UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Stoichiometry control mechanisms of bias sputtered zinc oxide films Brett, Michael Julian 1985

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1985_A1 B73.pdf [ 5.06MB ]
Metadata
JSON: 831-1.0085519.json
JSON-LD: 831-1.0085519-ld.json
RDF/XML (Pretty): 831-1.0085519-rdf.xml
RDF/JSON: 831-1.0085519-rdf.json
Turtle: 831-1.0085519-turtle.txt
N-Triples: 831-1.0085519-rdf-ntriples.txt
Original Record: 831-1.0085519-source.json
Full Text
831-1.0085519-fulltext.txt
Citation
831-1.0085519.ris

Full Text

STOICHIOMETRY CONTROL MECHANISMS OF BIAS SPUTTERED ZINC OXIDE FILMS by MICHAEL JULIAN BRETT B . S c , Queen's U n i v e r s i t y , 1979 M . S c , U n i v e r s i t y of B r i t i s h Columbia, 1981 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Ph y s i c s Department We accept t h i s t h e s i s as conforming to the r e q u i r e d standard THE UNIVERSITY OF BRITISH COLUMBIA June 1985 © Michael J u l i a n B r e t t , 1985 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Physics  The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date June 20, 1985 i i ABSTRACT This thesis reports the f i r s t detailed study of the stoichiometry control mechanisms and physical properties of ZnO films deposited by dc planar magnetron sputtering of a Zn target in a reactive Ar/02 atmosphere. Control of f i l m stoichiometry was achieved using a subsidiary rf discharge at the substrate and a reactive gas baffle surrounding the target. The reactive gas b a f f l e was shown to enhance f i l m oxidation by decreasing the metal flux to the substrate and increasing the oxygen p a r t i a l pressure near the substrate. Rutherford backscattering analysis of f i l m stoichiometry demonstrated that the effect of the rf discharge was to increase the O/Zn composition r a t i o . This oxidation was shown to occur through p r e f e r e n t i a l resputtering and p r e f e r e n t i a l evaporation of excess Zn and by a c t i v a t i o n and ion plating of oxygen species. Resputtering and evaporation rates were found to be enhanced above that expected for bulk Zn, due to the weak bonding of surface adatoms during fil m growth. Conducting ZnO films produced at various values of the rf-induced substrate bias voltage were characterized for e l e c t r i c a l , o p t i c a l and s t r u c t u r a l properties using H a l l probe,. X-ray d i f f r a c t i o n , electron microscope, and v i s i b l e and infrared spectroscopy techniques. Films deposited at low substrate bias (0 to -50V) were found to have a large Zn excess (15%) res u l t i n g in low electron m o b i l i t i e s (1 cm2/Vs), high r e s i s t i v i t i e s (10"2 flcm) and were strongly absorbing in the v i s i b l e . Films deposited at high substrate bias were nearly stoichiometric, o p t i c a l l y transparent and had high electron m o b i l i t i e s (15 cm2/Vs) resulting in low r e s i s t i v i t y (10"3 Qcm). The o p t i c a l properties of transparent conducting films for wavelengths 0.4 to 20 /im were modelled by the Drude theory of free electrons using measured e l e c t r i c a l transport properties. The o r i g i n a l goal of t h i s work, to develop a heat mirror coating suitable for manufacture, was achieved by bias sputter deposition of ZnO onto uncooled polyester sheet at deposition rates approaching 75 nm/min. The best heat mirror films had a transmission to solar energy of 75% and an 85% r e f l e c t i o n of 300 K blackbody r a d i a t i o n . i v TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS iv LIST OF TABLES v i LIST OF FIGURES v i i ACKNOWLEDGEMENTS ix CHAPTER 1 - INTRODUCTION 1.1 M o t i v a t i o n f o r T h i s Work 1 1 .2 Heat M i r r o r s .. 2 1.3 Drude Theory of Free E l e c t r o n s 5 1 .4 Choice of ZnO 8 1.5 P r o p e r t i e s of ZnO 11 1.6 D e p o s i t i o n Processes 13 CHAPTER 2 - HEAT MIRROR DEPOSITION 2.1 Apparatus and Experimental Method 18 2.2 Operating C h a r a c t e r i s t i c s 24 2.3 ZnO Heat M i r r o r s 30 2.4 ITO Heat M i r r o r s 34 2.5 Summary 36 CHAPTER 3 - THE REACTIVE GAS BAFFLE 37 CHAPTER 4 - THE SUBSTRATE RF DISCHARGE 4.1 P r o p e r t i e s of an r f Glow Discharge 42 4.2 E f f e c t s of an r f Discharge 47 4.3 F i l m S t o i c h i o m e t r y 50 4.4 M a t e r i a l Re-emission 54 4.5 Zinc Re-emission by Evaporation 57 4.6 Zinc Re-emission by S p u t t e r i n g 64 4.7 Oxygen A c t i v a t i o n and P l a t i n g 68 4.8 Summary and D i s c u s s i o n 72 CHAPTER 5 - ZnO FILM CHARACTERIZATION 5.1 F i l m Degradation 77 5.2 F i l m S t r u c t u r e 80 5.3 E l e c t r i c a l P r o p e r t i e s 91 5.4 O p t i c a l P r o p e r t i e s 95 CHAPTER 6 - SUMMARY " 112 CHAPTER 7 - APPENDIX 7.1 O p t i c a l Constants of Transparent F i l m s 115 7.2 O p t i c a l Constants of Absorbing F i l m s ..117 REFERENCES 120 vi LIST OF TABLES Table Page 1.1 Typical properties of oxide heat mirrors 9 4.1 Mass measurement of bias sputtered Zn films ... 57 4.2 Comparison of re-emission rate from ZnO with bulk Zn evaporation rate 63 4.3 Resputter of ZnO films after deposition 67 4.4 • Resputter yi e l d s of Zn and ZnO films 68 4.5 Thicknesses of bias sputtered T i 02 films 75 5.1 Grain size of bias sputtered ZnO films 86 5.2 E l e c t r i c a l properties of bias sputtered ZnO films 92 5.3 E l e c t r i c a l properties of ZnO films produced by other workers 92 v i i LIST OF FIGURES Figure Page 1.1 R e f l e c t i v i t y of a free electron plasma 7 1.2 ZnO energy le v e l diagram 12 2.1 Schematic of sputtering chamber 19 2.2 The reactive gas b a f f l e 20 2.3 Substrate holder for p l a s t i c sheet 22 2.4 Variations of cathode voltage and pp(02) with f( 02) 26 2.5 Cathode current/voltage c h a r a c t e r i s t i c 28 2.6 R e s i s t i v i t y and i r r e f l e c t i v i t y for bias sputtered ZnO ". 31 2.7 Heat mirror c h a r a c t e r i s t i c s of ZnO films 32 2.8 Heat mirror, c h a r a c t e r i s t i c s of ITO films 35 3.1 pp(02) enhancement by the reactive gas b a f f l e . 38 3.2 Substrate and target discharge interaction .... 40 4.1 Energy d i s t r i b u t i o n of ions at the substrate .. 46 4.2 Rutherford Backscatter spectrum for ZnO 51 4.3 Stoichiometry of bias sputtered ZnO 53 4.4 Variation of material re-emission with substrate bias and f ( 02) 55 4.5 Substrate temperature and rf power measurement . 59 4.6 Material re-emission from ZnO by evaporation .. 62 4.7 Material re-emission from Zn by sputtering .... 65 4.8 Material re-emission from ZnO by sputtering ... 66 v i i i 4.9 Ef f e c t s of oxygen activation 70 5.1 Degradation of r e s i s t i v i t y for heated ZnO 78 5.2 Degradation of i r r e f l e c t i v i t y for heated ZnO . 79 5.3 X-ray d i f f r a c t i o n spectra for films deposited at various f ( 02) 82 5.4 . X-ray d i f f r a c t i o n spectra for bias sputtered Zn. 84 5.5 X-ray d i f f r a c t i o n spectra for bias sputtered ZnO 85 5.6 E f f e c t s of annealing of ZnO 88 5.7 Scanning electron photomicrograph of columnar structure in ZnO 89 5.8 Scanning electron photomicrograph showing the effect of bias on surface roughness 90 5.9 Geometry of films for Hall probe measurements . 92 5.10 Depiction of Manifacier's technique for deducing o p t i c a l constants 98 5.11 Optical constants of bias sputtered ZnO films . 99 5.12 Ir r e f l e c t i v i t y of bias sputtered ZnO films ...101 5.13 Drude prediction of i r r e f l e c t i v i t y for ZnO-5 .103 5.14 Drude prediction of o p t i c a l constants for ZnO-5.104 5.15 Drude prediction of o p t i c a l constants and i r r e f l e c t i v i t y for ZnO-1 106 5.16 Transmission electron photomicrographs of bias sputtered ZnO 108 5.17 Maxwell-Garnett prediction of i r r e f l e c t v i t y for ZnO-1 111 7.1 Cross section of ZnO on quartz structure 116 ix ACKNOWLEDGEMENTS I am especially grateful to my supervisor, Dr. R. Parsons, for the support, advice and assistance that he has provided throughout my years of study as a graduate student. I feel fortunate to have had a r i c h and rewarding association with many other people in the semiconductor lab, and would l i k e to extend thanks to John A f f i n i t o , Normand F o r t i e r , Richard McMahon, Brian S u l l i v a n , Juan Rostworowski, and Kevin Betts for their friendship and assistance. I greatly appreciate the f i n a n c i a l assistance that was provided by the Natural Sciences and Engineering Research Council, and by the University of B r i t i s h Columbia. 1 CHAPTER ONE: INTRODUCTION 1.1 MOTIVATION FOR THIS WORK R a d i a t i v e heat l o s t through s i n g l e or double glazed windows in Canadian homes during the winter season may c o n s t i t u t e up to 50% of the t o t a l heat l o s s in the b u i l d i n g [ 1 ] . However, the thermal r e s i s t a n c e or 'R-value' of a window may be d r a m a t i c a l l y improved by the a p p l i c a t i o n of a heat mirror c o a t i n g to the inner g l a z i n g s u r f a c e s . Such a heat m i r r o r , i d e a l f o r c o l d c l i m a t e s , i s a t h i n f i l m m a t e r i a l that i s t r a n s p a r e n t to s o l a r energy (0.30 to 2.0 urn wavelength) but r e f l e c t s room temperature blackbody r a d i a t i o n (5 to 40 um wavelength). My t h e s i s work was i n i t i a l l y motivated by a d e s i r e to develop a heat m i r r o r m a t e r i a l and a corresponding high r a t e t h i n f i l m d e p o s i t i o n technique s u i t a b l e f o r l a r g e s c a l e manufacturing. These goals were f u l f i l l e d by the s u c c e s s f u l production of a new heat m i r r o r c o a t i n g , ZnO, by b i a s s p u t t e r i n g onto p l a s t i c sheet at a d e p o s i t i o n r a t e of 75 nm/min. In order to obta i n a deeper understanding of the d e p o s i t i o n p r o c e s s , r e s e a r c h was then expanded to i d e n t i f y and analyze the mechanisms through which the b i a s s p u t t e r technique enabled c o n t r o l of the o x i d a t i o n of s p u t t e r e d z i n c and subsequent o p t i m i z a t i o n of heat m i r r o r f i l m s . To complete t h i s study, the e l e c t r i c a l , o p t i c a l and s t r u c t u r a l p r o p e r t i e s of f i l m s were f u l l y c h a r a c t e r i z e d i n r e l a t i o n to f i l m s t o i c h i o m e t r y and the d e p o s i t i o n p r o c e s s . 2 Organization of t h i s thesis follows the chronological order of research. A review of heat mirror technologies i s provided in Chapter 1, along with reasons for the choice of fil m material and deposition process. The apparatus and techniques used to produce ZnO heat mirrors, and a preliminary evaluation of these heat mirrors, are described in Chapter 2. Chapters 3 and 4 are devoted to an in-depth study of the bias sputter process and Chapter 5 reports a f u l l characterization of ZnO films of varied degrees of oxidation. 1.2 HEAT MIRRORS Two general classes of heat mirror are presently recognized; heterostructure thin f i l m stacks and single layer semiconductor coatings. A heterostructure heat mirror i s a metal film sandwiched between d i e l e c t r i c coatings. - A very thin (10 to 20 nm) metal f i l m , such as gold or s i l v e r , provides infrared r e f l e c t i v i t y and i s partly transparent in the v i s i b l e . This transparency i s enhanced by a n t i r e f l e c t i o n coatings of a transparent and protective d i e l e c t r i c such as T i 02. A wide variety of metals and d i e l e c t r i c s have been used in heterostructure fa b r i c a t i o n and include I8nm Ti02/I8nm Ag/I8nm T i 02 [2], 45nm Bi203/I3nm Au/45nm B i203 [3], and 15nm ZnO/14nm Ag/I5nm ZnO [4]. Some of these films were o r i g i n a l l y developed as energy saving incandescent l i g h t bulb coatings. The Southwall Corporation [10] has already overcome the severe technological d i f f i c u l t i e s of thickness and composition control for heterostructure coatings manufactured in large 3 quantities on r o l l s of p l a s t i c sheet. Typical o p t i c a l c h a r a c t e r i s t i c s for heterostructures are a maximum transmission of 85% at wavelength 500 nm and a re f l e c t i o n exceeding 80% for wavelengths greater than 3 urn. By increasing the metallic layer thickness, heterostructures may be configured as 'cooling load' heat mirrors for hot climates. Lower transmission of solar energy and r e f l e c t i o n of atmospheric blackbody radiation . would reduce the air-conditioning load of a large o f f i c e b u i l d i n g . Lampert [5] gives a review of the deposition techniques and performance of heterostructure heat mirrors. Single layer heat mirror films are a subset of a larger class of materials known as transparent conductors. Applications of transparent conductors are not limited to radiative i n s u l a t i o n , but include thin f i l m devices such as heating elements for windshields, a n t i s t a t i c coatings, gas sensors, and electrodes for electrochromic or l i q u i d c r y s t a l d i s p l a y s . The most common transparent conductors are the semiconducting oxides Sn02, l n203 , and Cd2SnOi,. A review of research on these materials prior to 1975 i s given by Vossen [6], and the f l o u r i s h of recent a c t i v i t y in preparation and characterization of transparent conductors has been reviewed by Chopra et a l . [7], Jarzebski [8], and Lampert [5]. An imcomplete l i s t of some less common transparent conductors includes ZnO, CdS, Cu2S, NiO and W03. A l l these single layer materials are v i s u a l l y transparent 4 due to an o p t i c a l bandgap exceeding 2.8 eV. Conduction electrons are generated by dopants (such as Sn in l n203) or through deviations from stoichiometric composition (such as oxygen vacancies in Sn02). To achieve infrared r e f l e c t i o n necessary for a heat mirror, exceptionally low r e s i s t i v i t i e s , P<3X10~3 Bern, are required. As c a r r i e r m o b i l i t i e s in transparent thin films are t y p i c a l l y about 20 cm2/Vs, large c a r r i e r densities greater than 1 02° cm"3 are necessary. For films of thicknesses greater than 400 nm, strong infrared r e f l e c t i o n at wavelengths greater than about 4 nm is created by the plasma of conduction electrons, as predicted by the Drude model of free electrons [9]. I decided to study single layer heat mirror films due in part to the challenging problem of precise stoichiometry control during deposition - and to the lack of existing high rate manufacture of such f i l m s . Single layer heat mirrors have previously been fabricated in research deposition systems using only Sn02, l n203 and Cd2Sn04 [5]. As a result of my thesis research, good transparency and strong infrared r e f l e c t i v i t y have been obtained in ZnO [11,12] and i t i s now a recognised heat mirror material. Also, I have f u l l y characterized a high rate planar magnetron sputtering technique suitable for deposition of ZnO onto low temperature glass or p l a s t i c substrates [13]. These developments have occurred simultaneously with reports by other workers of heat mirror quality ZnO (p<3xl0~3 ficm) produced by spray pyrolysis [7], reactive evaporation [14], and rf sputtering'[15,16]. 