoming material. The broadening of a difEaction

line (full width at half maximum) cm be related to the grain sîze t within the coating.

The interpretation of experimental observations can be cornplex depending not

only on the type of X-ray measurement (0-20 or glancing angle) or if the substrate has a

surface roughness resulting fiom a grain size larger than that in the coating. As shown in

Section 4, this was the case in the present work.

2.3 Industrial Work

The sputtering of MoS, has been investigated by Teer Coating Ltd. [9]. A

specific tool of this group has been the use of a high curent, low voltage rnagnetron

source cdied an unbdmced rnagnetron. Teer have patented a commercial process using

this technique to create MoS2/titanium coatings ( M û S P ) which are said to have

2. BACKGROUND

supenor tribologicai properties. The initial proposai for this project was to construct an

unbalanced magnetron system for the fabrication of films. However, earlier work at

Queen's [8] indicated the possibility of the creation of Type 1 and Type II films using

conventional magnetron sputtering under controlled conditions.

Following a private communication fiom D.G. Teer [IO] that an unbalanced

magnetron process was not essentiai and the acquisition of an Materials Research

Corporation cryogenically pumped sputtering machine with 6" targets, it was decided to

investigate the secondary electron emission properties of layered MoS, thin films

prepared using conventional magnetron sputtering. The results have shown a

characteristic acicular type of deposit with primaïy texture in the Type 1 direction.

2. BACKGROUND

References

H. Seiler, Journal of Applied Physics, 54(l983), R1-18.

J.F.Shackelford, Introduction to Materials Science for Eneineers, Macmillan hiblishing Company, New York, 1985, p632-36

M.R. Loveîi, M.M. Khonsari and R.D. Marangoni, Transactions of the ASME, 118(1996), ~858-64

P.D. Fleishcauer, Thin Solid Film, l%(l %y), p309-22

Joint Cornmittee on Powder Diflkction Standards # 17-74!, #24-5 1 5, #3 7- l492

G. Weise, N. Mattem, H. Hermann, A. Teresiak, 1. Bacher, W. Bruckner, H.D Bauer, H. Vinzelberg, G. Reiss, U. Kreissig, M. Mader, P. Markschlager, Thin Solid Films, 298(1997) p98- 1 O6

J. Moser, H. Liao, and F. Lévy, Journal of Physics D: Applied Physics, 23(1990), p 624-6

W. Levy, Surnmer student report for Dr. M. Sayer, Physics Department, Queen's University, 1995

D.G. Teer, V. Bellido-Godes, and J. Hampshire International Confitence on Metallurgical Coatings and Thin Films (San Diego, 1997).

Pnvate communication fiom: Teer Coatings Ltd., Hartlebury Trading Estate, Hartlebury, Worcs DY 10 4JB, U.K.

3. EXPERIMENT

3.1 Overview of Sputtering

The process of sputtering uses a high energy plasma created in fiont of a target

material. This process is illustrated schematically in Figure 3.1. Ions fiom the plasma are

accelerated towards the target by a DC or RF potential and their impact displaces target

matenal into the vacuum for subsequent deposition on the substrate. Sputtering is

beneficial for the deposition of multi-component coatings because dl cornponents of the

MoS, target are displaced simultaneously. This is not absolute as the concentration of the

sulphur does Vary a small amount with some sputtering parameters [Il.

In simple DC sputtering, when the target and substrate conductivity are low, the

process of ion interaction will lead to a build up of charge on the surface of the target

which will eventually cause sputtering to cease. A solution is to use an altemating field

so that the target is altemately bombarded by positive ions and negative electrons and

therefore remains neutrai. A fiequency of 13.4 MHz in the radiofkequency range is

conventionally chosen as being available as a fiequency aliocated to commercial

radiofiequency generators.

Magnets 1 Electrons

\ S puttered Target atoms I

plasma ions Substrate

Figure 3.1: Representation of magnetron sputtering process whereby plasma ions scrike a target to fiee material for deposition on the substrate.

A plasma contains both electrons and ions. Sputtering is exclusively done by

ions. The light electrons are ineffective for sputtering and in fact tend to reduce

sputtering rates by reduction of effective ion currents. However, electrons are required to

keep the ion plasma neutrai and to sustain the plasma To achieve a wide range of

sputtering conditions the electton currents are confined using permanent magnets to a

ring path in which ionization is enhanced. The ions are then accelerated towards the

target without interference by the electrons. This is termed magnetron sputtering.

17

3. EXPERIMENT

The onginal sputtering of MoS, was made on a Vactec Sputtering unit with 4"

targets. The Materid Research Corporation Model 8667 Sputtering machine donated by

Alcan Research Laboratories is an industrial grade machine with 6" targets, cryogenically

pumped so that contamination is minimized. Results fiom this machine should be

directly transferrable to industrial applications. As part of this work, the machine was

made operational, targets were purchased and installed and extensive trouble shooting

was carried out. A temperature controlled substrate heater was installai. A schernatic

drawing of the MRC unit is s h o w is Figure 3.2. Three targets were installed to provide

the capability of depositing layered films.