5 1.3 DRUPE THEORY OF FREE ELECTRONS The i n f r a r e d r e f l e c t i o n propert ies of a semiconductor heat mirror may be modelled by the theory of free electrons as developed by Drude, Zener and Kronig [9] . Conduction electrons are free to move throughout the l a t t i c e , with the e f f e c t s of scat ter ing modelled simply by a damping y equal to the r e c i p r o c a l mean free time between sca t ter ing events. The d e r i v a t i o n of o p t i c a l constants i s e a s i l y i l l u s t r a t e d in one dimension, where the equation of motion for an electron in an o s c i l l a t o r y f i e l d i s ; m*d2y + m * 7 d y = -eE 0 exp( iut ) (1.1) d T * dt where m* is the electron e f f e c t i v e mass. The complex conduct iv i ty a i s ca lcula ted from the current dens i ty ; J = -Nedv = oE 0exp( icot) (1.2) dt where N i s the density of e l e c t r o n s . The complex d i e l e c t r i c constant follows immediately from the c o n d u c t i v i t y ; e = e 1 + i e 2 = eoO + i a / e 0 w (1.3) where £ w i s the high frequency l a t t i c e c o n t r i b u t i o n to the d i e l e c t r i c constant. The o p t i c a l constants n and k for the doped semiconductor are then given by; e, = n 2 - k 2 = - Ne 2 1 e0m* (co 2+7 2 ) (1.4) e2 = 2nk = Ne 2 y e0m* odj' + y* ) The damping 7 i s interpretated as 7=e/m*M, where M is the e lec t ron m o b i l i t y . For strongly absorbing f i lms in the IR, the t o t a l r e f l e c t i o n i s simply the r e f l e c t i o n at the a i r - f i l m 6 interface and i s ca lcula ted from; r = (n-1) 2+ k 2 (1.5) (n+1)2+ k 2 This c a l c u l a t i o n of r e f l e c t i v i t y requires that f i lms have a thickness that i s large compared to the skin depth for i n f r a r e d photons and is s a t i s f i e d by thicknesses greater than 0.3 /xm for t y p i c a l transparent conductors. The plasma frequency, u? , characterizes the o p t i c a l propert ies of the electron plasma by separating the r e f l e c t i v e low frequency behaviour from the transparent high frequency behaviour. wp i s defined when n 2 -k 2 =0 and i s given by; «P = Ne 2 - T2 (1.6) Groth and Kauer [18] have rewritten equation (1.4) using only u>p, 7 and [18]. Figure 1.1 shows the behaviour of r e f l e c t i o n with wavelength for ^=4 ( t y p i c a l for a semiconducting oxide) and at various values of the r a t i o cjp /7 . The p o s i t i o n of- the r e f l e c t i o n edge i s determined p r i m a r i l y by the number of c a r r i e r s , N, whereas the steepness of the edge i s c o n t r o l l e d by the r a t i o up/y. To ensure r e f l e c t i o n of room temperature IR r a d i a t i o n (6 to 25 Mm wavelength), a r e f l e c t i o n edge near X=3 Mm i s d e s i r a b l e , and corresponds to a c a r r i e r density of approximately 3 x l 0 2 ° cm" 3 in indium t i n oxide. A steep r e f l e c t i o n edge i s a lso necessary, requir ing a small value of 7 or a large e lec t ron m o b i l i t y . For the example of indium t i n oxide, a value of n-20 cm 2 /Vs gives o>p/7=2. In summary, the c h a r a c t e r i s t i c s of an o p t i c a l l y transparent semiconductor f i l m for heat mirror a p p l i c a t i o n s 7 Figure 1.1 The r e f l e c t i v i t y of a free electron plasma for various ra t i o s of plasma frequency CJ to electron damping 7 . 8 must be; 1) a f i l m t h i c k n e s s exceeding 300 nm, 2) a c a r r i e r d e n s i t y near 102 0 cm"3 and 3) a m o b i l i t y exceeding 10 cm2/Vs. Often only t h e ' r e s i s t i v i t y , p=l/Ne*j ( t y p i c a l l y 10~3 ftcm) i s quoted as a f i g u r e of merit for a transparent conductor or heat m i r r o r . 1.4 CHOICE OF ZnO An i d e a l s i n g l e l a y e r heat m i r r o r f i l m f o r c o l d c l i m a t e s should possess the f o l l o w i n g c h a r a c t e r i s t i c s . S t a r t i n g m a t e r i a l s should be inexpensive, and the f i l m should d i s p l a y no degradation of o p t i c a l p r o p e r t i e s with weathering. The f i l m should be o p t i c a l l y t r a n s p a r e n t , which i m p l i e s a bandgap great e r than 3 eV, and t h i s transparency should extend to near 2 Mm wavelength to reap f u l l b e n e f i t s from the s o l a r energy spectrum. For commercial a p p l i c a t i o n s , i t i s advantageous to depo s i t heat m i r r o r s on r o l l s of p l a s t i c sheet, r e q u i r i n g a d e p o s i t i o n technique that does not r e l y on high temperature s u b s t r a t e s . An examination of the s t a t e of the a r t of heat mi r r o r m a t e r i a l s and technology a v a i l a b l e in 1981 (at the s t a r t of my t h e s i s research) p r o v i d e s m o t i v a t i o n and some j u s t i f i c a t i o n f o r pursuing z i n c oxide as a heat m i r r o r c a n d i d a t e . I n 2 0 3 i s the most common m a t e r i a l f o r semiconductor heat m i r r o r s . C o n d u c t i v i t y i s achieved through non-stoichiometry (oxygen d e f i c i e n c y ) , or by doping with about 5 atomic % F or Sn. ITO i s a common a b b r e v i a t i o n f o r indium t i n oxide ( I n 2 03: S n ) . High c o n d u c t i v i t y f i l m s have been produced by 9 sputtering [29,30,31], reactive evaporation [26,27], chemical vapour deposition (CVD) [32] and spray pyrolysis [17,28,33]. Table 1.1 shows the t y p i c a l c h a r a c t e r i s t i c s of ITO and other Table 1.1 Typical properties of oxide heat mirrors. Material R e s i s t i v i t y (ficm) Bandgap (eV) IR Reflection (% at 10mn) ITO 4x10-" 3.6 85 Sn02 7x10" • >3.5 80 CTO 4x10-" 2.9 84 oxide heat mirrors. These figures are only representative of the c h a r a c t e r i s t i c s of the best conducting f i l m s , since f i l m properties vary greatly with deposition technique. It i s seen that ITO i s an excellent heat mirror. The only disadvantage ITO may have i s the high cost of bulk indium, approximately $10/ounce for 99.99% purity in 1985 [19], which may l i m i t the cost effectiveness of ITO compared with other coatings. Conducting Sn02 has been doped through oxygen deficiency, but the films with lowest r e s i s t i v i t i e s (<10~3 ficm) are doped with F or Sb and have been deposited only by CVD or spray pyrolysis [5], techniques requiring substrate temperatures in excess of 600 K. These high deposition temperatures may be necessary for substitutional incorporation of dopant atoms at e l e c t r i c a l l y active Sn s i t e s [8]. Films of Sn02:F have been shown to exhibit good thermal s t a b i l i t y of e l e c t r i c a l properties after subjection to temperatures exceeding 500 K [20]. 10 C d2S n O „ (CTO) films of r e s i s t i v i t y 4x10" * Gem have been prepared by diode sputtering [21,22] and planar magnetron sputtering [23] without heated substrates. E l e c t r i c a l conduction i s n-type, created by oxygen vacancies. CTO i s known to have good d u r a b i l i t y and chemical resistance [24]. The bandgap i s estimated at 2.1 eV, but the Moss-Burstein effect of bandgap expansion due to conduction band f i l l i n g can increase the bandgap up to 2.9 eV [25], or a wavelength of 430 nm. However, the transmission in the wavelength region 350 to 450 nm remains s i g n i f i c a n t l y lower than that of ITO and Sn02 heat mirrors. It i s seen from the previous discussion that ITO, CTO and Sn02 a l l have some c h a r a c t e r i s t i c that renders them unsuitable as an " i d e a l " heat mirror. ZnO was of interest as a new potential heat mirror material because i t i s inexpensive, has an o p t i c a l bandgap near 3.2 eV, and, prior to the start of my research in 1981, had been deposited by reactive evaporation [34] and spray p y r o l y s i s [35] at r e s i s t i v i t i e s between 10"3 and 10"2 Ocm. Later, in 1981, Webb and Williams [15] showed that low r e s i s t i v i t y ZnO (2x10"3 Ocm) could be produced by rf sputtering. These developments were motivated by the application of ZnO to transparent window electrodes for solar c e l l s and, as such, were not concerned with optimization of infrared r e f l e c t i o n properties. 11 1.5 PROPERTIES OF ZnO The c h a r a c t e r i s t i c s of single c r y s t a l and sintered ZnO were the subject of intensive study in the 1950's due to i t s varied use as a phosphor, enamel pigment and c a t a l y s t . These developments were reviewed by Heiland et a l . in 1959 [36], and have formed a strong groundwork for understanding the o p t i c a l , e l e c t r i c a l and mechanical properties of ZnO thin f i l m s . ZnO i s a d i r e c t bandgap, n-type semiconductor that c r y s t a l l i z e s in the hexagonal wurtzite l a t t i c e . Figure 1.2 shows the electronic energy levels of important native imperfections in ZnO, as estimated by Kroger [37]. The f i r s t donor levels of oxygen vacancies and zinc i n t e r s t i t i a l s are located about 0.05 eV below the conduction band edge. Previously, Zn i n t e r s t i t i a l s were believed to be active in providing conduction electrons, as evidenced primarily by experiments of Zn d i f f u s i o n into ZnO single c r y s t a l s [36]. But the chemical nature of the shallow donor in ZnO i s presently controversial due to ESR detection of donors from oxygen vacancies [38], Dopants of Ga and In form e l e c t r i c a l l y active shallow donors that reside s u b s t i t u t i o n a l ^ at Zn s i t e s [39,40]. The electron e f f e c t i v e mass has been measured by Baer [41] using Faraday rotation and i s found to be 0.24mo. Recently there has been great interest in stoichiometric thin films of ZnO for use in acoustic wave devices [42,43] and thin f i l m pressure transducers [44]. ZnO i s p i e z o e l e c t r i c and possesses a high electromechanical coupling c o e f f i c i e n t . Whereas conducting films require doping through Zn excess, 12 Conduction Band 1.05 eV | Zn; 1.05 eV .45 eV Znt 3.2 eV 1.8 eV Valence Band F i g u r e 1.2 E l e c t r o n i c energy l e v e l s of n a t i v e i m p e r f e c t i o n s i n ZnO at temperature 300 K. 13 films for acoustic wave devices are ide a l l y defect-free and have a preferred orientation of p o l y c r y s t a l s . 1.6 DEPOSITION PROCESSES A deposition process ideal for commercial manufacture of heat mirrors should be high rate with the a b i l i t y to coat p l a s t i c substrates, should have good control of film composition and uniformity throughout the coating, and should u t i l i z e inexpensive target materials. A brief survey of standard deposition techniques i s given below, followed by comparison to, and description of, planar magnetron sputtering. In chemical vapour deposition (CVD), gases are introduced to a chamber at a constant flow rate where they react and form a coating on a heated substrate. Roth and Williams [45] have deposited ZnO by reacting gases of diethyl zinc and oxygen on substrates with temperatures in the range 560 to 760 K. Deposition rates approached 40 nm/min, which i s considered slow for CVD [46]. Fundamentally similar to CVD i s spray p y r o l y s i s , where chemicals are sprayed onto a heated substrate in aerosol form. A high consumption of chemical occurs due to vapourization of many aerosol droplets before they react on the substrate. Aranovich et a l . [35] have sprayed an aqueous solution of ZnCl2 onto substrates of temperature 620 to 760 K. Coatings of ZnO were formed and a waste product of HC1 gas li b e r a t e d . Typical deposition rates were 60 nm/min. 14 Morgan and Brodie [14] have recently described a reactive evaporation process capable of depositing high conductivity ZnO. Zn metal was evaporated through a low pressure oxygen discharge onto unheated glass substrates at deposition rates of about 20 nm/min. The oxygen pressure, power of the discharge, and Zn evaporation rate were controlled independently to optimize the oxidation of z i n c . It i s seen that the fundamental l i m i t a t i o n of CVD and spray pyrolysis i s the requirement of heated substrates to activate the deposition process, rendering them unsuitable for coating of p l a s t i c sheet. The reactive evaporation technique shows excellent promise i f the d i f f i c u l t i e s of composition and thickness control over large areas can be overcome, and i f the deposition rate may be increased. Alternatives to the above-mentioned deposition processes are the wide variety of sputtering configurations and techniques, reviewed in d e t a i l by the standard texts of Vossen and Kern [47] and Bunshah [48]. Of these, magnetron sputtering (planar or c y l i n d r i c a l geometry) i s the recognised high rate sputter process and w i l l be described in some d e t a i l . A glow discharge plasma (usually of Ar) i s created in front of a cathode (the target material) by rf or dc ex c i t a t i o n . Electrons flowing from the cathode cross the positive space charge sheath or darkspace surrounding the cathode and then ionize Ar atoms in the glow region by electron impact i o n i z a t i o n . These Ar* ions bombard the cathode, and by momentum transfer target atoms are sputtered 15 from the surface and may subsequently deposit on a substrate situated beyond the glow discharge plasma. The plasma i s self-substaining through the emission of secondary electrons from the target due to Ar* ion bombardment. Magnets behind the target generate a magnetic f i e l d above the target (hence 'magnetron') that constrains electrons to movement near the target surface, thus increasing the e f f i c i e n c y of Ar ionization and subsequently the ejection rate of target material. A further advantage of a magnetron source i s that the substrate i s not heated by stray electrons from the cathode. Dc sputtering i s r e s t r i c t e d to conducting targets, such as Zn metal, whereas rf sputtering must be used for non-conducting pressed oxide targets such as ZnO. Reactive sputtering involves the addition of gas(es) that react at the target and/or the substrate to form a desired compound. Webb and Williams [15] have used rf reactive magnetron sputtering from a ZnO target to deposit conducting ZnO films at deposition rates near 6 nm/min. A serious disadvantage of sputtering from an oxide target i s the low sputter y i e l d (atoms ejected per bombarding Ar* ion) compared to that of metal targets. Zinc has one of the highest sputter y i e l d s of any element, approximately 4.6 at 300 V [137], whereas ZnO has a y i e l d of 1.1 [85], Also, oxide targets tend to be more expensive and more d i f f i c u l t to machine than metal targets. Reactive magnetron sputtering from a metal target in an Ar/02 atmosphere was the deposition technique used for my 16 thesis work. It has deposition rates of t y p i c a l l y 20 to 100 nm/min and does not necessarily require heating of the substrate to activate f i l m deposition. The geometry of a planar magnetron system may be e a s i l y scaled up to provide deposition and control over large areas [49]. Zinc targets are r e l a t i v e l y inexpensive and easy to machine to s p e c i f i c a t i o n s . The major d i f f i c u l t y of this technique l i e s in maintaining a metal target surface to ensure high rate sputtering, while f u l l y oxidizing the sputtered zinc that arrives at the substrate. During magnetron sputtering of ITO and Sn02, Howson et a l . [50] promoted f i l m oxidation by developing a subsidiary rf glow discharge at the substrate. This technique, known as ion p l a t i n g or bias sputtering, has been used for many years to improve the mechanical- properties of films and the t r i b o l o g i c a l properties of machine tools and bearings. The rf power i s usually monitored by the negative dc s e l f - b i a s voltage that occurs at the substrate (hence 'bias sputtering'). A f u l l review of the ion platin g process i s given in Chapter 4, where i t i s more timely. It s u f f i c e s to say here that the major mechanism of ion platin g was believed by Howson to be generation of ionized oxygen species and acceleration of these species towards the negatively biased f i l m . Of less importance, the bombarding species were believed to add surface energy to the f i l m and consequently enhance c r y s t a l growth [51]. In tandem with an rf discharge at the substrate, Maniv et 17 a l . [23] u t i l i z e d a reactive gas b a f f l e at the target, which assisted in keeping the metal target surface free from oxidation in the Ar/02 sputtering gas mixture. A metal box with a series of slot shaped apertures was b u i l t surrounding the planar magnetron target. This geometry decreased the oxygen flux to the target by providing a getter surface for oxygen near the target. In summary, i t was decided to investigate the potential of transparent, conducting ZnO_as a heat mirror, deposited by reactive planar magnetron sputtering. Metal targets are inexpensive and ZnO ' i s known to have a suitable bandgap of 3.2 eV. Zn would be sputtered from a metal target surface to enable high rate deposition, with a reactive gas ba f f l e protecting the target surface from oxidation. Substrates would be unheated and oxidation of sputtered Zn would be assisted by the application of a subsidiary rf discharge at the substrate. 18 CHAPTER TWO: HEAT MIRROR DEPOSITION 2.1 APPARATUS AND EXPERIMENTAL METHOD Figure 2.1 shows a schematic cross section of the planar magnetron sputtering chamber. The chamber was pumped by an o i l d i f f u s i o n pump and a freon cold trap. Regulation of pumping speed during sputter experiments was accomplished with a variable o r i f i c e . A 15 cm diameter Zn target of purity better than 99.99% was firmly clamped to a water cooled backing plate containing the magnetron assembly. The electromagnet was used to confine the glow discharge plasma in the shape of a torus d i r e c t l y in front of the target, creating sputtering erosion of the target in a ring pattern. The t o t a l area of the erosion region was 40 cm2. A c i r c u l a r reactive gas ba f f l e was attached to the ground shield that was positioned concentrically around the target. The b a f f l e i s shown in more d e t a i l in Figure 2.2. It was fabricated from aluminum, with a series of 0.64 cm diameter holes d r i l l e d in a concentric pattern through the front surface, so that the r a t i o of open area to t o t a l area of the ba f f l e was about 36%. The separation of the baff l e from the target surface was 4.5 cm, chosen so that the ba f f l e would not d i r e c t l y interfere with the glow region of the discharge. When in place, the grounded b a f f l e also served as the anode, in l i e u of the water cooled anode ring shown in Figure 2.1. The discharge was powered by a 5 kW, f u l l wave r e c t i f i e d , constant current dc power supply (Plasma-Therm 5000D, 19 Mass spectrometer rf power capacitance manometer t to diffusion pump F i g u r e 2.1 A schematic c r o s s s e c t i o n of the s p u t t e r i n g chamber. 20 Figure 2.2 The r e a c t i v e gas b a f f l e . About 36% of the top b a f f l e s u r f a c e i s open. 21 0-1000V). Pressures during sputtering were measured with an MKS Barytron capacitance manometer, which had a minimum resolution of 0.004 Pa. A UTI-100C quadrupole mass spectrometer was mounted in a side arm near the substrate holder. Flows of high purity Ar and 02 into the chamber were controlled to ±1% by independent leak valves (Granville P h i l l i p s #203) and measured by independent flow meters (Hastings #All-5 or A l l - 1 0 ) . The substrate holder and shutter assembly were e l e c t r i c a l l y isolated from the chamber and accomodated six 1x2 inch substrates of Corning 7059 glass or commercial grade quartz. Another substrate holder shown in Figure 2.3 was designed to demonstrate low temperature deposition of films onto 2 inch wide r o l l s of p l a s t i c sheet. The p l a s t i c was wound onto a c o l l e c t o r spool from a f r i c t i o n - r e s i s t i v e feeder spool, thereby stretching the sheet over an aluminum block positioned in front of the target. P l a s t i c s used in t h i s r o l l e r system included 6-foot lengths of 13, 25, 50, or 125 Mm thick polyester and 150 urn thick polyethylene. A secondary discharge at the substrate was created by an rf generator (ERAtron HFS-8005A at 13.56 MHz) coupled to the substrate holder through a matching network. This discharge was monitored by the negative dc s e l f - b i a s voltage occurring at the substrate and by the net forward power. A microcomputer system was used to control and/or for datalogging of a l l aspects of the sputtering and deposition processes including gas flow rates, t o t a l pressure, substrate 22 Polyester \ Sheet Chain drive To gearbox and feedthrough Figure 2.3 A schematic of the substrate holder used for f i l m deposition onto r o l l s of p l a s t i c sheet. 23 bias, and cathode voltage, current and power. This control enabled a maintainence of precise operating conditions throughout f i l m deposition. Before deposition, glass or quartz substrates were washed in trichlorethylene and isopropanol [52], P l a s t i c substrates were mounted a s ~ i s without any cleaning. After loading the substrates, the sputtering chamber was baked overnight (about 70°C) and pumped to a base pressure of 4x10"5 Pa. The d i f f u s i o n pump was throttled and Ar admitted at 3.0 seem (cm3/min at standard temperature and pressure) to a t y p i c a l operating pressure of 1.0 Pa. To erase the e f f e c t s of target history [53] and ensure repeatable starting conditions, the target was sputtered at 150 W in a pure Ar discharge for one hour. The desired flow of oxygen was admitted and again the discharge was l e f t to eq u i l i b r a t e for one hour before f i l m deposition. High p a r t i a l pressures of oxygen, pp(O2)>0.05 Pa, were measured with the capacitance manometer by subtracting the known Ar p a r t i a l pressure from the t o t a l pressure. The mass spectrometer, ca l i b r a t e d at high pp(02) with the capacitance manometer, was used to measure small oxygen p a r t i a l pressures, pp(O2)<0.05 Pa. A t y p i c a l experiment involved depositing films at various values of substrate bias, with a l l other parameters held constant. Deposition times were nominally 20 minutes per f i l m , giving f i l m thicknesses of the order of 800 nm. After sputtering, the chamber was baked and b a c k f i l l e d with dry nitrogen gas to atmospheric pressure to avoid condensation on the chamber walls. 24 Film thickness measurements were made primarily with a Talysurf 4 mechnical stylus instrument using sharp steps created in the fi l m by masking during deposition. Some thickness measurements were supplemented by measurements using o p t i c a l interference techniques, photography of cross-sections using a scanning electron microscope, and measurement of the mass of the deposited f i l m . These measurements generally agreed to within 10%. Film r e s i s t i v i t y measurements were made with a 4 point probe. Infrared r e f l e c t i o n measurements for the wavelength range 2.5 to 20 jim were made with a Perkin-Elmer 598 Infrared Spectrometer. Reflection and transmission spectra for wavelengths 350 to 2500 nm were taken by a Beckman UV5270 spectrophotometer with an integrating sphere attachment. The o p t i c a l spectra had absolute errors of ±2% for infrared measurements and ±1.5% for measurements with the Beckman spectrophotometer. Film morphology and structure were studied with the use of an X-ray d i f f T a c t o m e t e r , a scanning electron microscope, and a transmission electron microscope. 2.2 OPERATING CHARACTERISTICS In order to control fil m stoichiometry during growth, the re l a t i v e a r r i v a l rates of Zn atoms and 02 molecules at the substrate must be regulated. Such control would be achieved i f one could smoothly and monotonically vary the p a r t i a l pressure of 02 when reactively sputtering Zn from a metal target. However, Figure 2.4 shows that abrupt t r a n s i t i o n s and 25 pronounced hysteresis of 02 p a r t i a l pressure and cathode voltage occur as a function of the flow of oxygen, f ( 02) , admitted to the chamber. A target power of 150 W, an Ar flow rate of 3.0 seem and an Ar p a r t i a l pressure of 1.0 Pa were kept constant throughout t h i s experiment. These hysteresis e f f e c t s during reactive sputtering are well known [54,55] and are explained below. The p a r t i a l pressure of oxygen in the chamber pp(02), i s determined by the flow of oxygen f ( 02) , the pumping rate of 02 by the d i f f u s i o n pump, and the gettering action of sputtered Zn. Gettering i s the chemical combination of Zn deposits with oxygen and has a rate determined by the sputter rate of the target which, in turn, i s highly sensitive to the oxidation state at the target surface. It i s known that an oxide layer can form on the target.surface through two mechanisms [56]; chemisorption of neutral oxygen and ion plating of 02+ and 0+ species accelerated towards the surface. At low values of flow f ( 02) , pp(02) i s kept small mainly because of getter pumping of oxygen by sputtered Zn, and the target surface remains m e t a l l i c . As f ( 02) and consequently pp(02) are increased, an oxide layer begins to form on parts of the target surface where sputter erosion i s weakest. Since oxides have a much lower sputter y i e l d than metals, the sputter rate is reduced and creates a corresponding reduction in gettering, thus permitting further target oxidation. This positive feedback cycle ultimately produces runaway oxidation of the target surface. A sudden increase of pp(02) i s seen in Figure 26 0.6 0.4 D CM O fr 0 ^ tr p o o r ] n • • a a a n n a n o a METALLIC MODE 700 1 •°©§88<9gg° ooooo o o o u o j « 6 0 0 l • s o £ 5 0 0 r i-OXIDE I MODE ooooo oo°oooo I I 1 2 3 4 OXYGEN FLOW (cc/min) Figure 2.4 Hysteresis in the target voltage and oxygen p a r t i a l pressure as a function of increasing and decreasing oxygen flow rate. The target power was kept constant at 150 W. 27 2.4, coincident with a sudden decrease of cathode voltage while operating at constant power. The l a t t e r e f f e c t i s due to both the increased secondary electron c o e f f i c i e n t of zinc oxide on the target and the higher t o t a l sputtering pressure creating an increase in the cathode current density. In the reverse d i r e c t i o n of decreasing f ( 02) , the opposite t r a n s i t i o n occurs with marked hysteresis. It i s unfortunate that stable operating points do not exist in the t r a n s i t i o n regions, because these regions are prec i s e l y those of interest in producing oxygen de f i c i e n t f i l m s . Figure 2.5 shows the corresponding abrupt t r a n s i t i o n in current I versus cathode voltage V at constant f(O2)=2.0 seem. At voltages before the abrupt current decrease the target surface i s oxidized, while at large voltages the surface i s m e t a l l i c . A f f i n i t o and Parsons [56] have shown that for reactive gases that do not chemisorb on the metal target, target coverage i s controlled by only the ion plating mechanism. Consequently (due to the slower response of target coverage to ion plating f l u c t u a t i o n s ) , stable operating points can be maintained in the t r a n s i t i o n region. They were able to deposit films of stoichiometry A1NX for any value of x<1 by precisely c o n t r o l l i n g the n i t r i d e coverage of the aluminum target. Since oxygen chemisorbs onto Zn (as well as for heat mirror materials In, Sn and Cd), control of fil m stoichiometry for ZnO cannot be accomplished by varying the target surface composition, but must occur through control of oxidation of the growing f i l m at the substrate. 600 700 800 900 VOLTAGE (V) F i g u r e 2.5 The c u r r e n t - v o l t a g e c h a r a c t e r i s t i c of a Zn t a r g e t operated an Ar/02 d i s c h a r g e . T o t a l pressure was 1.0 Pa with f(Ar)=3.0 seem and f(O2)=2.0 seem. 29 Figure 2.4 shows the two fundamental operating modes of a Zn target in an Ar/02 discharge; metallic target surface and oxide target surface. Films deposited by operation in the oxide mode are characterized by high r e s i s t i v i t y , stoichiometric composition due to the large pp(02), and very low deposition rates. To obtain films with s l i g h t oxygen deficiency suitable as transparent conductors, normal target operation was in the metallic mode at f( 02) very close to the tra n s i t i o n such that pp(02) was maximized (about 0.03 Pa) near the substrate while maintaining a metal target surface of high deposition r a t e . These p a r t i a l pressures of 02 were not large enough to enable complete oxidation of sputtered zinc at the substrate. This d i f f i c u l t y was overcome through the use of a reactive gas baffl e at the target and an rf discharge at the substrate. As described in section 1.3, by providing a getter surface, the baffl e allowed a p a r t i a l pressure gradient of oxygen across the b a f f l e and, thereby, s u f f i c i e n t oxygen at the substrate for complete oxidation of the Zn f l u x . The rf discharge was believed to promote f i l m oxidation by excitation and ion plating of oxygen species. A nominal operating condition for the rf discharge was -100 V se l f - b i a s at a power of 20 W over a substrate holder of surface area 1070 cm2. 30 2.3 ZnO HEAT MIRRORS Many unsuccessful attempts were made to produce transparent ZnO from the metallic operating mode without the use of the baff l e and the substrate rf discharge. Films produced were brown or black in colour with r e s i s t i v i t i e s not lower than 10~2 Bern. Confirmation of t h i s d i f f i c u l t y in production of transparent, conducting ZnO by planar magnetron sputtering was recently provided by Maniv et a l . [57]. However, transparent films could be made by t r i p l i n g the standard target to substrate distance of 10 cm. The deposition rate at the substrate decreased by a factor of about 10, allowing ample time for each deposited Zn atom to oxidize in the small p a r t i a l pressure of 02. Such a technique required three hours for deposition of a fi l m of thickness 0.7 nm and obviously i s not desirable for commercial or even research purposes. A l l further ZnO films described in this thesis were produced using both the reactive gas ba f f l e and an rf discharge, unless e x p l i c i t l y stated otherwise. Figure 2.6 shows ZnO f i l m r e s i s t i v i t y and infrared r e f l e c t i v i t y at 10 am wavelength plotted as a function of the se l f - b i a s voltage induced at the substrate by the rf discharge. These f i l m s , of nominal thickness 900 nm, were deposited on 125 nm thick polyester sheet at a cathode power of 150 W, with f(02)=2.3 seem and f(Ar)=3.0 seem. At low negative bias the films were characterized by high r e s i s t i v i t i e s and a dark metallic appearance. As the rf power, and subsequently the dc bias, are increased, a 31 Figure 2.6 ZnO r e s i s t i v i t y ( c i r c l e s ) and infrared r e f l e c t i v i t y at 10 ym wavelength (squares) as a function of substrate bias for films deposited on polyester sheet. Figure 2.7 Transmission and r e f l e c t i o n c h a r a c t e r i s t i c s for ZnO heat mirrors deposited on both glass and polyester substrates. Normalized to 100% are the solar irradiance (AM1) and the infrared emission from a blackbody at 300 K. 33 t r a n s i t i o n i n f i l m c h a r a c t e r i s t i c s i s observed to low r e s i s t i v i t y and good o p t i c a l t r a n s p a r e n c y . A maximum i n i n f r a r e d r e f l e c t i v i t y occurred f o r the f i l m of lowest r e s i s t a n c e . T h is behaviour was t y p i c a l f o r that of many hundreds of ZnO f i l m s s p u t t e r e d . The f i l m d e p o s i t i o n rate of about 75 nm/min was l a r g e r than the rate of f i l m s s p u t t e r e d on g l a s s s u b s t r a t e s because the g l a s s s u b s t r a t e holder had a shutter i n f r o n t of the f i l m s that e f f e c t i v e l y reduced the d e p o s i t i o n r a t e by shadowing s p u t t e r e d Zn atoms a r r i v i n g at obliq u e angles of i n c i d e n c e . F i g u r e 2.7 shows t r a n s m i s s i o n and i n f r a r e d r e f l e c t i o n c h a r a c t e r i s t i c s of heat m i r r o r q u a l i t y ZnO produced on both 125 Mm t h i c k p o l y e s t e r and Corning 7059 g l a s s s u b s t r a t e s . For e v a l u a t i o n of heat m i r r o r performance, the s o l a r i r r a d i a n c e and the i n f r a r e d emission from a blackbody at 300 K are shown. The f i l m on p o l y e s t e r was 900 nm t h i c k with r e s i s t i v i t y 3x10"3 flcm whereas the ZnO on g l a s s was 700 nm t h i c k and had a r e s i s t i v i t y of 2x10"3 ficm. The i n t e r f e r e n c e s t r u c t u r e , seen in the tra n s p a r e n t region of these s p e c t r a , w i l l be u t i l i z e d in Chapter 5 to determine the o p t i c a l constants of ZnO. It should be noted that the t r a n s m i s s i o n c h a r a c t e r i s t i c s of both f i l m s were obtained using a double beam spectrometer with an uncoated s u b s t r a t e i n the re f e r e n c e beam. For the ZnO c o a t i n g on g l a s s , the average t r a n s m i s s i o n to s o l a r energy i s about 75% and the percent r e f l e c t i o n of r a d i a t i o n from a 300 K blackbody i s estimated as 85%. In comparison, the t r a n s m i s s i o n to s o l a r energy of uncoated g l a s s i s 92% and the average i n f r a r e d r e f l e c t i v i t y i s 7%. 