66 1 saae l . - - I ïi *-Sputtcr Controis j o o g o c ,

Cham ber 1 1 3 t h ( , I c

Systern Contml

RF Powcr Supply

Main Unit

- - Chambec Top View Chamber. Side Vicw

Figure 3.2: MRC Model 8667 Sputtering System

3. EXPERIMENT

3.2 Sample Preparation

Most fiims were produced on alUrnuiun substrates similar to the matenai used in

the construction of satellites as received fiorn ComDev Ltd. This material is shown later

to have a grain structure on the order of I pm (see Figure 4.4). In accordance with a non-

anaiysis agreement no compositional anaiysis is available. Samples were also produced

on commercial grade aluminum, stainless steel and S i 4 substrates. Aluminum and

srainless steel substrates were sanded with 220 grit sandpaper, then 600 grit sandpaper

and fmalIy polished to a 6pm surface finish using a diamond suspension. Substrates were

cleaned with acetone and then placed in a beaker of ethyl alcohol in an ultrasonic bath

for 5 minutes pnor to deposition. Placing the polished substrates on a hotplate to remove

m e r organic contamination was unnecessary since the sample was heated in the

vacuum chamber.

3.3 Sputtering Procedure

Samples were cleaned and mounted with a temperature controlled sample holder

which could be heated electricaiiy to > 300°C. The temperature was monitored using a

K-type (chromel-alumel) therxnocouple and controlied using an Omega CN-2000

Controlier. The 6" MoS, targets were mounted horizontaily and the sample was placed

beneath the target. M e r evacuathg the chamber to 6 x 10 mTorr using a cryopump the

sample was raised to the desired temperature. High purity argon was introduced and the

3. EXPERIlMENT

plasma was ignited at a pressure of 50 mTorr. Sputtering took place at a variety of argon

pressures. On completion of sputtering, the chamber was allowed to cool before nitrogen

gas was introduced and the sample eventudly removed.

3.1.2 Sputtering Parameten

The sputtering parameters that were varied in the production of the MoS, films

were the pressure of the argon gas, the RF power and the temperature of the sample. The

pressure of the argon gas could be varied fiorn 1 mTorr to 100 mTorr. Initidly,

increasing the pressure of the argon gas increased the deposition rate. Eventuaiiy,

however, increasing the argon gas pressure lowers the deposition rate because it enhances

the scattering of sputtered material [l]. The point of dimlliishing retums is dependent on

the other sputtering conditions.

The RF power could be set anywhere between O and 500W. Increasing the power

m e r accelerates the ions thus increasing the deposition rate. The disadvantage in using

hi& power is that under high incident flux the crystalli7ation is poor because atoms have

insufficient time to reorder themselves after striking the substrate.

The substrate couid be heated up to 3 80°C. Higher temperature facilitates

crystaUization of the MoS, This is because a higher substrate temperature increases the

mobility of atoms dong the substrate and film d a c e .

3. EXPERIlMENT

3.13 Optimhation of Sputtering Parameters

A literature review was conducted to determine the 'optimum' starting parameters

for growth of well crystallized MoS, films (2 1 [3][4]. The recommended parameters Vary

considerably. This is most likely due to the differences in configuration between the

sputtering machines. A common recommendation for a beginning range for s p u t t e ~ g is

show in Table 3.1.

Table 3.1: Sputtering Parameters Suggested fiom the Literature

II RF Power II 100W-300W 1) 1

Substrate Temperature 11 150°C - 300°C 11

Sputtering Panmeter

(

The communication with G.H. Teer [SI suggested that low pressure and low

power provide the best Type II results. It was therefore decided to produce n I m s within

the range of sput te~g parameters mentioned above as weii as at lower power and

pressure.

1 Sugeested Range U

3.4 Characterization of Thin Films

Argon Pressure I I

The films were characterized using glancing angle x-ray dZûaction QCRD) to

determine the crystai structure, a Talystep profilorneter to determine the deposition rate

1 10 *Torr - 30 mTon II

and a scanning electron microscope (SEM) to evaluate the surface morphology.

3.4.1 X-Ray Diffraction Experiments

A 12kW Rigaku rotating copper anode x-ray genenitor was used to take XRD

spectra of the films. Even using glancing angle geometry, peaks fiom the substrate are

present in the spectra due to the small thickness of the film (0.5pm - 1.5p.n). Two

important difhction peaks (the Type II [O021 peak at 28,=14O and the Type I [100] peak

at 28,=34") are disthguished for MoS, films.

Estimates of the grain size within the films were made using the following

where B is the broadening of the difiction line (full width at half maximum) at the angle

8, Ac, is the x-ray wavelength (0.1 54 12nm in this case) and t is the grain size.

3.4.2 Talystep Profilorneter

Measurement of the thickness of the films was done using a Rank Precision

Industries Ltd Talystepm profilorneter. The Talystep is a stylus instrument which

measures thickness by travershg the stylus over an edge between the film and the

substrate. Vertical movement of the stylus is arnplified and recorded on paper.

When a sample was sputtered onto a substrate, a srnail piece of glass was also

3. EXPERIlMENT

placed on the heated sample holder in the chamber. Glass was used for measuring

thickness as it has a lower surface roughness than the metal substnites. A step was

produced by an aluminum shadow mask.

3.43 Scanning Electron Microscope (SEM)

The JEOL 840 Scanning Electron Microscope (SEM) uses an electron gun and

various focussing mechanisms to scan the surface of a sample with a beam of electrons.

The reflectedlbackscattered and secondary electrons are collected by backscattered and

secondary electron detecton to form an image of the sample through modulation of the

spot intensity on the video screen. The spot is scanned on the screen in step with the scan

of the sample.

Plan view micrographs were taken in this work. Attempts were made to create

suitable cross sections of the films by cutting the film and substrate and polishing the

edge. However, poiishing the edge removed the nIm fiom the substrate due to poor

adhesion and therefore no cross-section SEM rnicrographs were taken.