34 2.4 ITO HEAT MIRRORS To enable comparison of ZnO to a standard heat mirror material, indium t i n oxide (ITO) heat mirrors were fabricated using the same deposition technique. The Zn target was substituted for an alloy target of composition 90% In and 10% Sn. Films were sputtered on glass and 13 m^ thick polyester substrates at a cathode power of 150W, f(Ar)=1.50 seem, f(02)=1.87 seem and a t o t a l gas pressure of 1.1 Pa. Figure 2.8 shows the solar transmission and infrared r e f l e c t i o n c h a r a c t e r i s t i c s of a 400 nm thick ITO f i l m on glass with r e s i s t i v i t y 9x10"" Rem. As with ZnO, the transmission spectra were obtained with an uncoated substrate in the spectrometer reference beam. Films produced on polyester had i d e n t i c a l infrared r e f l e c t i o n and r e s i s t i v i t y values. At the time of publication of these r e s u l t s , such r e s i s t i v i t i e s were believed to be the lowest reported for ITO on p l a s t i c sheet. The ITO films had heat mirror c h a r a c t e r i s t i c s very similar to the ZnO films; the ITO coating on glass had an average integrated solar energy transmission of 76% and an 87% r e f l e c t i o n of 300 K blackbody r a d i a t i o n . One advantage of the ITO films was the lower r e s i s t i v i t i e s obtained than in ZnO. Thus a thinner layer of ITO heat mirror could be used to exhibit the same infrared r e f l e c t i o n as ZnO. 35 WAVELENGTH (microns) F i g u r e 2.8 Transm i s s i o n and r e f l e c t i o n c h a r a c t e r i s t i c s of an ITO heat m i r r o r on a g l a s s s u b s t r a t e . Normalized to 100% are the s o l a r i r r a d i a n c e (AM1) and the i n f r a r e d emission from a blackbody at 300 K. 36 2.5 SUMMARY The properties of the f i r s t ZnO heat mirrors deposited by a sputter technique have been reported in t h i s chapter. During high rate, reactive sputter deposition onto p l a s t i c substrates, a reactive gas b a f f l e was used to control target oxidation and a substrate rf discharge enabled optimization of f i l m properties. The best ZnO heat mirrors had a transparency to solar radiation of 75% and an average r e f l e c t i o n of 85% for 300 K r a d i a t i o n . The only other report of heat mirror properties of ZnO i s by Morgan and Brodie [14], who produced films of transparency 80% and infrared r e f l e c t i v i t y 66% by reactive evaporation. The d e t a i l s of their c alculation of transparency were not given. Despite successful ZnO heat mirror deposition, the mechanisms of control of f i l m c h a r a c t e r i s t i c s by the rf discharge and of target oxidation by the b a f f l e were not at a l l well understood. For commercial application of ZnO coatings, a greater understanding of the deposition technique would be required. Thus the emphasis of my research was sh i f t e d from heat mirror optimization and testing to performing a series of experiments to demonstrate the effectiveness of the b a f f l e and, most important, to identify processes induced by the rf discharge that 'control and enhance f i l m properties. 37 CHAPTER THREE: THE REACTIVE GAS BAFFLE The technique of using a getter surface or reactive gas ba f f l e surrounding the target in order to enhance substrate oxidation was f i r s t used by S c h i l l e r et a l . [53] to produce oxides of T i and Ta and by Maniv et a l . [23] for Cd2SnO<, f i l m s . However, these authors did not provide any data concerning the c h a r a c t e r i s t i c s or operative mechanisms of the b a f f l e , but reported only f i l m properties produced with/without b a f f l i n g . The b a f f l e was presumed to decrease oxygen flux to the target by providing a getter surface and also decrease metal flux to the substrate, thus increasing the r a t i o of 02/metal atoms s t r i k i n g the substrate surface. The following investigations were undertaken to c l e a r l y demonstrate the effectiveness of a reactive gas b a f f l e . The p a r t i a l pressure of oxygen, pp(02), near the substrate was measured as a function of the flow of oxygen admitted to the chamber for cases of sputtering with and without a ba f f l e surrounding the target. The results are shown in Figure 3.1 for a cathode power of 150 W and an Ar p a r t i a l pressure of 0.75 Pa in both cases. The cathode was operating in the metallic mode of target coverage. The deposition rate for Zn sputtered in a pure Ar discharge with a baf f l e was found to vary between 25 and 40% of the deposition rate without a b a f f l e . The va r i a t i o n was due to accumulation of sputtered deposits and clogging of holes in the baff l e over the course of many experiments. C l e a r l y , the use of a b a f f l e creates a substantial (Tj ^ & 0.07 to £ 0.05 a. ao3 CM 0.01 o B A F F L E * NO . B A F F L E • • 2 4 6 0 2 FLOW RATE (cc/fnin) F i g u r e 3.1 Enhancement of oxygen p a r t i a l p ressure near the s u b s t r a t e through the use of a r e a c t i v e gas b a f f l e . The t a r g e t power was 150 W f o r both c a s e s . 39 increase in oxygen p a r t i a l pressure at the substrate, and decreases the deposition rate of sputtered metal. The combination of these e f f e c t s provides a four or f i v e - f o l d increase in 02/Zn a r r i v a l rates at the substrate, enhancing the oxidation of growing f i l m s . The baffle i s able to u t i l i z e about 65% of the sputtered metal flux to getter oxygen in the target v i c i n i t y . Even for low flow rates of 02 (2 seem), the remaining metal flux passing through the b a f f l e is not enough to completely getter a l l oxygen in the much greater chamber volume, resul t i n g in an increased oxygen p a r t i a l pressure near the substrate. In addition, the long molecular mean free path (about 1cm at 1Pa) and decreased molecular conductance presented by the baffle help maintain an 02 pressure d i f f e r e n t i a l between the substrate region and the target region. A further advantage of the baffle was discovered in conjunction with use of an rf discharge at the substrate. Kominiak [58] has observed minor interactions between an rf powered discharge at the target and a dc biased substrate. Similar interactions between a dc target discharge and an rf substrate discharge might hamper the precise control required for bias sputtering of films of s p e c i f i c stoichiometry. In Figure 3.2 the dc target current is shown as a function of substrate bias for cases with and without a reactive gas b a f f l e . In both cases the target power i s kept constant at 150 W. A small increase in target current with increasing substrate bias i s seen for the case of no b a f f l e . This dc 40 F i g u r e 3.2 E f f e c t of the r f s u b s t r a t e d i s c h a r g e on the dc t a r g e t o p e r a t i n g c h a r a c t e r i s t i c s . The t a r g e t power was h e l d constant at 150 W f o r cases with and without a r e a c t i v e gas b a f f l e . 41 c u r r e n t i n c r e a s e i s a t t r i b u t e d to a feedin g of ions from the r f discharge to the dc d i s c h a r g e , due to the l a r g e s p a t i a l extent of the r f glow r e g i o n . C l e a r l y the r f d i s c h a r g e has a n e g l i g i b l e e f f e c t on the I-V c h a r a c t e r i s t i c s of the t a r g e t when the b a f f l e i s i n c l u d e d . Thus an added b e n e f i t of a r e a c t i v e gas b a f f l e i s to reduce the r f - d c plasma i n t e r a c t i o n and subsequent i o n - f e e d i n g by s e p a r a t i n g the two plasmas with a p a r t i a l l y open grounded p l a n e . 42 CHAPTER FOUR: THE SUBSTRATE RF DISCHARGE 4.1 PROPERTIES OF AN RF GLOW DISCHARGE Several authors have recently provided excellent reviews of the properties of an rf glow discharge, s p e c i f i c a l l y for applications to sputtering [59,60,61], In my sputtering system, rf power i s applied to the substrate through a standard L-C matching network that includes a blocking capacitor in s e r i e s . In t h i s configuration, the substrate i s c a l l e d a 'capacitively coupled' electrode with the series capacitor preventing a dc short to ground. Electrons have a much higher mobility in a glow discharge plasma than massive argon ions. Thus, a large electron current can flow during the positive half of an rf voltage applied at an electrode. However, during the negative c y c l e , only a r e l a t i v e l y small ion current can flow. To conserve charge the substrate must s e l f - b i a s to a negative dc voltage u n t i l the net current averaged over each cycle i s zero. The negative dc potential creates a darkspace or sheath above the surface of the substrate electrode, in the same manner as the dc powered cathode. The voltage established at the substrate may be written; V(t) = V0 + v,sina>t (4.1) where the absolute value of the dc voltage, V0, i s not greater than V,. For my system, f = co/27r = 13.56 MHz. A simple model of the rf discharge assumes that the low mobility ions only see the dc voltage, whereas electrons may o s c i l l a t e in the rf 43 f i e l d . These o s c i l l a t i n g electrons can acquire s u f f i c i e n t energy to ionize argon atoms, thus maintenance of the glow discharge does not rely solely on secondary electrons emitted from the substrate. The voltage across the sheath at the substrate , Vs , i s determined by the difference between the plasma voltage VP and the bias voltage, V0 ; Vs =Vp -V0. It i s known that the plasma f l o a t s at a potential higher than the potential of any surrounding surfaces. Coburn and Kay [63] have shown that Vp i s dependent on the r a t i o of target area to the area of a l l other surfaces in contact with the glow. The sputter system of Figure 2.1 has a substrate to chamber ground r a t i o of 1070cm2/6000cm2 or 0.17. This large asymmetry of electrodes leads to small plasma voltages, VP=10 V, at a bias of -100 V [63]. Thus to a f i r s t approximation, the voltage across the substrate sheath may be equated to the bias voltage, Vs =V0. It i s evident that for Vp>0, a sheath w i l l also be formed at the ground electrode (chamber w a l l ) . The r a t i o of electrode sheath voltages has been predicted by Koenig and Maissel [66] to vary with the fourth power of the inverse area r a t i o ; Vs1/^2 = ( A 2 ) V ( A , ) » (4.2) Experimental results confirm t h i s general behaviour but have shown the exponents in (4.2) to be closer to 1 [59]. Of interest in this work, i t i s seen that for strongly asymmetric electrodes, the smaller electrode (substrate) w i l l receive high energy bombardment whereas the large electrode (chamber 44 wall) receives a p a r t i c l e flux of much lower energy. To a s s i s t in interpretation of the ef f e c t s of an rf discharge, i t i s useful to estimate the energy d i s t r i b u t i o n and flux of ions a r r i v i n g at an rf powered substrate. Due to the separate current contributions of electrons and ions, the ion current may not be calculated from the power-voltage c h a r a c t e r i s t i c of the discharge, but from c h a r a c t e r i s t i c s of the substrate sheath. Thornton and Penfold [62] have modified the Child-Langmuir law of space charge limited current to enable calc u l a t i o n of the ion current density to the substrate for r f discharges; JD2 = 46(40/Mjk (Vs/1000)V2 (4.3) where D i s the maximum sheath thickness in mm, J i s the ion current density in mA/cm2, M i s the ion mass in atomic units and Vs i s the voltage across the sheath. To obtain an estimate of the ion current to the substrate, an observed darkspace D=6 mm and M=40 (Ar) were substituted. For a dc bias V0=-100 V the ion plating current i s J=3X101* ions/cm2s, i f i t i s assumed that VS=100 V. Ions accelerated across the substrate sheath lose energy through two processes; symmetric charge transfer c o l l i s i o n s and through rf modulation of the sheath voltage. The l a t t e r mechanism s h a l l be discussed f i r s t . If the t r a n s i t time for an ion to cross the sheath i s less than the period of the rf o s c i l l a t i o n (74ns), then upon reaching the substrate the ion w i l l have an energy related to the rf phase at that instant. The spread of the ion energy d i s t r i b u t i o n , dE, due to rf 45 modulation was given by Okamoto and Tamagawa as [64]; dE = (63V"i/37rfD) (2eV0/M)'^ (4.4) where V0 i s the dc bias, V, i s the amplitude of rf modulation of the sheath voltage (approximately equal to V0 for an asymmetric configuration), f =13.56 MHz, and M i s the ion mass. Substituting values of V0=V1=100 V yields an energy spread of only 11 eV and a corresponding ion transit time of 870 ns. These results are compatible with the previous comments concerning the slow response of heavy Ar+ ions. The energy spread i s small enough to be neglected in further c a l c u l a t i o n s . The ion energy d i s t r i b u t i o n created by symmetric charge transfer c o l l i s i o n s in the sheath ( A r * + A r — » A r + A r * ) was shown by Davis and Vanderslice [65] to depend on the sheath thickness to ion mean free path r a t i o , D/L. Although this model was developed for dc discharges, i t i s used here to estimate the response of Ar* ions to the dc component of the substrate voltage. Using the assumptions that the e l e c t r i c f i e l d decreases l i n e a r l y from the cathode to the edge of the glow and that newly created ions start from r e s t , they derived the following expression for the d i s t r i b u t i o n of ion energies at the cathode; dN = N^ D 1 exp[-D/L( 1-( 1-V/VS )'/z)] (4.5) dV Vs L 2(1-V/VS )'£ where N is the number of ions of energy eV, and V5 is the voltage across the sheath. Using a charge transfer cross section for Ar of 3x10"15 cm2 at 100 eV [65] gives a mean free path of 3 mm at 1 Pa pressure. Figure 4.1 shows the energy 46 • • ' 1 L _ 20 4 0 6 0 8 0 100 ION ENERGY (eV) Figure 4.1 Energy d i s t r i b u t i o n of ions a r r i v i n g at the substrate, calculated from the theory of Davis and Vanderslice using a sheath thickness to ion mean free path r a t i o of 2 and a substrate bias of -100 V. Not indicated on t h i s graph, 13% of the ions suffer no c o l l i s i o n s and ar r i v e at the substrate with energy 100 eV. 47 d i s t r i b u t i o n for D/L = 6mm/3mm. The d i s t r i b u t i o n of ion energies i s found to be nearly constant from zero v o l t s to the f u l l sheath voltage, and 13% of the ions suffer no c o l l i s i o n s . A further experimental confirmation of the model of Davis and Vanderslice was recently made by Machet et a l . [67], In summary, the s i g n i f i c a n t properties of an rf-induced bias of -100 V at the substrate are the bombardment by a flux of 3 x l 01 a ions/cm2s with a d i s t r i b u t i o n of energies widely spread about an average of 65 eV, but not exceeding 100 eV. This characterization i s pertinent to an understanding of the e f f e c t s of substrate bias described in the following section. 4.2 EFFECTS OF AN RF DISCHARGE The substrate ion plating or bias sputter method used in th i s work i s a s p e c i f i c technique of a wide variety of existing ion assisted deposition processes. The f i r s t recent use of f i l m deposition using acclerated ions was by Mattox in 1964 [68] to enhance the f i l m adhesion of metallic a l l o y s . Subsequently ion assisted processes have been rapidly adopted in research and industry and include ion p l a t i n g , plasma activated reactive evaporation, plasma assisted chemical vapour deposition and plasma n i t r i d i n g . Recent reviews of the properties and effects of these processes are given by Mattox [69,70], Greene and Barnett [71], Thornton [72] and Coad and Dugdale [73]. Ion bombardment of a growing film can affect e s s e n t i a l l y a l l f i l m properties. E f f e c t s include resputtering, defect 48 production, inert gas incorporation, heating, stress and crystallographic changes, disruption of surface morphology and changes in f i l m composition. Although a large fraction of research e f f o r t has been expended on studies of bombardment e f f e c t s , e s p e c i a l l y on metal or metal a l l o y f i l m s , these effects are not as yet completely understood. Ion bombardment effects may be generally c l a s s i f i e d by the energy of bombardment [72,69]. At low ion energies <50 eV, loosely bonded impurity or surface atoms may be desorbed by sputtering. For instance, incorporation of impurity Ar atoms in the fi l m may be minimized. For incident ion energies 50 to 150 eV, sputtering and heating of the growing f i l m w i l l increase, coincident with possible crystallographic disruption and/or composition change. At high ion energies >>150 eV implantation of bombarding species w i l l occur and subsequently greatly disrupt or amorphize f i l m c r y s t a l l i n i t y . Such defect production may be counteracted by s i g n i f i c a n t f i l m heating and sputtering. Ion assisted deposition of metals i s known to densify coatings and improve adhesion to the substrate. Densification is characterized by an elimination of columnar or dendritic morphology through the mechanisms of enhanced adatom mobility and material r e d i s t r i b u t i o n by resputtering. The composition dependence on substrate bias for metal a l l o y films has been modelled by several workers [74,75,76], who have shown that metal constituents with higher elemental sputter y i e l d s are p r e f e r e n t i a l l y resputtered from the growing f i l m surface. 49 Ion platin g has been used to optimize the properties of many react i v e l y sputter deposited metal oxides and n i t r i d e s including ITO [51], T i 02 [77], CTO [23] and TiN [78,79] but su r p r i s i n g l y l i t t l e investigation of the mechanisms of such 'reactive ion p l a t i n g ' has been performed, as most researchers have been concerned with characterization of fil m properties. An exception i s the work of Barnett and Greene [89], who have successfully used a resputter model similar to the models used for metal a l l o y s , to predict the composition of GaAs films sputtered from a GaAs target. Also, Winters and Kay [80] have bias sputtered W, Au and Ni in Ar/N2 gas atmospheres. They found an increase in N content with bias for Au, presumably due to implantation of N2*. W and Ni had a decreasing N content with bias, believed to be due to the large sputter y i e l d of chemisorbed nitrogen gas for these metals. For reactive ion platin g of oxide films i t has been assumed but not yet proven that fil m oxidation i s enhanced through bombardment of 02+ ions and that surface energy created by energetic ions enhances nucleation and c r y s t a l growth [51,23]. The following sections undertake an in-depth study of the reactive ion plating process for ZnO and demonstrate that ion p l a t i n g does indeed increase the oxygen content of the deposited f i l m , as assumed by other workers. The stoichiometry of bias sputtered films, was measured and a series of experiments were performed to identify three processes that enhance f i l m oxidation; p r e f e r e n t i a l resputtering of Zn, p r e f e r e n t i a l evaporation of Zn and 50 activation or ion plating of oxygen species. 4.3 FILM STOICHIOMETRY Three sets of ZnO films of thicknesses 0.29 to 0.37 urn intended for Rutherford Backscattering (RBS) analysis of stoichiometry were deposi-ted in the metallic mode of Figure 2.4 at various values of applied substrate b i a s . The f i r s t set of films were sputtered onto glass substrates and sent to C. Evans and Associates in San Mateo, C a l i f o r n i a for RBS analysis. Due to oxygen and other elements in the glass substrates, O/Zn stoichiometry ratios could not be deduced. However, a carbon impurity concentration of less than 1% was detected in the f i l m s . Furthermore, the f i l m deposited at -100 V bias was found to have less carbon impurity than the film deposited without a substrate discharge. The two other sets of films were deposited on spectrographic grade graphite stubs and then analyzed at McMaster University in Hamilton by D. Stevanovic. Figure 4.2 shows the backscattered y i e l d versus energy for a t y p i c a l RBS spectrum obtained from the films on graphite substrates. Backscattered He* ions from a 2.03 MeV He* beam at normal incidence were detected at a scattering angle of 1 6 0 ° . Zn nuclei present a larger cross section and scatter He* at higher energies than oxygen due, respectively, to a larger nuclear charge and mass. Helium ions scattered from the back side of the film are detected at lower energy than those scattered from the surface due to a gradual energy loss by 51 3 0 0 0 -2 0 0 0 -UJ >-1000 J ENERGY (MeV) Figure 4.2 A Rutherford Backscatter spectrum obtained from a ZnO f i l m deposited on a graphite substrate. The backscattered yi e l d s of He* ions from zinc and oxygen atoms in the f i l m and from carbon atoms in the substrate are indicated. 52 electronic interactions as the ion traverses the f i l m . This energy loss creates rectangular shaped y i e l d spectra. Since the scattering cross sections are small and multiple scattering effects are n e g l i g i b l e , the horizontal plateaus of the y i e l d spectra indicate a composition that i s uniform with depth. The r a t i o of oxygen to zinc in the f i l m i s simply obtained from the r a t i o of areas under the spectra multiplied by a cross section correction factor [81]. The detection l i m i t for impurities i s generally of the order of 1% or l e s s . No impurities were detected within these l i m i t s , however, implanted Ar would be p a r t i c u l a r l y d i f f i c u l t to detect as i t s y i e l d spectrum partly overlaps that of z i n c . Figure 4.3 gives the oxygen to zinc r a t i o for each film and the corresponding r e s i s t i v i t y for each f i l m . The data points indicated by squares are from one of the sets of films on graphite substrates. They have smaller composition error estimates than the other data indicated by c i r c l e s due to less t a i l i n g in the experimental spectra and a lower background y i e l d . The films at -7 V bias (no rf power) are v i s u a l l y dark brown, the f i l m at -40 V bias i s a l i g h t e r brown and the remaining films show increasing c l a r i t y with increasing b i a s . This confirms previous assumptions that the colour change was due to an increase in the O/Zn r a t i o [11]. Notice that the best conducting films are transparent and nearly stoichiometric with a s l i g h t zinc excess, compatible with the suggested mechanisms of ZnO conduction by oxygen vacancies or zinc i n t e r s t i t i a l s . In summary, the substrate bias i s shown 53 Figure 4 .3 The composition of bias sputtered zinc oxide as determined by Rutherford Backscatter a n a l y s i s . The r e s i s t i v i t y measurements correspond to those films whose compositions are indicated by c i r c l e s . 54 to have a s i g n i f i c a n t e f fect on the composition of reactively ion plated ZnO f i l m s , increasing the oxygen content from Zn0.7 6 to nearly stoichiometric ZnO. 4.4 MATERIAL RE-EMISSION The effect of substrate bias on deposition rate has been studied for several d i f f e r e n t oxygen flow rates as shown in Figure 4.4. A l l films were deposited for 20 minutes in the metallic operating mode at a target power of 150 W and a t o t a l sputtering gas pressure of 1.0 Pa. The p a r t i a l pressure of oxygen in the v i c i n i t y of the substrate, pp(02), i s shown for each flow rate. The films deposited at f(O2)=0 seem are, of course, pure metallic z i n c . It was previously mentioned that the Zn deposition rate at constant target power slowly decreases over the course of several experiments due to the clogging of holes in the b a f f l e . These experiments were done in order of increasing f ( 02) and the decrease in deposition rate of Zn i s estimated not to exceed 15% during sputtering of the 24 films reported here. A substantial decrease in thickness of the Zn metal films was observed as the substrate bias (and consequently t o t a l rf power) was increased. At high b i a s , the deposition rate i s apparently near zero. Also with increasing b i a s , films at f(O2)=0.75 seem change in appearance from metallic black to almost c l e a r , films at f(O2)=1.80 seem change from black to c l e a r , and films at f(O2)=2.80 seem change from brown to c l e a r . At -7 V b i a s , t h i s i s indicative of Zn excess in the 55 pp(O2)=-046 Pa SUBSTRATE BIAS (V) Figure 4.4 Evidence of p r e f e r e n t i a l re-emission of excess zinc from growing f i l m s . A l l films were deposited at 150 W t o t a l power in the metallic operating mode for the various substrate bias and oxygen flow rates indicated. 56 films decreasing with increasing f ( 02) , whereas at high substrate bias a l l films (except Zn films) approach stoichiometric ZnO. The observed material loss i s confirmed by the report of Maniv [57] of an unexplained decrease in Zn deposition rate with increasing substrate discharge power during reactive planar magnetron sputtering of ZnO. Figure -3.2 showed that the reactive gas baffle prevented interactions between the rf and the dc plasmas, so that a change in conditions at the target due to the rf discharge i s not the cause of the deposition rate decrease. Igasaki and Mitsuhashi [78] reported a small deposition rate decrease with bias during reactive ion plating of TiN f i l m s . They were able to attribute t h i s decrease to the d e n s i f i c a t i o n of films deposited at high bias. Such an explanation seems unlikely for the films of Figure 4.4 due to the doubling or t r i p l i n g of film density that would be required. However, an experiment, described below, was conducted to test for d e n s i f i c a t i o n effects in bias sputtered Zn f i l m s . Zn films were sputtered onto glass substrates at various values of substrate bias. The glass substrates were c a r e f u l l y weighed with a Mettler H41 balance of resolution 10 ygm before deposition, after deposition, and after etching off the Zn coating in HC1. Also after deposition, the f i l m thicknesses were measured with a profilometer. Table 4.1 shows that the film thickness has a strong correlation with the f i l m mass. The densities of a l l f i l m s , assuming homogeneity over 1x2 inch 57 Table 4.1 Mass measurement of bias sputtered Zn f i l m s . Bias (V) Thickness (<xm) Mass (mg) Density (gm/cm3) -6 -50 -60 -70 0.39 0.29 0.16 0.47±.02 3.21±.0 6 5.3±0.3 2.60 5.2 1.93 5.2 1.46 7.0±1.2 substrates, compare reasonably well with the tabulated bulk figure of 7.1 gm/cm3. The s l i g h t d e n s i f i c a t i o n of the film at -70 V bias may in fact be due to less e f f i c i e n t material removal near the edges of the substrate. The deposition rate decrease with bias i s attributed not to a change of target discharge conditions or film density but to re-emission of material from the growing f i l m . Zn ri c h films (low f ( 02) or low bias) show a large material loss with increasing bias whereas stoichiometric ZnO (high f ( 02) or high bias) i s resistant to material l o s s . This phenomenon is interpreted as p r e f e r e n t i a l emission of excess z i n c . 4.5 ZINC RE-EMISSION BY EVAPORATION Re-emission of zinc from a growing f i l m could occur by two processes; resputtering or evaporation. In this section experiments are described in which the heating action of the substrate bias i s measured, and then duplicated by a radiant heater in order to observe any evaporation effects independent of resputtering processes. Substrate temperature measurements were made as close to 58 the surface of the growing f i l m as possible. A thin Zn film (50 nm) of resistance 10 0 was deposited in a four point probe pattern on a glass substrate. Contacts to 0.03 mm diameter wires were made with s i l v e r paint. The Zn fi l m and contacts r were coated with 100 nm of e l e c t r i c a l l y insulating A l203 and then c a l i b r a t e d for temperature c o e f f i c i e n t of resistance in an oven. Four point probe measurements of Zn r e s i s t i v i t y during ZnO fi l m deposition on top of the Zn/Al203 structure enabled determination of the substrate surface temperature. Figure 4.5 shows both the t o t a l rf power applied to the substrate and the film surface temperature as a function of the substrate self b i a s . A maximum fi l m temperature near 400 K i s observed at the high bias value of -120 V. These low temperatures are in general agreement with the results of Sundgren et a l . [82] who reported a temperature of 420 K for uncooled substrates during reactive bias sputtering of TiC. Maniv et a l . [57] report uncooled substrate temperatures not exceeding 373 K during reactive bias sputtering of ZnO. The self bias voltage developed on Maniv's substrate holder (-10 V) was less than the bias voltages used in my experiments. The four point probe temperature results agreed to within 15 K with subsequent measurements made by attaching a 0.03 mm chromel-alumel thermocouple wire to the surface with s i l v e r paint. Successful deposition of ZnO coatings onto polyethylene p l a s t i c f i l m , which softens s i g n i f i c a n t l y at 370 K, provided further proof of the low substrate 59 SUBSTRATE BIAS (V) Figure 4 . 5 The substrate temperature and average rf power density at the substrate shown as a function of the applied substrate bias. 60 temperatures attained. An estimate of substrate heating was found by summing heat flux contributions to the fi l m and then assuming that cooling occurs only by a radiative process. There exist three major contributions to heating: 1. Atoms of sputtered Zn condensing on the substrate are expected to provide approximately 15eV/atom [83]. This corresponds to a power density of 100 W/m2. 2. A t y p i c a l rf power density, as seen in Figure 4.5, i s 200 W/m2. 3. The temperature of the Zn target surface was estimated at 403 K, using data from indium targets where the In surface melted (429 K) at s l i g h t l y higher cathode powers than used for Zn. The radiative heat flux from the hot Zn target provides about 60 W/m2 at the substrate. Thus a generous estimate of t o t a l power density at the substrate for a -100 V bias i s J=360 W/m2. For radiative cooling, the f i l m surface w i l l e q u i l i b r a t e at a temperature T given by; J = ea(T4-T0*) (4.6) where e i s the average f i l m e m i s s i v i t y , a i s the Stefan-Boltzmann constant, and T0 i s the temperature of the chamber walls. The chamber walls are cooled and heat only s l i g h t l y to about T0=315 K. For an emissivity e=0.5, a fi l m temperature T=387 K i s reached. This rough estimate of f i l m surface temperature compares well with the experimental data. At f i r s t consideration, these temperatures seem unlikely 61 to cause evaporation loss from the deposited f i l m . A radiant heater behind the substrate was c a l i b r a t e d using the thermocouple and four point probe thermometers, and used to test for evaporation e f f e c t s . ZnO f i l m s were deposited in the m e t a l l i c mode (f(0 2 )=2.5 seem) with no substrate rf discharge but at temperatures corresponding to those induced by the b i a s . Figure 4.6 shows these r e s u l t s ; f i l m thicknesses p lo t ted against temperature and the e f f e c t i v e bias that would be required to produce such temperatures. Films deposited at low temperatures were brown in colour and became clear at high temperatures, i n d i c a t i n g a t r a n s i t i o n from a Zn excess material to s toichiometr ic ZnO and a p r e f e r e n t i a l loss of Zn s i m i l a r to that of Figure 4.4. Also shown in Figure 4.6 are data points from f i lms made in the target oxide coverage mode with subsequent high p a r t i a l pressure of oxygen near the substrate . These f i lms are transparent and showed n e g l i g i b l e material l o s s . Table 4.2 Comparison of re-emission rate from ZnO fi lms with bulk Zn evaporation r a t e . Temperature ZnO Bulk Zn (K) Re-emission Evaporation (nm/min) (nm/min) 340 5 360 12 380 18 0.0037 400 0.027 440 0.80 480 1 4 Consult ing known evaporation rates for bulk zinc [84], 4 0 0 £ c to to 3 0 0 y 2 0 0 2 LI 1 0 0 BIAS E Q U I V A L E N T (V) - 4 0 - 9 0 -120 1 1 T ™ OXIDE M O D E -) METALL IC -\ M O D E 1 i I l I 310 3 3 0 3 5 0 3 7 0 3 9 0 SUBSTRATE T E M P E R A T U R E (K) F i g u r e 4 . 6 Evidence of m a t e r i a l re-emission from z i n c oxide f i l m s d e p o s i t e d i n the m e t a l l i c and oxide t a r g e t modes. No s u b s t r a t e b i a s was a p p l i e d , however the s u b s t r a t e was heate to temperatures e q u i v a l e n t to those induced by a s u b s t r a t e d i s c h a r g e . 