3.5 Secondary Electron Emîssion Measurements Using a SEM

3.5.1 Sample Pnparation

For measurements of the secondary electron yield (6) on aluminum or stainless

steel, the d a c e s were polished to a &or finish prior to coating. This minimized

3. EXPERIMENT

spurious effects due to edges and asperities on the surface. The sample, for which 6 was

to be measured, was attached to an SEM specimen mount wîth conducting tape.

3.5.2 Secondary Etectron EmUsion Measurement Overview

In order to measure the SEE of a thin film of MoS,, a vacuum environment with a

source of electrons and the ability to measure the nurnber of electrons absorbed or emitted

by the film is required. The secondary electron emission apparatus was constnicted by

rnodmg a JEOL-840 scanning electron microscope (SEM). This provided the electron

gun and the hi& vacuum environment.

It was necessary to interrupt the co~ect ion between the stage and the grouod with

the JEOL-840 in order to connect an ammeter to measure the curent Nnning fiom the

stage to ground. A Keithley 6 16 electrometer was used, confÏgured as a feedback-type

picoammeter in which the cumnt fiows through the feedback resistor of the voltage

amplifier. This instrument was capable of rneasuing currents as low as 1 ~ 1 0 * ' ~ amps. A

voltage source was also added to provide the option of a bias voltage to prevent

secondary electrons fiom escaping the sample. The apparatus was then ready to measure

secondary electron emission. Figure 3.3 shows a schematic of the apparatus.

3. EXPERIMENT

Vacuum cham ber4

Faraday Cage \

- Electron Gun

- Sample - stage

Figure 33: Secondary electron yield measurement using an SEM such that current absorbed by the stage may be measured as a function of the bias voltage on the stage.

In order to calculate 6, three currents are measured. First the total incoming

current Ip was recorded for various primary energies by directing the electron beam into a

Faraday cage. The Faraday cage was used to mesure the total incoming current by

preventing refiected or secondary electrons nom escaping the stage. The cage was

constnicted by attaching a 5mm aiuminum disc with a 2 0 ~ hole in the centre (schematic

diagram - Figure 3.4, SEM micrograph - Figure 3.5) to an aluminum sample holder.

Si& View

Figure 3.4: Scbematic d i a m representing the experirnental set-up used for measuring the total incoming current.

Figure 3.5: SEM micrograph of the hole in the aperture grid. The hole is black because no electrous escape.

Next, the electron bearn was directed at the sample. The curent measured (Io) is

the primary current minus the current due.to reflected electrons and secondary electrons.

Finaily a positive voltage of +5OV was applied behveen the stage and ground in order to

prevent the secondary electrons f?om escaping the material and the current measured (I,

was the incoming current minus the reflected electrons.

In order to determine how much voltage was necessary to prevent the secondary

electroas from escaping the material, a meamernent of the current absorbed as a hction

26

3. EXPE-NT

of bias voltage was taken. By calculating how much the current changes as the bias

voltage increases, a value for 1, was obtained. The merence between I,, and 4 gave the

current due to the secondary electrons 03. A complete data set (I,, 1, and 1, for a range

of E, values) showing the calculation of the secondary electron emission coefficient (6) is

given in Appendix A. The sample was a polished aluminum substrate.

The following equation gives the secondary electron yield fiom these two

measurements:

References

1. R Bichsel and F. Levy, Journal of Physics D: Applied Physics. 19 ( 1 986) p 1575-85

2. E.W. Roberts, B.J. Wiliiams and LA. Ogilivy, Journal of Physics D: Applied P hysics, 25 (1 W2), p A65-A70

3. RI. Christy, Thin Solid F h , C73, 1979, p 299-307

4. S. Pantz et al., Vacuum, 39 (7-8), 1989, p 735-8

5. Rivate communication fiom G.H. Teer of Teer Coatings Ltd., Hartlebury Trading Estate, Hartlebury, Worcs DY 10 4JB, U.K.

B.D. Cullity, Elements of X-ray Difiction, 1978,2" edition, p 284

4. FILM PRODUCTION

This chapter presents the effects of sputtering parameters on Type 1 and Type II

MoS, film production. The deposition rate of the MRC sputtering machine was

calculated fiom film thickness measurements made on the Talystep. This deposition rate

and the tirne of deposition was used to estimate the film thickness. In order to assess the

effects of different substrates the adhesion, crystallization and grain size of MoS,

deposited on silicon oxide, stainless steel, and duminurn substrates were analyzed using

x-ray difiaction. Findy, the surface morphology of the MoS, was examined with the

scanning electron microscope (SEM). The secondary electron emission properties are

discussed in Chapter 5.

4.1 Deposition rate

Films of MoS, were sputter deposited onto glass substrates which were partially

masked with duminum foil to provide a step profiie which could be measured using the

Talystep. The deposition rates were calculated and are given in Table 4.1, below.

4. FILM PRODUCTION

Table 4.1: Deposition Rates Measured as a Function of Argon Pressure and RF Power

Pressure

These results are graphed in Figure 4.1 and show that the deposition rate is

strongly dependent on the power and only weakly dependent on the argon pressure.

Estimates of film thickness were made for the MoS, Nms produced at higher argon

pressures by extrapolating from these graphs. The change in surface morphology when

the sputtering pressure rose above lOmT (discussed in Section 4.4.2) must have had an

effect on the thickness of the N m s . Therefore the film thicknesses stated are ody

estimates. Accurate thickness is not important in terms of secondary electron emission

since secondary electrons that will escape are created within the first 10 - 100nrn of the

d a c e . Any variation in film thickness above lOOnm will therefore not be important.