63 Table 4.2 was constructed to compare the predicted bulk Zn evaporation loss with the experimental loss of material from ZnO f i l m s . The predicted bulk evaporation rate i s seen to be more than 2 orders of magnitude less than that observed experimentally. These data are explained by proposing that adatoms of Zn are evaporated from the growing f i l m before oxidation, and at an enhanced rate due to the weak bonding nature of such surface adatoms. Zinc atoms that are oxidized and become part of the bulk f i l m w i l l be stable and much less prone to evaporation. Consequently, a Zn excess material w i l l show more Zn evaporation than stoichiometric ZnO. S i m i l a r l y , under f i l m growth conditions where zinc oxidation i s rapid (high pp(02), oxide mode in Figure 4.6) there w i l l be much less Zn loss than under growth conditions where oxidation of Zn adatoms occurs much more slowly (low pp(02), metallic mode of Figure 4.6). .Notice that the loss of material of the heated ZnO f i l m (Figure 4.6, metallic mode) i s somewhat larger than that of the bias equivalent ZnO at about f(02)=2.5 seem in Figure 4.4. This difference in material re-emission rate is associated with an oxygen ion p l a t i n g and a c t i v a t i o n effect that w i l l be discussed in section 4.7. 64 4.6 ZINC RE-EMISSION BY SPUTTERING To independently i d e n t i f y any e f f e c t s of r e s p u t t e r i n g , s u b s t r a t e s were cooled to e l i m i n a t e evaporation p r o c e s s e s . C o o l i n g was accomplished by bonding the g l a s s s u b s t r a t e with s i l v e r epoxy to a water coo l e d copper p l a t e . Subsequent temperature measurements i n the cooled mode showed that the su b s t r a t e s u r f a c e d i d not exceed 313 K at the r f power l e v e l s used. F i g u r e 4.7 shows the t h i c k n e s s dependence on s u b s t r a t e b i a s f o r pure Zn f i l m s s p u t t e r e d on both c o o l e d and uncooled s u b s t r a t e s at a t a r g e t power of 150 W f o r d e p o s i t i o n times of 30 minutes. Substrate c o o l i n g i s seen to reduce the Zn re-emission l o s s , p r o v i d i n g f u r t h e r evidence of the evapo r a t i o n p r o c e s s . F i g u r e 4.8 shows the r e s u l t s of an i d e n t i c a l experiment on ZnO f i l m s produced i n the m e t a l l i c mode at t a r g e t power 150 W with f(02)=2.35 seem. ZnO f i l m s at low s u b s t r a t e b i a s v o l t a g e s were black whereas those at high b i a s were c l e a r , i n d i c a t i v e of a r e l a t i v e i n c r e a s e i n the oxygen content with b i a s . I a s c r i b e the m a t e r i a l l o s s on the co o l e d f i l m s to p r e f e r e n t i a l r e s p u t t e r i n g of z i n c and the added m a t e r i a l l o s s on the uncooled f i l m s t o e v a p o r a t i o n . A f u r t h e r experiment was performed to compare these observed r e s p u t t e r r a t e s with s p u t t e r r a t e s expected from bulk z i n c and bulk ZnO. A s e r i e s of four black ZnO f i l m s were sp u t t e r e d under i d e n t i c a l c o n d i t i o n s with no s u b s t r a t e b i a s a p p l i e d . A f t e r d e p o s i t i o n of a l l f i l m s , one f i l m (#3) was exposed to a s u b s t r a t e d i s c h a r g e of -175 V b i a s f o r 120 65 Figure 4 . 7 Evidence of re-emission from bias sputtered Zn f i l m s . Film thicknesses are shown for films deposited on both cooled and uncooled substrates. 66 Figure 4.8 * Evidence of re-emission from ZnO films bias sputtered in the metallic target mode. Film thicknesses are shown for films deposited on both cooled and uncooled substrates. 67 Table 4.3 Resputter of ZnO films after deposition. ZnO Film Resputter After Thickness Deposition (urn) no no -175 V no 0.27±.02 2 3 4 0.26 0.17 0.27 minutes. Table 4.3 shows the resu l t i n g thicknesses of each f i l m . From the material loss of f i l m #3, a resputter rate was determined. A similar experiment was performed where a pure Zn fil m was resputtered at -175 V bias after deposition. From these data, an average sputter y i e l d for bulk Zn and bulk ZnO was calculated using the rf plasma c h a r a c t e r i s t i c s of section 4.1. Due to the d i s t r i b u t i o n of bombarding ion energies and the dramatic decrease of sputter y i e l d for energies below 50 eV, t h i s calculated y i e l d is a very conservative estimate of the true y i e l d at f u l l bias energy. The resputter r e s u l t s are summarized in Table 4.4. The sputter y i e l d s for Zn and ZnO during deposition were calculated from Figures 4.7 and 4.8. I was unable to find tabulated data for the low energy sputter y i e l d of Zn or ZnO, however a value of 0.8 at 175 V i s t y p i c a l for a high y i e l d metal such as Zn [85]. The observed resputter y i e l d during deposition i s much larger than that measured for bulk ZnO or bulk Zn afte r deposition and also larger than the 'tabulated value'. Consistent with the explanation of the evaporation process, i t i s proposed that the growing f i l m displays an 68 Table 4.4 Resputter y i e l d s of Zn and ZnO f i l m s • Fi lm Conditions Bias (V) Resputter Rate (nm/min) Y i e l d Zn ZnO resputter af ter deposit ion -175 -175 2 0.8 0.3 0.1 Zn ZnO resputter during deposi t ion -1 60 -160 17 20 3.0 2.6 enhanced sputter y i e l d for unoxidized zinc adatoms. 4.7 OXYGEN ACTIVATION AND PLATING The rf discharge surrounding the substrate induces bombardment by energetic ions of A r + , 0 2 + , and 0" (ion pla t ing) and w i l l create exci ted and atomic oxygen species through c o l l i s i o n processes (oxygen a c t i v a t i o n ) . Teer [86] has shown that a large amount of the energy in "a glow discharge is d i s s i p a t e d by energetic n e u t r a l s . To determine i f p l a t i n g and a c t i v a t i o n are s i g n i f i c a n t in stoichiometry c o n t r o l , I have chosen a substrate geometry that w i l l minimize the processes of resputtering and evaporation. An rf discharge was appl ied to a f ine mesh screen placed 1.5 cm in front of the e l e c t r i c a l l y i s o l a t e d glass substrate . In t h i s mode, energetic bombardment of the substrate was reduced and most heat d i s s i p a t e d at the screen. However, due to the proximity of the substrate to the screen discharge, the substrate would s t i l l be subjected to lower energy neutral and ion bombardment, low energy e lec t ron bombardment, and exposed 69 to a plethora of activated species. This geometry i s similar to that of Morgan et a l . [14], who evaporated Zn through an oxygen gas discharge in order to enhance the growth of ZnO at the substrate. A Zn fil m sputtered in pure Ar at various values of screen self bias provided a sensitive test of the magnitude of resputtering and evaporation e f f e c t s . Figure 4.9 shows that no measurable Zn re-emission occurred. However, reactive sputtering of ZnO in the metallic mode with the screen discharge s t i l l exhibited a change in f i l m appearance from dark brown to very l i g h t brown and a decrease in f i l m r e s i s t i v i t y as the screen bias was increased (Figure 4.9). No ZnO material loss was observed. Notice that the lowest r e s i s t i v i t y achieved i s somewhat higher than that attained with the rf discharge at the substrate (Figure 4.3). I conclude that oxygen act i v a t i o n and/or ion platin g are s i g n i f i c a n t in c o n t r o l l i n g ZnO f i l m stoichiometry by enhancing the oxidation of sputtered z i n c . At present, the r e l a t i v e importance of p l a t i n g and activation processes i s d i f f i c u l t to d i s t i n g u i s h . The stoichiometry change inferred from the re s u l t s of Figure 4.9 (where ion p l a t i n g of the substrate i s reduced) indicates that plasma ac t i v a t i o n processes such as energetic neutral bombardment a s s i s t f i l m oxidation. The ion plating contribution to oxidation may be evaluated more p r e c i s e l y . The oxygen ion current to the substrate was estimated from the rf plasma c h a r a c t e r i s t i c s discussed in section 4.1. The t o t a l ion current in the discharge at -100 V 70 Figure 4 .9 Results from application of an rf discharge to a screen in front of the substrate. Film thicknesses for Zn films and r e s i s t i v i t y for ZnO films deposited in the metallic target mode are shown as a function of screen bias. 71 bias was found to be J = 3X101b ions/cm2s. If the number of oxygen ions created is proportional to the oxygen p a r t i a l pressure, then J(02) = 1.5xl01 3 ions/cm2s. Hecq. et a l . [87] have shown that oxygen ionizes p r e f e r e n t i a l l y in high pressure Ar/02 discharges, but the effect i s not s i g n i f i c a n t at the low pressures used in this--investigation. The r a t i o of the rate of a r r i v a l of Zn atoms to oxygen molecules (due to pp(O2)=0.05 Pa) and to oxygen ions (due to the ion current) is estimated, as 1:65:0.01 respectively. This r a t i o demonstrates two important points. F i r s t , the s t i c k i n g c o e f f i c i e n t of molecular oxygen on zinc i s not large (about 0.015). Second, for the ion plating mechanism to be s i g n i f i c a n t , the sticking c o e f f i c i e n t of bombarding oxygen ions should be much larger than that of free oxygen molecules. This i s l i k e l y the case, since near energies 50 eV or greater, ions w i l l be implanted into the growing f i l m [88]. Assuming a unity sticking c o e f f i c i e n t for oxygen ions, the resultant increase in oxygen content of the f i l m would be about 2%. Additional evidence for the a c t i v a t i o n / p l a t i n g mechanisms was provided in Figure 4.6, where heated ZnO showed a sur p r i s i n g l y large material loss compared to that expected from a substrate discharge. The heated ZnO was lacking a c t i v a t i o n / p l a t i n g processes during growth that would have enhanced the formation of stable oxide and resisted further material l o s s . 72 4.8 SUMMARY AND DISCUSSION Resputter processes at the substrate have been observed and modelled by a number of researchers, most recently by Barnett and Greene for GaAs deposition [89], and by Dove et a l . for metal a l l o y films [76]. None of these models required enhancement factors due to unexpectedly large resputter yields and, evaporation mechanisms were not e x p l i c i t l y considered. In other work, resputter processes have been attributed to substrate bombardment by neutrals and negative ions produced at the target [90,91,92]. Such bombardment i s unlikely in my experiments since the target to substrate separation of about 10 mean free path lengths ensured that energetic species from the target would thermalize before reaching the substrate. Also, the observed resputter v a r i a t i o n with bias i s opposite to that expected for negative ion bombardment. In reactive bias sputtering by Poitevin et a l . [79] and by Igasaki and Mitsuhashi [78], no change in stoichiometry of TiN films resulted from variations in the substrate b i a s , and no material re-emission was observed. Reactive bias sputtering of TiC by Sundgren et a l . [82] showed an increase in carbon content of the films as the bias was varied from -200 V to 0 V. This was attributed to ion platin g of CH^ molecules. Murti has rf sputtered transparent ZnO films from a zinc oxide target [138]. With increasing substrate bias, he observed a change in f i l m r e s i s t i v i t y from 10" 8cm to 10"1 Ocm and attributed this effect to p r e f e r e n t i a l resputtering of oxygen from the f i l m , in apparent contradiction to my r e s u l t s . 73 However, in Murti's work, the bias voltage on the substrate holder was obtained by means of a voltage divider network from the same power supply used to supply the target voltage. Consequently, as the bias voltage was increased, the voltage at the target and the f i l m deposition rate decreased. Such a change in f i l m deposition rate confuses the interpretation of results by a l t e r i n g the a r r i v a l rate of Zn and 0 species at the substrate. Furthermore, extensive studies of reduction at oxide surfaces by Kim et a l . [139] have shown that ZnO i s not reduced by Ar + ion bombardment. The results of Jones et a l . [93] are most relevant to my work on ZnO. During deposition of Si02 by sputtering, high material re-emission c o e f f i c i e n t s of up to 0.85 were observed. Substrate temperatures (623 K) and rf power density (4 W/m2) were s i g n i f i c a n t l y higher than those used here for ZnO. Re-emission c o e f f i c i e n t s were found to be an increasing function of substrate temperature and bias and were explained by a model of evaporation and resputtering of loosely bound surface atoms. I have adopted th i s model to explain the observed loss of Zn at low temperatures and bombardment energies. A notable s i m i l a r i t y of Si and Zn i s that both materials oxidize very slowly, as judged by formation times of oxide monolayers on clean metal [94,95]. Thus adatoms may remain unattached and mobile on a growing film for longer periods of time, leading to a greater pr o b a b i l i t y of emission. Other pertinent c h a r a c t e r i s t i c s of Zn are i t s low melting point (693 K) and very high vapour pressure (10~8 torr at 74 400 K) due in part to the weak Zn-Zn bond strength of 0.3 eV. This is the lowest bond strength of any metal except Hg. For comparison, the vapour pressure of t i n (melting point 505 K and bond strength 2 eV) i s 12 orders of magnitude lower than Zn [96]. Weakly bonded surface monomers or dimers of Zn should exhibit even larger vapour pressures and e f f e c t i v e l y higher emission p r o b a b i l i t i e s . However, when oxidized, Zn i s strongly bound to oxygen (2.8 eV bond strength), has a ne g l i g i b l e vapour pressure and a higher bulk sputtering y i e l d . In further work by Jones et a l . [97] i t was observed that re-emission was essential to obtain films of best q u a l i t y , as judged by the pin-hole breakup thickness phenomenon, which i s a test of resistance to erosion by a c i d . This increase in f i l m quality was suggested by Jones to be due to re-emission of atoms in non-optimum positions that would otherwise be trapped by further deposits to form defected material. Analagous to t h i s r e s u l t , I note that the best ZnO f i l m s , as judged by lowest r e s i s t i v i t y , were not made by stoichiometry control with only the oxygen pla t i n g / a c t i v a t i o n mechanism (Figure 4.9) but by a substrate discharge that also caused Zn material l o s s . Re-emission e f f e c t s may optimize e l e c t r i c a l properties by removing impurities such as carbon (as observed in RBS measurements) or implanted Ar. Winters and Kay [98] have observed that the concentration of implanted Ar i s minimized for substrate biases near -90 V. If the anomalous properties of Zn such as bond strength and vapour pressure are the fundamental reasons for the 75 observed enhanced re-emission phenomena, then a metal with large bond strengths, low vapour pressure and a known a f f i n i t y for oxygen should not exhibit re-emission during reactive bias sputtering. Titanium was chosen as a representative metal of this c l a s s . T i has a bond strength of 1.5 eV, a vapour pressure more than 14 orders of magnitude lower than Zn, and is used extensively as a getter pump for oxygen [99']. Table 4.5 Thicknesses of bias sputtered T i 02 f i l m s . Bias Thickness (V) (Mm) r7 0.391.03 -66 0.42 -100 0.43 -125 0.40 -150 0.39 -175 0.39 A Ti metal target was react i v e l y sputtered in an Ar/02 discharge, and films with a s l i g h t oxygen deficiency (sub-stoichiometric T i 02) were deposited in the metallic target mode at various values of substrate b i a s . Table 4.5 shows the result i n g thicknesses of films deposited at 400 W target power. No material loss was observed for a bias up to -175 V, which exceeds the substrate bias values normally used to produce ZnO f i l m s . The T i 02 films were v i s u a l l y dark blue, indicative of a Ti excess [140]. A decrease of t h i s visual absorption with increasing bias was attributed to a reduction of the Ti excess through ion plating of oxygen. It has been pointed out by Koenig and Maissel [66] that 76 resputtering of a film during deposition i s dependent on the d i v i s i o n of applied rf voltage between the darkspaces of the substrate and the ground (chamber) electrode. As pointed out in section 4.1, t h i s d i v i s i o n is dependent on the r e l a t i v e areas of the electrodes. I have shown that substrate bias and temperature not exceeding -100 V and 370 K, respectively, allow oxidation control and cause some b e n e f i c i a l material removal but do not reduce deposition rates to i n e f f i c i e n t l e v e l s . Consideration of these effects becomes c r i t i c a l l y important when designing sputter systems for reactive bias sputter deposition of ZnO. It may be necessary to t a i l o r substrate geometry and cooling in order to obtain the desired bias, temperature and current density c h a r a c t e r i s t i c s , subsequently optimizing resputtering, evaporation and oxygen activation e f f e c t s . In summary, I have shown that the effect of the substrate discharge i s to enhance the oxygen content of the growing f i l m . This occurs by p r e f e r e n t i a l evaporation and resputtering of zinc and by oxygen ion platin g and a c t i v a t i o n . The resputtering and evaporation processes were found to be s i g n i f i c a n t l y enhanced over that expected for bulk Zn, due to the weak bonding nature of unoxidized Zn atoms. In l i g h t of th i s understanding of the deposition process and a knowledge of f i l m stoichiometry, the following chapter describes the f u l l characterization of o p t i c a l , e l e c t r i c a l and structural properties of bias sputtered ZnO f i l m s . 