4. FILM PRODUCTION

1.6

1.4- (A) SmTon Argon Pressure (8) lOmTorr Argon Pmssun

3 1.0- - S g 0.8 - 8 0.6 - r 8 0.4 - CL

Power (W) Power 0

1.6 (C) SOW RF power (O) 1OOW RF power

G 1.4 -

Argon Ptessure (mTorr) Argon Pressure (mTorr)

Figure 4.1: Determinhg the effect of argon pressure and RF power on deposition rate.

4. FILM PRODUCTION

4.2 Silicon Oraide Substrate

The following table lists the range of sputtering conditions used in attempts to

produce thin films of MoS, on silicon oxide (SiO,) substrates.

Table 4.2: Summary of Results for Samples Prepared on Si02 Substrates.

1 Film Specifics 1 Results 1

Al1 attempts to produce films on SiO, or Si substrates were failures. Adhesion to

Si02 was not achieved with any sputtering conditions. A thin film of alurninum was

placed on the SiO, to hcrease adhesion. It was found that both the aluminum and the

MoS, layer separated nom the substrate after being heated. This was most ke1y due to

the residual stress between the film and substrate which results during the cool-down

phase.

Sample

L - Power (Watts)

50 - 500

Pressure (mTorr)

5 - 50

Time (min)

20-120

Temp (OC)

300

XRD Spectrum

NIA

Comment.

zero adhesion

4. FILM PRODUCTION

4.3 Stainless Steel Substrate

The following table lists the range of sputtering conditions used in atternpts to

produce thui films of MoS, on polished stainless steel substrates.

Table 4.3: Summary of Results for Samples Prepared on Stainless Steel Substrates.

Results

Comments

-

XRD Spectrum

N/A

Amorphous (Figure 4.1)

weak Type 1

- - - - - -- - - -

Film Specifics

Du11 Black, film flaked off imrnediately

I I I I I 1 (Figure43

Grey, film flaked off after 1 week

Sample

-

A

B Grey

Time (min)

30

60

75

Stainless steel provided better film adhesion than SiOdSi. However the adhesion

was still only acceptable when the power and pressure were Iowered to 100 Watts and 10

mTorr respectively. An XRD spectnun of sample A sputtered at these parameters, is

shown in Figure 42. The strong peak is fiom the stalliless steel substrate. The MoS,

contributes a broad, low intensity peak centered at 28,,=34*. This indicates a weakly

organized, nanocrystalline Type I fïim. A very small amount of crystalli7ntion was

achieved when the sputtering conditions were lowered to 50 W and 6 mTorr (Sample B,

Figure 4.3).

Power (Watts)

300

100

50

Temp ( O C )

300

300

300

Pressure (mTorr)

30

10

6

4. FILM PRODUCTION

5 10 15 20 25 30 35 40 45

28 for Cu radiation (degrees)

1000

Figure 4.2: Smnple A: XRD Spectra of MoS, nIm on stainless steel. The peak at 20,y14" is due to the substrate. The broad, low intensity peak centered at 20,=34" is indicative of a weakly organized type 1 MoS,.

i - IOOW, IOmTorr, 60 min, 300°C .

4. FILM PRODUCTION

10 15 20 25 30 35 40

20 for Cu radiation (degrees)

Figure 4.3: Sample B: XRD spectra of Type 1 MoS2 on stainiess steel. AU peaks are indicative of the substnite except for the broad peak at 2eCu=34O which is due to type 1 Mas,. Note that this is a stronger peak than that in Figure 4.2. AU other peaks present are due to the stainless steel substrate.

4. FILM PRODUCTION

4.4 Aluminum Substrate

The foilowing table lists the range of sputtering conditions used to produce thin

fh of MoS, on aluminum (Al) substrates.

Tabie 4.4: Summary of Resuits for Sarnples Prepared on Aiuminum Substrates.

Results 1 I Film Specifica - Temp (C) - 300

Comments

Weak Type 1 (Figure 4.5)

Type I (Figure 4.5

& 4.6)

Dark Grey

Dark Grey (ComDev substrate)

-

Type I (Figure 4.6

& 4.8)

Dark Grey (ComDev substrate)

- - -

Dark Grey (ComDev substrate)

Shiny Black (ComDev substrate)

Shiny Bfack (ComDev substrate)

Type I (Figure 4.6)

Type I (Figure 4.10)

TY pe (Figure 4.8 & 4.10 )

Type I (Figure 4.12)

Shiny Black (Polished Al substrate)

TypeDT'II (Figure 4.12)

J was burnished (turned grey) (Polished Al subsfrate) 1

4. FILM PRODUCTION

Alumiraum substrates yielded the best MoSz films. In order to understand the

results, SEM micrographs of the two types of aluminum substrates were taken for

cornparison. Micrographs at 10,000 times magnification for the ComDev and polished

aluminum substrates are shown in Figure 4.4.

The fht films, produced in the range of 300 - 500 watts and 10 - 40 mTorr,

produced linle or poor adhesion. It is important to note that the higher RF power resdts

in a higher incident flux of sputtered material. This Iimits the time available for atom

rearrangernent on the substrate surface. In addition, films produced with a power less

than 100 W had low deposition rates, requiring several hours to produce fiims, and were

therefore not practical fkom an indushial perspective.

Films with acceptable adhesion were produce in the range of 100 to 300 Watts

and lOmTorr to 30mTorr. These films will be discussed in the following sections.

4. FILM PRODUCTION

ma';puncation: (a) ~ o ~ e ~ a l u m i n u m substrat= and (b) polished aluminum substrate

4. FILM PRODUCTION

4.4.1 Determining the Effect of Substrate Temperature

In order to determine the effect of temperature, two samples were produced with

identical argon pressure and RF power (1 OmTorr and 100W) but at different substrate

temperatures: Sample C at 150°C and Sample D at 300°C. Their XRD spectra are shown

in Figure 4.5. The peak produced at 28,=34" is due to type 1 MoS,. The absence of any

type II peaks (20,=14") indicates that the film is entirely type 1. The intensity of the

peak increases as the temperature increases, signifving an increase in crystalluation.

While the overail intensity of the peak increased with increasing temperature, the width

of the peak also increased. At low temperattues, the atoms are more stationary and the

resulting film has a large percentage of amorphous material. At higher temperatmes, the

atoms can rearmnge themselves on the surface and the resulting nIm has an increase in

the percentage of crystalline material.

Calculation of the grain size of the film using equation 3.1 and the peak width

gives girain sizes of 3.0IOSnm and 2.2k0.5nm for Sarnples C and D. The two grain sizes

are w i t . error of each other and therefore do no indicate a trend in grain size as the

temperature changes. The rough surface of the ComDev aluminum may play a significant

role in the peak broadening.

4. FILM PRODUCTION

10 15 20 25 30 35 40

20 for Cu radiation (degrees)

Figure 4.5: XRD spectni of samples C and D. MoS, on aiuminum whose sputtering conditions are identicai (100W, 1 OmTorr and 0.5hr) aside fiom substrate temperature. Note that the intensity of the Type 1 MoS, peak at 20,=34O increases as the temperature inmases. The MZD spectra for the 300°C sample has been offset by 1 SCPS for clarity.

4. FILM PRODUCTION

4.4.2 Determinhg the Effect of Argon Pressure

in order to determine the effect of argon pressure on the crystallization and

d a c e morphology, the next set of MoS, films were sputtered at different argon

pressures: sample D at 1 OmTorr, sample E at 2OmTorr and sample F at 30mTorr ont0

ComDev aluminum. The XRD spectra are shown in Figure 4.6. Once again, the ody

peak produced by the MoS, is the Type 1 peak at 20,,,=34O. The peak intensity, and

therefore the crystallization, decreases with decreasing argon pressure. The peak

broadening, according to equation 3.1, indicates that the grain size increases with

increasing argon pressure. The grain sizes for sarnples D, E and F are 2.20.5nm,

3.8kO.Snm and 5.7+0.5nm respectively.

SEM micrographs taken of these samples at ~10,000 magnification are shown in

Figure 4.7. Sample D shows a dense film with grain size too small to discem The film

mimics the surface roughness of the underlyhg substrate.

There is a signincant change in the surface morphology in the samples E and F.

The film has taken a loosely packed, needle-like or acicular fom. In the case of sample

E, there is very little evidence of the substrate roughness in the film. The crystals visible

on sample F are more tightly packed, redting in the substrate roughness being evident

The orientation of the needles with respect to the substrate clearly varies signüicantly.

This will influence the observed broadening.

4. FILM PRODUCTION

Sample F: 30mTorr

Sample E: 20mTorr

20 for Cu radiation(degrees)

Figure 4.6: XRD spectra of samples D, E and F. MoS, on aliuninum whose sputtering conditions are identicai (100W, 0.5hr and 300°C) aside fkom argon pressure. Note that the intensity of the Type 1 MoS, peak at 20,,=34" decreases as the argon pressure decreases. For cl*, the XRD spectra for the 2OmTorr and 3OmTorr samples are offset by ZOCPS and 40CPS, respectively.

4. FILM PRODUCTION

p&~sure on the d a c e morphology: (a) Sample D: 1 OmTon, (b) Sample E: 2OmTorr and (c) Sample F: 30mTorr. AU of the samples were sputtered for 0.5 hours at IOOW and 300°C.

4. KILM PRODUCTION

4.43 Determining the Effect of RF' Power

In order to examine the effect of the RF power on the MoS, film, samples E and

H were sputtered with different powers (100 and 200 Watts respectively). Since

increasing the RF power increases the deposition rate, their individuai sputtering times

were adjusted to maintain a consistent thickness (approximately 0.8pm). Their XRD

spectra are plotted in Figure 4.8. The sample produced at lOOW yields less crystallization

than the sample produced at 200W. Calcdations made using equation 3.1 show that

grain size changes very little. The grain size is 3.8*0.5nm for sample E and 4.510.5nm

for sample H.

Figure 4.9 shows SEM micrographs taken of both samples at ~10,000

magnification. The acicular structures are of similar size suggesting that the grain sizes

are comparable. The main ciifferences between the two samples is the degree of

crystaliization and the sputtering rate.

4. FILM PRODUCTION

Sample H: 200W, 1t4hr 1 \ Sampte E: IOOW, 1/2hr

28 for Cu radiation (degrees)

Figure 4.8: Cornparison of XRD spectra for samples E and K. MoSz films sputtered at different RF power values. The sample sputtered at lOOW has a larger amorphous background, with evidence of crystallization. The sample sputtered at 200W shows increased crystallization. For clarity the spectra for the sample sputtered at 200W has been offset by SOCPS.

4. FILM PRODUCTION

M& films sputkred at an argon pressure of 2OmTorr and substrate temperature of 300°C for (a) Sample E: 0.5 hous at lOOW and (b) Sample H: 0.25 hours at 200W.