77 CHAPTER FIVE: ZnO FILM CHARACTERIZATION 5.1 FILM DEGRADATION A t y p i c a l experiment to test for the ageing or weatherability of a heat mirror f i l m is to expose the fil m to high temperatures (370 K) and humidity while i r r a d i a t i n g the film with intense u l t r a v i o l e t radiation [100], Such tests are believed to simulate normal use degradation properties, but at an accelerated rate. To give some indication of the s t a b i l i t y of sputtered ZnO fil m properties, a very simple experiment was designed as follows. A small oven was maintained at a temperature of 365 K and approximately 90% r e l a t i v e humidity. Two i d e n t i c a l ZnO samples of good conductivity were deposited on glass substrates. One f i l m was kept as a control at room temperature and humidity, whereas the other was kept in the oven for a period of 40 days. Measurements of fil m r e s i s t i v i t y and infrared r e f l e c t i v i t y were made at periodic inter v a l s for both samples. No measurable change in properties of the control f i l m was observed after the 40 day i n t e r v a l . However, Figures 5.1 and 5.2 show s i g n i f i c a n t changes in r e s i s t i v i t y and infrared r e f l e c t i v i t y of the heated f i l m . The increase in fil m r e s i s t i v i t y corresponded to a decrease in the infrared r e f l e c t i v i t y . Also, the heated film was less transparent due to a cloudy or milky appearance and, in some regions of the film small flakes of ZnO had separated from the substrate and could be blown away. 78 1 0 2 0 3 0 TIME (DAYS) Figure 5.1 Degradation of r e s i s t i v i t y for a ZnO f i l m maintained at 95°C and 90% r e l a t i v e humidity. 79 4 8 12 16 WAVELENGTH (jjm) Figure 5.2 Infrared r e f l e c t i o n c h a r a c t e r i s i t i c s of a ZnO f i l m before and after a 40 day exposure to a temperature of 95°C and a r e l a t i v e humidity of 90%. 80 The degradation e f f e c t s i n the h o t , humid environment may be due t o f u r t h e r o x i d a t i o n of the ZnO f i l m ( i n c r e a s i n g the r e s i s t i v i t y ) and by the a t t a c k of water vapour ( p o s s i b l y s l i g h t l y a c i d i c ) at the g r a i n boundaries of the c o a t i n g . Maniv et a l . [57] and H i c k e r n e l l [101] have re p o r t e d very r a p i d e t c h r a t e s for ZnO i n weak a c i d . Work by Minami et a l . [102] compared the changes i n c o n d u c t i v i t y of ZnO and ITO t r a n s p a r e n t conductors due to high temperature exposure. ITO was r e s i s t a n t to change f o r temperatures up to 700 K, but ZnO f i l m s showed an i n c r e a s e i n r e s i s t i v i t y even a f t e r b r i e f (10 minute) exposure to temperatures exceeding 470 K. From my experiment i t i s c e r t a i n l y not p o s s i b l e to p r e d i c t the l i f e t i m e of a ZnO heat m i r r o r or transparent e l e c t r o d e i n normal use. However, conducting ZnO does not appear p r o m i s i n g as a s t a b l e c o a t i n g f o r long term use. F i l m d e g r a d a t i o n may be i n e v i t a b l e as the conduction p r o p e r t i e s r e l y on non-stoichiometry through oxygen d e f i c i e n c y . In comparison, good q u a l i t y s t o i c h i o m e t r i c ZnO f i l m s f o r a c o u s t i c t r a n s d u c e r a p p l i c a t i o n s are known to e x h i b i t very s t a b l e f i l m p r o p e r t i e s [101]. 5.2 FILM STRUCTURE The m i c r o s t r u c t u r e of b i a s s p u t t e r e d ZnO "films i s m o d i f i e d by the e f f e c t s of both e n e r g e t i c ion bombardment and by s t o i c h i o m e t r y change. To gain an independent understanding of each e f f e c t , X-ray d i f f r a c t i o n measurements were made on b i a s s p u t t e r e d z i n c f i l m s (no s t o i c h i o m e t r y change) and on ZnO 81 f i l m s s p u t t e r e d at v a r i o u s f ( 02) without s u b s t r a t e b i a s (no bombardment e f f e c t s ) . T h i s study has enabled a q u a l i t a t i v e a n a l y s i s of the b i a s induced s t r u c t u r e m o d i f i c a t i o n i n ZnO films." ZnO f i l m s were s p u t t e r e d in the m e t a l l i c t a r g e t mode on g l a s s s u b s t r a t e s at v a r i o u s v a l u e s of oxygen flow f ( 02) . F i g u r e 5.3 shows the hexagonal w u r t z i t e X-ray d i f f r a c t i o n spectrum f o r each f i l m , using 0.154 nm Cu-Ka r a d i a t i o n . S t r u c t u r e s i n the d i f f r a c t i o n s p e c t r a were not observed ou t s i d e of the given angle range, with the o c c a s i o n a l exception of a broad peak at 2 0 = 7 2 . 7 ° , c h a r a c t e r i s t i c of the ZnO 004 l i n e . Zn excess f i l m s at low f ( 02) d i s p l a y broad d i f f r a c t i o n peaks of low i n t e n s i t i e s , i n d i c a t i v e of a m i c r o c r y s t a l l i n e or near amorphous m a t e r i a l . No peaks due to m e t a l l i c Zn were i d e n t i f i e d , presumably due to the large d i s o r d e r i n the f i l m s and the gr e a t e r volume f r a c t i o n of ZnO. A s i m i l a r lack of metal d i f f r a c t i o n peaks in n o n - s t o i c h i o m e t r i c f i l m s was observed by A f f i n i t o et a l . [103] for A1/A1N composites at volume f r a c t i o n s of A l up to 0.45. At high f ( 02) , the n e a r l y s t o i c h i o m e t r i c ZnO f i l m s have a good c r y s t a l s t r u c t u r e with a s t r o n g l y p r e f e r r e d o r i e n t a t i o n of the c - a x i s p e r p e n d i c u l a r to the s u b s t r a t e . A s i g n i f i c a n t l a t t i c e s t r a i n of 0.8% i s seen, i n d i c a t e d by the s h i f t of the 002 peak to 3 4 . 2 0° from i t s u n s t r a i n e d value of 3 4 . 4 7 ° . Such s t r a i n and a p r e f e r r e d basal o r i e n t a t i o n are t y p i c a l l y observed i n sputtered ZnO c o a t i n g s [104,105]. Some of the common e f f e c t s n o t i c e d i n s p u t t e r e d elemental 82 J I I I I L 36 34 32 2xTHETA (DEGREES) Figure 5 .3 X-ray d i f f r a c t i o n spectra of zinc oxide films deposited at various values of oxygen flow r a t e . With increasing f ( 02) , the composition of the films approach stoichiometric ZnO. A substrate discharge was not used during deposition. 83 f i l m s as the s u b s t r a t e b i a s i s in c r e a s e d are f i l m d e n s i f i c a t i o n , a l a r g e decrease i n c r y s t a l l i t e s i z e and a s i g n i f i c a n t d i s r u p t i o n of the morphology and c r y s t a l o r i e n t a t i o n [106,107]. These e f f e c t s are b e l i e v e d to be due in part to a sp u t t e r r e d i s t r i b u t i o n of f i l m m a t e r i a l [108]. In F i g u r e 5.4, X-ray d i f f r a c t i o n s p e c t r a are shown for Zn f i l m s s p u t t e r e d i n an Ar discharge at 1 Pa t o t a l pressure and at v a r i o u s v a l u e s of s u b s t r a t e b i a s . Zn c r y s t a l l i z e s i n a hexagonal c l o s e s t packed form. F i l m s at high b i a s were sputtered f o r longer time p e r i o d s such that a l l f i l m t h i c k n e s s e s agree to w i t h i n 20%. Even'a low b i a s of -50 V i s seen to cause a dramatic decrease i n c r y s t a l l i n i t y of the f i l m and a change from a p r e f e r r e d b a s a l o r i e n t a t i o n of c r y s t a l l i t e s to a more random o r i e n t a t i o n . Two important e f f e c t s of s u b s t r a t e b i a s on f i l m m i c r o s t r u c t u r e have been observed i n the pre v i o u s two experiments. F i r s t , a r e d u c t i o n of Zn excess in the f i l m s leads to o r i e n t e d p o l y c r y t a l l i n e f i l m s showing l a r g e s t r a i n whereas e n e r g e t i c ion bombardment reduces and randomizes f i l m c r y s t a l l i n i t y . Both e f f e c t s must be i n c l u d e d to e x p l a i n the m i c r o s t r u c t u r e of b i a s s p u t t e r e d ZnO shown i n F i g u r e 5.5. These f i l m s were s p u t t e r e d from the standard m e t a l l i c t a r g e t mode at t o t a l pressure 1 Pa and at f(02)=2.5 seem. The o p t i c a l and e l e c t r i c a l c h a r a c t e r i s t i c s are s i m i l a r to those reported i n F i g u r e 2.6. Films at -6 V and -40 V b i a s are v i s u a l l y dark with a Zn excess. F i l m s at higher b i a s are transparent and good con d u c t o r s . The f i l m at -6 V b i a s (no r f 84 J 1 1 — i 1 i ' • ' • 4 4 4 0 3 6 2xTHETA (DEGREES) Figure 5.4 X-ray d i f f r a c t i o n spectra of bias sputtered Zn f i l m s . Deposition times were chosen such that a l l films were of thickness 0.65 ym. 85 ZnO FILMS 002 —1 I 1 l l l i 38 36 34 32 2 * THETA (DEGREES) Figure 5.5 X-ray d i f f r a c t i o n spectra of bias sputtered zinc oxide f i l m s . E l e c t r i c a l and o p t i c a l characterization of these films are detailed in Table 5.2 and Figure 5.11 respectively. 8 6 power) i s an un s t r a i n e d Zn excess m a t e r i a l s i m i l a r to those seen in F i g u r e 5.3. At -40 V b i a s , l i t t l e s t o i c h i o m e t r y change has o c c u r r e d , and bombardment induced randomization has degraded the o r i e n t a t i o n and c r y s t a l l i n i t y of the f i l m . For f i l m s at -65 V and -90 V, r a p i d improvements i n s t o i c h i o m e t r y (see RBS r e s u l t s of Figure 4.3) overwhelm bombardment induced d i s o r d e r and a d i f f r a c t i o n p a t t e r n i n d i c a t i v e of a w e l l o r i e n t e d , h i g h l y s t r a i n e d , near s t o i c h i o m e t r i c f i l m i s seen. For even higher b i a s , the s t o i c h i o m e t r y cannot improve f u r t h e r , and the e f f e c t s of ion bombardment are p r e v a l e n t a g a i n , producing a r e d u c t i o n i n c r y s t a l l i n i t y . Table 5.1 Gr a i n s i z e of b i a s s p u t t e r e d ZnO f i l m s . Sample B i a s Grain S i z e (V) (nm) ZnO-1 -6 36 ZnO-2 -40 23 ZnO-3 -65 52 ZnO-4 -90 42 ZnO-5 -115 43 ZnO-6 -140 39 Th i s q u a l i t a t i v e e x p l a n a t i o n of ZnO d i f f r a c t i o n s p e c t r a assumes that inhomogeneous s t r a i n broadening of peaks i s not a major e f f e c t . Some proof f o r t h i s assumption i s presented l a t e r . With t h i s assumption, an estimate of g r a i n s i z e can be made from the Scherrer r e l a t i o n f o r g r a i n s i z e broadening of d i f f r a c t i o n peaks [109]; d = O.9X/Bcos0 (5.1) where X=0.154 nm i s the wavelength of Cu-Ka r a d i a t i o n , d i s 87 the g r a i n diameter, B i s the f u l l width at h a l f maximum and 8 i s the Bragg a n g l e . The r e s u l t s of such a c a l c u l a t i o n using the 002 peak are shown in Table 5.1. The g r a i n s i z e does not change d r a m a t i c a l l y as a f u n c t i o n of s u b s t r a t e b i a s . The best conducting f i l m s (-90V, -115V) have a g r a i n s i z e of about 40 nm. An annealing experiment was ab l e to provide some i n s i g h t concerning the extent of inhomogeneous s t r a i n broadening. A good q u a l i t y ZnO conductor was annealed i n a i r at 720 K f o r 3 hours. F i g u r e 5.6 shows the 002 d i f f r a c t i o n peak and f i l m r e s i s t i v i t y before and a f t e r a n n e a l i n g . The l a r g e i n c r e a s e i n r e s i s t i v i t y i s b e l i e v e d to be due to the e l i m i n a t i o n of oxygen vacancy or Zn i n t e r s t i t i a l donor l e v e l s as s t o i c h i o m e t r i c composition i s a t t a i n e d through oxygen d i f f u s i o n . S i g n i f i c a n t s t r a i n r e l i e f has occurred i n the annealed f i l m , i n d i c a t e d by a s h i f t i n peak p o s i t i o n c l o s e r to 3 4 . 4 7 ° . Gawlak and A i t a [105] have observed s i m i l a r s t r a i n r e l i e f i n ZnO annealed at 600 K. S u r p r i s i n g l y , i n s p i t e of s t r a i n r e l i e f , my f i l m s show very l i t t l e change in i n t e n s i t y or peak h a l f - w i d t h at f u l l maximum. It i s u n l i k e l y that homogeneous s t r a i n r e l i e f would occur without some inhomogeneous s t r a i n r e l i e f ( i n d i c a t e d by a re d u c t i o n of peak w i d t h ) , t h e r e f o r e I conclude that the ZnO f i l m s do not possess s i g n i f i c a n t inhomogeneous s t r a i n . A l s o , from a symmetry argument, i t seems u n l i k e l y that a f i l m composed of w e l l a l i g n e d p a r a l l e l c r y s t a l l i t e s should e x h i b i t s t r a i n of d i f f e r e n t magnitudes i n d i f f e r e n t c r y s t a l l i t e s . F i l m s f o r scanning e l e c t r o n microscope (SEM) 88 34.5° i if. 34.5° l » ' II Not I lAnnealed I lAnnealed j 1600 ncm I \5xl0"3ricm f~--IFWHM / l.25±.03° h~IFWHM / 1.251.03° 36 35 34 33 36 35 34 33 2* THETA (DEGRFFS) F i g u r e 5.6 The 002 X-ray d i f f r a c t i o n peak of a ZnO film before and after annealing at 720 K for 3 hours. Figure 5.7 An SEM photomicrograph of a transparent, conducting, bias sputtered ZnO f i l m . The f i l m i s viewed in cross section at an angle of 45° from the substrate normal. F i g u r e 5.8 SEM photomicrographs comparing the s u r f a c e roughness of ZnO f i l m s d e p o s i t e d with and without a s u b s t r a t e d i s c h a r g e . 91 i n v e s t i g a t i o n were spu t t e r e d on g l a s s s u b s t r a t e s . In order to view the f i l m s i n c r o s s s e c t i o n the s u b s t r a t e s were c l e a v e d . E t c h i n g techniques were not used to d e f i n e g r a i n boundaries. F i g u r e 5.7 shows an SEM photograph of a ZnO f i l m deposited at -100 V b i a s , viewed at an angle of about 4 5° from the s u b s t r a t e normal. As i s common for vacuum deposited c o a t i n g s at low temperatures, a roughly columnar m i c r o s t r u c t u r e i s v i s i b l e . One of the columns has been d i s p l a c e d by the c l e a v i n g process to a p o s i t i o n p a r a l l e l to the s u b s t r a t e s u r f a c e . The column diameter of about 200 nm i s compatible with (not smaller than) the g r a i n s i z e of 40 nm estimated from X-ray d i f f r a c t i o n s t u d i e s . F i g u r e 5.8 compares c r o s s - s e c t i o n a l SEM photographs of ZnO f i l m s d e p o s i t e d with and without a s u b s t r a t e b i a s . The b i a s s p u t t e r e d f i l m e x h i b i t s a smoother s u r f a c e . Such b i a s induced smoothing i s b e l i e v e d to be due to a r e d i s t r i b u t i o n of s p u t t e r e d m a t e r i a l from evaporation and r e s p u t t e r i n g e f f e c t s , as observed by Thornton f o r s p u t t e r d e p o s i t e d metal f i l m s [ 108]. 5.3 ELECTRICAL PROPERTIES To enable e l e c t r i c a l t r a n s p o r t s t u d i e s , f i l m s were d e p o s i t e d in a H a l l probe geometry by masking as shown in F i g u r e 5.9. F i l m s on quartz s u b s t r a t e s were de p o s i t e d simultaneously f o r the study of o p t i c a l p r o p e r t i e s , a l l o w i n g a c o r r e l a t i o n of t r a n s p o r t and o p t i c a l p r o p e r t i e s that w i l l be addressed i n s e c t i o n 5.4. A f t e r p r o d u c t i o n of ZnO f i l m s in 92 Figure 5.9 ZnO samples for characterization by Hall probe techniques were deposited in thi s geometry by u t i l i z i n g a mask during deposition. Table 5.2 E l e c t r i c a l properties of bias sputtered ZnO f i l m s . Sample Substrate Bias (V) R e s i s t i v i t y (10"3ficm) Hall Constant (cm3/C) Mobility (cm2/Vs) Carrier Density (1 0 2 O c i r r 3 ) ZnO-1 -6 20 -.019 • 1 .0 3.3 ZnO-2 -40 24 -.031 1.3 2.0 ZnO-3 -65 4.5 -.055 12 1 .1 ZnO-4 -90 1 .9 -.034 18 1 .8 ZnO-5 -1 15 2.1 -.040 1 9 1 .6 ZnO-6 -140 2.4 -.037 15 1 .7 Table 5.3 E l e c t r i c a l properties of ZnO films produced by other groups. Deposition R e s i s t i v i t y Hall Mobility Carrier Reference ' Technique Constant Density (lO-3J2cm) (cm3/C) (cm2/Vs) ( l 02 Oc n r3) rf sputter reactive evap rf sputter rf sputter 7 13 0.7 1 12 2 -.07 45 0.8 14 7 8 1 . 