4. FILM PRODUCTION

4.4.4 Determinhg the Effect of Sputtering Time

Samples G and H are compared to observe changes in crystallization and grain

size as the sputtering tirne was increased from 15 minutes to 30minutes. By increasing

the sputtering t h e , film thickness increased fiom approximately 0 . 7 0 ~ to 1 . 4 ~ . XRD

spectra for sarnpies G and H are plotted together in Figure 4.10. Again using equation

3.1 the grain sizes are shown to be close: 5.7k0.5nm for sample G and 4.5*0.5nm for

sample H.

SEM micrographs of the two samples are shown in Figure 4.1 1. Once again, the

acicular structures are similar, as suggested by the grain sizes. The only difference in the

surface morphology is that sample H reveais the surface rnorphology of the substrate

more than sample G. This is due to the increase in film thickness. As the film begins to

deposit, the £ilm conforms to the surface roughness of the mbsûate. As the film

continues to grow, the valleys are eventually filled in such that the surface roughness

decreases.

4. FILM PRODUCTION

I O 15 20 25 30 35 40

20 for Cu radiation (degrees)

Figure 4.10: Cornparison of XRD spectra for samples G and H. MoS2 nIms produced with same sputtering conditions (200W, 2OmTorr and 3OO0C) except for Merent lengths of time (0-25 hours and 0.5 hours). The decrease in the Type 1 peak at 20,,=34O with sputtering tirne conesponds to a decrease in crystallization. The narrowing of the peak indicates an increasing in grain size with increasing sputtering tirne. The sample sputtered for 0.5 hours has been offset by 6OCPS for clarity.

4. FILM PRODUCTION

- - - MoS, sputtered at an RF power of200W, an argon pressure of 20mTorr, a substrate temperature of ?OO°C and a sputte~g t h e of (a) Sample H: 0.5 hours and (b) Sample G: 0.25 hours.

4. FILM PRODUCTION

4.5 Mechanical Alignment of Coating Structure

Since Type iI films could not be created by varying the s p u t t e ~ g conditions,

Type 1 films were produced and the film bumished with a soft cloth to flatten out the

piatelets. This has been previously demonstrated in the literature when creating films for

tribological testing [l]. Films were sputtered for 2 hours with an RF power of 150W7 an

argon pressure of 25 mTorr and a substrate temperature of 300°C. The XRD spectrum of

such films (prior to bumishing) is illustrateci in Figure 4.12. An XRD spectrum of the

burnished surface (Figure 4.12) shows that the bumishùig created a small peak at

approxirnately 20c,=140. This peak, dong with the now diminished peak at 20,u=3407

indicates that the resultant films is a combination of Type 1 and Type II MoS,.

As discussed in Figure 2.6, it has been shown that equal amounts of Type 1 and

Type II MoS, will result in a Type 1 peak (2ec,=14O) that is 100 times stronger than the

Type II peak (28,=34O)[2]. As can be seen by the low intensity of the 14O peak in Figure

4.12, the amount of Type II nIm produced by bumishing the sample is very smd.

4. FILM PRODUCTION

Type 11 [O021 peak

-

20 for Cu radiation (degrees)

Figure 4.12: Cornparison of XRD spectra for the same sample pnor to and after burnishing with a cloth (Kimwipe-L). There is a very small decrease in the type I peak at 2ecU=34O. The burnishing produces a peak at 28c,=140 which shows a small amount of Type II MoS, has been produced. For clarity the bumished spectnim has been offset by 6OCPS.

4. FILM PRODUCTION

SEM rnicrographs of the sarnple before and after bumishing are shown in Figure

4.13. As expected, bumishing the sample reduces the overall surface roughness. This

becomes important when determining secondary electron emission properties. This is

f.urthtr explained in Chapter 5.

morphology of a sample sputtered for one hour at an RF power of 150W, an argon pressure of 20mTorr and a mbstrate temperature of 300°C (a) pnor to and (ô) after bumishing.

4. FlLM PRODUCTION

4.6 Summary of Film Production

The major conclusion fiom the XRD and SEM studies is that the particle

orientation is primarily Type 1 and the particle size is very srnail. By cornparison with the

results of Levy [3] using the 4" target VacTec machine, the 6" target MRC machine was

less effective in creating the two types of crystal structure. A possible explanation for

this is that the greater uniformity of the field in the 6" diameter system does not

encourage crystallization of the parallel texture.

References

1. P. Fleishcauer and R Bauer, Tribohgy Tranactions, V3 1,2, p 239-250

2. J. Moser, H Liao and F. Levy, Journal of Physics D: Applied Physics, 23(l WO), p 624-626

3. W. Levy, Summer student report for Dr. M. Sayer, Physics Department, Queen's University, 1995

5. SECONDARY ELECTRON YIELD

The secondary electron emission apparatus was first calibrated by mesuring

secondary electron yield (6) as a f'unction of primary energy (EJ for staidess steel and

aluminum. These values were then compared to the published values of the maximum

yield (6y and the correspondhg primary energy @,O"). The resdts of rneamrernents of

the secondary electron yield (6) as a hct ion ofprimary energy (Ed are presented

graphically .

5.1 Determination of V,

Measurernents were taken of the current absorbed by a polished aluminum

sampIe as a function of the applied stage bias voltage (see Figure 5.1). By calculahg

how much 'extra' current was absorbed as the voltage was increased in steps of 6V it

was found that no significant change in the cunent occurred beyond a bias voltage of

42V.

5. SECONDARY ELECTRON YELD

Stage Bias Voltage (V)

Figure 5.1: Determuiing the necessary bias voltage to be appiied to the stage to ensure the capture of a i l secondary electrons.

5. SECONDARY ELECTRON YIELD

5.2 Calibration measurements and Stainless Steel

The secondas, electron yield (6) was measured for stainless steel (see Figure 5.2).

As shown in the table 5.1, the value of 6- fdls within experimental error of the

published value [l 1.

Table 5.1: Cornparison of Experimental and Published values of PU and E" for Stainless Steel.

- - - -

Substrate 6max EPmx(eV)

The values of 6- and E" are within experimentai error of the published values.

The large uncertainty value for the EPm is due to energy step size limitations on the

SEM.

5. SECONDARY ELECTRON YIELD

Figure 5.2: The secondary electron yield (6) as a bction of primary electron energy (Ep) for stainless steel.

5. SECONDARY ELECTRON YIELD

5.3 Calibration measurements on Aluminum

The secondary electron yield (6) measurements of polished and ComDev

aluminum are shown in Figure 5.3. A cornparison of the value of 6- and E F measured

by experiment and compared with the values fkom the literature [2] is s h o w in Table 5.2

below:

Table 5.2: Cornparison of experimental and published values of 6- and E" for alUrninum.

Substrate 1) 11 Enmax (eV)

II Experiment

The results for the alurninum are equal to the established values within stated

uncertainties. The difference between the two samples can be accounted for by the

clifference in surface roughness which has a signincant effect on secondary electron yield

[3]. Figure 5.4 schematically depicts the secondary electron emission behaviour of rough

and smooth surfaces. Secondary electrons produced near edges, see Figure 5.4@), will

have extra s d a c e area fiom which to escape. A rough surface has more surface area.

This enables secondary electrons which would usually be reabsorbed by the material to

escape*

5. SECONDARY ELECTRON YIELD

+ Polished Aluminum -û- ComDev Aluminum

Figure 5.3: The secondary electron yield (6) as a hction of primary electron energy (EJ for polished and ComDev alumînum

5. SECONDARY ELECTRON YlELD

Figure 5.4: Schematic diagram illustrating the path of secondary electrons in materials with either a smooth or rough surface. Secondary electrons created in materiais with a rough surface will have more surface area through which to escape than a smooth d a c e .

Since the 6- and EPm for both aluminum and stainless steel substrates were

equal within experimental uncertainty, the calibration procedure was deemed complete.

The secondary electron yield (6) of the sputtered MoS2 fihm could now be meamred.

5. SECONDARY ELECTRON YIELD

5.4 Secondary Electron YieId Measurements of MoS, films

Measurements of the secondary electron emission yield (6) were made on three

films produced on ComDev aluminum substrates representing samples of different

surface morphology and crystallization (see Figure 5.5). AU of the samples were

sputtered for % hr at an RF power of 1 O0 W and a substnite temperature of 300 OC. The

argon pressure was varied to alter the surface morphology and crystallization. Film

thicknesses were equal within experimental uncertainties. SEM micrographs of the

substrate and the three samples are shown in Figure 5.6 for reference.

The first measurement was made on sample C sputtered at 1 OOW, 1 OmT, and

300 OC for % hr (SEM micrograph in Figure 5.6(a)). The sample morphology was small

grained (2.1k0.5nm) and had low crystallization. Its secondary electron yield is very

similar to that of the mbstnite at energies above 1 keV. At lower energies there is a

decrease in the secondary electron yield with respect to the substrate becaue the primary

eiectrons spend more t h e in the film and therefore the effects of the nIm wiil be more

pronounced.

The expianation for the reduction in 6 in the acicular films is analogous to that

given in section 5.3 which explains the fact that the polished aiuminum possess a lower

secondary electron yield than the ComDev aluminum. The MoS2 film decreased the

overail d a c e roughness thereby lowering secondary electron emission [3].

5. SECONDARY ELECTRON YIELD

I I I I I I I I I I

(a) - - ComDsv Alumlnum -- &O - ComOw Aluminum

- ComOev Alumlnurn - ComDev Aluminum + i W 30mT. 1Rhr. W C -0- 100W,lOmT.lrZhr.30aC

-& lOOW,20mT. t Rhr,3OOC -& 100W,30mT,1Rht,300C

Figure 5.5: Graphs of secondary electron yield as a function of primary electron energy for MoS2 films sputtered on ComDev aluminum substrates at (a) Sample C: 1 OOW, IOmT, 1/2hr, 300°C, (b) Sample E: 100W, 20mT, 1/2hr, 300°C, (c) Sample F: 100 W,30mT, 1/2hr,300°C, and (d) a cornparison of aU three results.

5. SECONDARY ELECTRON YIELD

Figure 5.6: SEM micrographs iUustrating the substrate and different MoS2 nIm d a c e morphologies created using the MRC sputtering system at an RF power of 100W, a substrate temperature of 300°C and a sputtering time of '/z hr: (a) ComDev alwninum substrate, @) 1 OmTon argon pressure, (c) 20mTorr argon pressure and (d) 30mTorr argon pressure.

5. SECONDARY ELECTRON YIELD

Samples E and F, sputtered at 20mT and 30mT respectively, have the same

acicular structure but different densities(Figure 5 4 c ) and 5.4(d)) . The difference in 6

between sample E and F can be accounted for by the roughness of the surface. Sample F

has greater overall surface roughness because it conforms more to the substrate suface.

The secondary electron yield of sarnples E and F differ significantly fiom the

yield for the substrate and sample C. The yield has been reduced by approximately 40%

at low primary energy and 10% at higher energy. This arnount of reduction of 6 is unique

to the acicular MoS, films.

The unique acicular form of the film accounts for this drop in the secondary

electron yield. Since the escape depth of secondary electrons is less than 100~1 , al1 of

the secondary electrons will be produced in the MoS, fih. Due to the porosity of the

film, secondary electrons will have to pass through several air-MoS2 interfaces to make it

to the surface of the material. This results in an increase in the effective work fùnction

that the secondary electrons must overcome to escape the surface.

5. SECONDARY ELECTRON YIELD

5.5 Secondary Electroa Emission Yield Measurements of Burnished

M0Sz films

Measurements of 5 were also made for the unbumished (Type I) and burnished

(type 1 / type II) MoS, (see Figure 5.7). The results show that bumishg the film reduced

the ef5ect that the film had on the secondary electron yield of the duminum. The arnount

of type II MoS, produced is negligible and therefore could not result in such a dramatic

change.

Burnishing the sample decreased overail surface roughness (see section 4.5). By

removing the acicular surface morphology, the secondary electron yield increased. This

is consistent with the conclusion of section 5.4 that the surface morphology is the primary

factor responsible for the change in secondary electron yield.

5. SECONDARY ELECTRON YIELD

-o- Burnished Type I

Figure 5.7: Secondary electmn yield(6) vs primary energy(Ep)for MoS, on Al for type I and bumished type 1 (The thick c w e represents d for the aluminum mbstrate)

References

D. Ruch, R. Moore, D. Manos, and S. Cohen, Journal of V a m m Science and Technology, 20(4), April 1982, pl3 13-13 16.

Handbook of Chernistw and Phvsics, (CRC, Boca Raton, 1993) 74& edition, Sect. 12, p107.

H. Seiler, 1983, Jounial ofApplied Physics, 54, p 1-18.

6. CONCLUSIONS

The goals of this thesis were to: (1) create type 1 and type II films of MoS, films

using the Material Research Corporation (MRC) sputtering machine and (2) to examine

the correlation between film and secondary electron yield (6) properties.

Parameter optimization on the MRC sputtering machine permitted the production

of crystallized type 1 films on aiuminum substrates. The crystallized type I MoS, films

showed two distinct surface morphologies that were dependent upon sputtering

conditions. The first type describes films that were small grained and had surface

morphologies which coafomed to the substrate roughness (Figure 5.6@)). The second

type describes films that were larger grained, with stronger crystallization, and possessed

a surface morphology formed of loosely packed needle-like 'acicular' structures varying

in length between 0 . 5 ~ and I . O p (Figure 5.6(c,d)). This second type of film

morphology was observed over a narrow range of sputte~g conditions (20 - 30mTorr

and 100 - 200 W).

The MRC machine proved ineffectuai at producing type II fihs. Attempts to

mechanidy align the type 1 platelets paralle1 to the subtrate via burnishing produced a

film containhg a negfigible amount of type II MoS,. This is probably due to the fact that

burnishing only &ers the uppermost platelets present at the surface leaving the

underlying platelets untouched.

6, CONCLUSIONS

The modified SEM was calibrated by accmtely measuring the secondary electron

emission yield (6) as a fuaction of primary electron energy (EJ of the two substrate

materials: stainless steel and aluminum. The calculated values for 6- and Epw for both

dumllium and stainless steel were equal to the values found in the literahue witbin stated

uncertainties. This permitted the accurate detemination of 6 as a function to E, for the

various sputtered thui films of MoS, produced within this work (see Table 6.1 for 6-

and E,mU). The behaviour of 6 for MoS, was best characterized as a fhction of the

surface morphology of the film.

The small grained non-acicular films had secondary electron yields that were

comparable to those of the substrate whereas the acicular films produced a large drop in

6. Increasing the density of the acicular structures resulted in decreasing secondary

electron emission. Since crystaiiized type 1 films produced a significant change in only

the acicular fonn it can be concluded that the surface morphology and not the crystal

texture is responsible for the changes in secondary electron emission properties of MoS,

thin films.

The acicular structure reduces the secondary electron yield of the aluminum

substrate by approxhately 40%. The effect on 6 c m be explained in terms of

interactions between the secondary electrons and the film structure. Due to the porosity

of the structure, many air-MoS, interfaces will be present and are encountered by the

secondary electrons while travening the film to escape through the surface. This results

6. CONCLUSIONS

in an increase in the effective work function that the secondary electrons m u t overcome

to escape the surface.

While burnishing the sarnple did produce a s m d amount of Type II MoS,, the

change in 6 for films before and after being burnished is most accurately explained in

terms of the change in the surface morphology. Buniishing the type 1 film had the effect

of Ulcreasing 6. Even though the bumishing did create a srnall amount of type II MoS,, it

was too little to account for the large change in 6. The burnishing removed the acicular

structure fiom the sdace thereby removing the air-MoS, interfaces near the surface.

This made it easier for secondary electrons to escape.

Thin films of MoS, sputtered using the MRC sputtering unit c m be engineered to

grow with a characteristic acicular surface morphology. Production of type 1 films

without the acicular morphology resulted in negligible changes in the secondary electron

emission properties of the aluminum substrates. it can therefore be concluded that these

fine grained needle-üke structures determine the secondary electron yield, not the aligned

texture of type 1 films.

i i v i n u r L V ~ L U W I I V I Y

TEST TARGET (QA-3)

APPLIED IMAGE , lnc fi 1 653 East Main Street - -. , Rochester. NY 14609 USA -- --= Phone: 716î482-0300 -- -- - - F a 716QW-5989