1 15 1 20 3.0 16 9 3 the m e t a l l i c t a r g e t mode at v a r i o u s values of s u b s t r a t e b i a s , the H a l l c o e f f i c i e n t was measured at room temperature with the use of a 13 kG electromagnet. From the H a l l c o e f f i c i e n t , R, the c a r r i e r m o b i l i t y u and c a r r i e r d e n s i t y N were c a l c u l a t e d using the r e l a t i o n s ; a = R / P (5.2) N = 1/Re where p i s the f i l m r e s i s t i v i t y and e i s the e l e c t r o n i c charge. Table 5.2 summarizes the e l e c t r i c a l c h a r a c t e r i z a t i o n f o r b i a s s p u t t e r e d ZnO f i l m s . Experimental e r r o r s in the measurements are about ±20%. A l l f i l m s were found to have l a r g e c a r r i e r d e n s i t i e s near 2 X 1 02° cm"3 and a c a l c u l a t i o n of degeneracy [110,111] gave a Fermi l e v e l EF approximately 0.35 eV above the conduction band edge. Samples ZnO-1 and ZnO-2 are v i s u a l l y dark and have a Zn excess of 10 to 20 atomic p e r c e n t . A l l other samples are transparent and expected to have a n e a r l y s t o i c h i o m e t r i c c o m p o s i t i o n . The d i s t i n g u i s h i n g f e a t u r e between dark and c l e a r f i l m s i s the low m o b i l i t y of 1 cm2/Vs f o r the dark f i1ms. S e l e c t e d r e s u l t s of e l e c t r i c a l t r a n s p o r t p r o p e r t i e s at room temperature f o r high c o n d u c t i v i t y t r a n s p a r e n t ZnO de p o s i t e d by other workers are shown i n Table 5.3. I t i s seen that the c a r r i e r d e n s i t i e s and m o b i l i t i e s i n my f i l m s are s i m i l a r to those produced by r f sp u t t e r d e p o s i t i o n from a oxide t a r g e t . However, only C a p o r a l e t t i [112] has d i s c u s s e d s c a t t e r mechanisms f o r conduction e l e c t r o n s and no 94 i n v e s t i g a t o r s have s t u d i e d black Zn excess oxide f i l m s . C a p o r a l e t t i has proposed a g r a i n boundary l i m i t e d m o b i l i t y f o r ZnO f i l m s , but u n f o r t u n a t e l y had no data on f i l m s t r u c t u r e to i n d i c a t e very small g r a i n s i z e s . From the t r a n s p o r t data of Table 5.2, I estimated the mean f r e e path of conduction e l e c t r o n s i n transparent f i l m s to be about 2 nm, using a r e l a x a t i o n time given by 7=m*/u/e for e l e c t r o n e f f e c t i v e mass m*=0.24mo. Since the g r a i n boundary s e p a r a t i o n (determined from X-ray and SEM work of s e c t i o n 5.2) i s at l e a s t 40 nm and thus much l a r g e r than the mean f r e e p a t h , I conclude that g r a i n boundary s c a t t e r i n g i s not the dominant s c a t t e r i n g p r o c e s s . A l s o , the order of magnitude change in the m o b i l i t y of my f i l m s with i n c r e a s i n g b i a s does not c o r r e l a t e with the n e a r l y constant g r a i n s i z e shown in Table 5.1. It should be s t r e s s e d that the e l e c t r i c a l p r o p e r t i e s (Table 5.2) and X-ray d i f f r a c t i o n s p e c t r a ( F i g u r e 5.4) are data taken from the same set of f i l m s . M a n s f i e l d [113] has extended the standard work of Conwell and Weisskopf on i o n i z e d impurity s c a t t e r i n g [114] to degenerate semiconductors and has d e r i v e d the f o l l o w i n g e x p r e s s i o n f o r c o n d u c t i v i t y ; a = 3h3 e2n (5.3) 1 67r2e2m*2f (x) with f ( x ) = l n d + x ) - x 1+x and x = (h2e/e2m*) (3n/87r/3 where e i s the s t a t i c d i e l e c t r i c constant (8.2 f o r ZnO), n i s the d e n s i t y of i o n i z e d impurity c e n t r e s , and a l l other symbols 95 have t h e i r normal meanings. I f the e l e c t r o n d e n s i t y (from donor c e n t r e s ) in transparent conducting ZnO i s equated to the i o n i z e d impurity c o n c e n t r a t i o n , t h i s leads to a m o b i l i t y M=70 cm2/Vs. A d d i t i o n a l s c a t t e r i n g c o n t r i b u t i o n s may come from defect or n e u t r a l impurity c e n t r e s such as Ar or C i m p u r i t i e s or from a small excess of u n i o n i z e d Zn. Using e x p r e s s i o n s f o r n e u t r a l impurity s c a t t e r i n g developed by Erginsoy [114], a n e u t r a l Zn impurity d e n s i t y of 101 9 cm"3 c o u l d l i m i t c a r r i e r m o b i l i t y to about 30 cm2/Vs. Such a combination of i o n i z e d and n e u t r a l impurity s c a t t e r i n g has been used to e x p l a i n the t r a n s p o r t p r o p e r t i e s of t r a n s p a r e n t , conducting ITO [116]. U n t i l the p r e c i s e s t r u c t u r a l nature of the l a r g e Zn excess i n dark, low m o b i l i t y ZnO i s known, i t i s only p o s s i b l e to s p e c u l a t e on the i d e n t i t y of s c a t t e r c e n t r e s . S i m i l a r metal excess i n s p u t t e r e d l n203 and AlN f i l m s [127,131] was found to occur i n the form of small i n c l u s i o n s or p r e c i p i t a t e s of m e t a l . It i s p o s s i b l e that the lower m o b i l i t i e s of ZnO-1 and ZnO-2 are due to Zn i n c l u s i o n s c a t t e r c e n t r e s and a l s o to the higher impurity c o n c e n t r a t i o n s of Ar and C that occur i n f i l m s d e p o s i t e d at low b i a s v a l u e s . 5.4 OPTICAL PROPERTIES Recently Greene [117] has pointed out the s c a r c i t y of d e t a i l e d c h a r a c t e r i z a t i o n of ZnO f i l m s , d e s p i t e the l a r g e amount of l i t e r a t u r e a v a i l a b l e on ZnO. T h i s s e c t i o n r e p o r t s the r e s u l t s of the f i r s t d e t a i l e d study of the o p t i c a l 96 p r o p e r t i e s of transparent conducting ZnO f i l m s and black Zn excess f i l m s . Other workers have c h a r a c t e r i z e d o p t i c a l p r o p e r t i e s by ju s t the t r a n s m i s s i o n spectrum i n the v i s i b l e and by the behaviour of the band edge with doping [16,1 1 2 , 1 1 8 , 1 1 9 , 1 2 0 ] . In a d d i t i o n , Morgan and Brodie [14] have reported an average f i l m r e f l e c t i v i t y f o r r a d i a t i o n from a 373 K blackbody. For both dark and transparent conducting ZnO f i l m s prepared at v a r i o u s v a l u e s of the s u b s t r a t e b i a s , I have measured the r e f r a c t i v e index and e x t i n c t i o n c o e f f i c i e n t f or the wavelength range 0.35 to 2.5 Mm, and the r e f l e c t i v i t y at wavelengths 2.5 to 20 Mm. The same set of f i l m s , on quartz s u b s t r a t e s , that were s t u d i e d by X-ray d i f f r a c t i o n ( F igure 5.5) and H a l l probe (Table 5.2) techniques were measured f o r t r a n s m i s s i o n and r e f l e c t i o n using the u l t r a v i o l e t , v i s i b l e and i n f r a r e d spectrophotometers d e s c r i b e d in s e c t i o n 2.1. In a d d i t i o n , a h i g h l y r e s i s t i v e (>10U Ocm) s t o i c h i o m e t r i c ZnO f i l m ( l a b e l l e d ZnO-7) was depo s i t e d i n the oxide mode of t a r g e t coverage with no s u b s t r a t e b i a s , and analyzed o p t i c a l l y . Determination of o p t i c a l c onstants was p o s s i b l e i n the wavelength range 0.35 to 2.5 Mm, where both t r a n s m i s s i o n and r e f l e c t i o n measurements were taken. Transmission measurements i n the middle i n f r a r e d (2.5 to 20Mm) were not p o s s i b l e due to the strong a b s o r p t i o n of the quartz s u b s t r a t e s . In the weakly absorbing wavelength r e g i o n s , the o p t i c a l c o nstants n and k were determined using a technique s l i g h t l y m o d i f i e d from that of M a n i f a c i e r et a l . [ 1 2 1 ] . T h i s technique 97 i s i l l u s t r a t e d i n the t r a n s m i s s i o n and r e f l e c t i o n c h a r a c t e r i s t i c s of sample ZnO-5 (deposited at -115 V bi a s ) i n Fi g u r e 5.10. Continuous curves were drawn connecting c o n s e c u t i v e i n t e r f e r e n c e maxima and c o n s e c u t i v e i n t e r f e r e n c e minima that enabled e x p l i c i t c a l c u l a t i o n of n and k using the independently determined f i l m t h i c k n e s s . In wavelength regions where f i l m s were s t r o n g l y a b s o r b i n g , standard equations f o r the t r a n s m i s s i o n and r e f l e c t i o n of a s i n g l e t h i n f i l m l a y e r on quartz s t r u c t u r e were s o l v e d f o r n and k using an i t e r a t i v e t e c h n i q u e . Both of these types of c a l c u l a t i o n are d e s c r i b e d in d e t a i l i n the Appendix, s e c t i o n s 7.1 and 7.2. Conducting f i l m t h i c k n e s s e s were chosen to be about 800 nm, t h i c k enough to e x h i b i t a few orders of i n t e r f e r e n c e s t r u c t u r e , yet t h i n enough to measure t r a n s m i s s i o n i n absorbing r e g i o n s . The r e s i s t i v e f i l m ZnO-7 was much t h i c k e r , 2.7 Mm, eo that i n t e r f e r e n c e s t r u c t u r e would extend i n t o the near i n f r a r e d region (1 to 2.5 Mm) where s t o i c h i o m e t r i c ZnO remains t r a n s p a r e n t . F i g u r e 5.11 shows the o p t i c a l c o n s t a n t s n and k that were c a l c u l a t e d f o r samples ZnO-1 (-7V b i a s ) , ZnO-2 (-40V b i a s ) , ZnO-5 (-115V bi a s ) and f o r the r e s i s t i v e sample ZnO-7. The e l e c t r i c a l p r o p e r t i e s of the conducting samples were summarized p r e v i o u s l y i n Table 5.2. The v i s u a l o b s e r v a t i o n of a t r a n s i t i o n form dark t o c l e a r ZnO with i n c r e a s i n g b i a s i s confirmed by the decrease of k values i n the v i s i b l e wavelength r e g i o n . The a b s o r p t i o n i n sample ZnO-7 was too small to be 98 Figure 5.10 Transmission and r e f l e c t i o n c h a r a c t e r i s t i c s of sample ZnO-5 on a quartz substrate. Continuous curves (indicated by the dashed line s ) were constructed connecting interference extrema to enable determination of the o p t i c a l constants of the f i l m . 99 Figure 5.11 Optical constants of ZnO f i l m s . Samples ZnO-1, ZnO-2 and ZnO-5 were deposited at substrate bias values of -6, -40, and -115 V respectively. Sample ZnO-7 was deposited in the oxide target mode and i s highly r e s i s t i v e . Experimental errors are 4% in n and 25% in k. 1 00 measured using my te c h n i q u e s , except for the few data p o i n t s near the band edge. T h i s sample had a constant r e f r a c t i v e index i n the v i s i b l e of 1.86±.08. R e f r a c t i v e index values f o r a ZnO s i n g l e c r y s t a l are shown by Mollwo [122] to vary between 1.98 and 2.14 in the wavelength region 670 to 430 nm. It i s o f t e n found that vacuum dep o s i t e d f i l m s have lower r e f r a c t i v e i n d i c e s than t h e i r corresponding bulk m a t e r i a l s , due in part to the lower d e n s i t y of d e f e c t e d or p o l y c r y s t a l l i n e f i l m s [ 123]. D e t a i l e d s t u d i e s of the band edge region (360 to 390 nm) were not made. Measurement of the strong a b s o r p t i o n near the band edge would r e q u i r e very t h i n f i l m s (100 nm) to ensure measureable t r a n s m i s s i o n . I n s p e c t i o n of the t r a n s m i s s i o n s p e c t r a of c l e a r f i l m s shows an o p t i c a l bandgap of approximately 3.3±0.1 eV. P r e c i s e measurements of the o p t i c a l bandgap i n conducting ZnO by other workers have given values between 3.2 and 3.5 eV [112,118]. These workers observed an i n c r e a s e of the o p t i c a l bandgap with i n c r e a s i n g c a r r i e r c o n c e n t r a t i o n , a t t r i b u t e d to the Burstein-Moss band f i l l i n g ef f e c t . F i g u r e 5.12 compares the i n f r a r e d r e f l e c t i o n s p ectra and r e s i s t i v i t y f o r samples ZnO-1, ZnO-3 and ZnO-5. As expected, the best conducting f i l m , ZnO-5, d i s p l a y s the highest r e f l e c t i v i t y . The r e f l e c t i v i t y s t r u c t u r e seen i n sample ZnO-1 i s due to the p r o p e r t i e s of the quartz s u b s t r a t e . As d i s c u s s e d i n s e c t i o n 5.3, b i a s s p u t t e r e d ZnO f i l m s can be c l a s s i f i e d i n t o two general groups. F i l m s at low b i a s F i g u r e 5.12 I n f r a r e d r e f l e c t i v i t y c h a r a c t e r i s t i c s of f i l m s ZnO-1, ZnO-3 and ZnO-5, d e p o s i t e d at s u b s t r a t e b i a s values of -6, -65 and -115 V r e s p e c t i v e l y . 1 02 (ZnO-1 and ZnO-2) have a l a r g e Zn excess r e s u l t i n g i n low e l e c t r o n m o b i l i t i e s , high r e s i s t i v i t y and poor i n f r a r e d r e f l e c t i v i t y , and are s t r o n g l y absorbing in the v i s i b l e . The ZnO f i l m s d e p o s i t e d at high s u b s t r a t e b i a s (ZnO-3 to ZnO-6) are n e a r l y s t o i c h i o m e t r i c , o p t i c a l l y t r a n s p a r e n t , have high e l e c t r o n m o b i l i t i e s r e s u l t i n g i n low r e s i s t i v i t y and good i n f r a r e d r e f l e c t i v i t y . The transparent conducting ZnO f i l m s are of most i n t e r e s t i n t h i s t h e s i s work. T h e i r o p t i c a l and e l e c t r i c a l p r o p e r t i e s may be r e l a t e d using the f r e e e l e c t r o n approx imat i o n . The Drude theory of f r e e e l e c t r o n s , i n t r o d u c e d i n s e c t i o n 1.3, has been used s u c c e s s f u l l y to model the i n f r a r e d r e f l e c t i o n p r o p e r t i e s of the transparent conductors ITO [124], CTO [ 2 9 ] , and Sn02 [ 2 0 ] . T h i s theory has not p r e v i o u s l y been a p p l i e d to conducting ZnO. The measured e l e c t r i c a l parameters of m o b i l i t y /u and c a r r i e r d e n s i t y N i n Table 5.2 have been used i n equations 1.4 and 1.5 to generate the Drude p r e d i c t i o n for o p t i c a l constants and r e f l e c t i v i t y . F i g u r e 5.13 shows the experimental and p r e d i c t e d i n f r a r e d r e f l e c t i v i t y of sample ZnO-5, using the independently measured e l e c t r i c a l parameters M=19 cm2/Vs and N=1.6X102 0 cm- 3. Instead of using the t a b u l a t e d value of 6^=3.73 f o r bulk ZnO, I used ew=3.46 c a l c u l a t e d from measurement of the o p t i c a l c onstants of the s t o i c h i o m e t r i c , r e s i s t i v e sample ZnO-7. The plasma wavelength was c a l c u l a t e d as Xp=2.8 ym. The good f i t to the experimental data shows that the f r e e e l e c t r o n theory of Drude prov i d e s a model for conducting ZnO. 103 -I 1 I I I I I | | _ 2 6 10 14 18 WAVELENGTH (pm) Figure 5 . 1 3 Comparison of the measured infrared r e f l e c t i v i t y of sample ZnO-5 with the r e f l e c t i v i t y predicted by the Drude free electron model. The measured e l e c t r i c a l transport parameters N = 1 . 6 X 1 0 2 0 cm"3 and M=19 cm2/Vs were used in the Drude model. 1 04 Figure 5.14 Comparison of the measured r e f r a c t i v e index n and absorption c o e f f i c i e n t a of sample ZnO-5 with the predictions of the Drude free electron model. The measured e l e c t r i c a l transport parameters N=1 .6X1 02 0 cm"3 and /tx=l9 cm2/Vs were used in the Drude model. 1 05 The Drude p r e d i c t i o n was a l s o extended i n t o the v i s i b l e wavelength r e g i o n . F i g u r e 5.14 shows the r e f r a c t i v e index n and the a b s o r p t i o n c o e f f i c i e n t a=47rk/X of sample ZnO-5. The values p r e d i c t e d from Drude theory d i s p l a y a rough f i t to the d a t a , the major di s c r e p a n c y being a l a r g e r p r e d i c t e d a b s o r p t i o n at small wavelengths (600nm<X<1500nm). However, at these wavelengths, where X<XP, the c l a s s i c a l Drude theory f a i l s . For high energy photons, the s c a t t e r i n g process i n v o l v e d in a b s o r p t i o n must be c o n s i d e r e d , and the value of the a b s o r p t i o n c o e f f i c i e n t becomes dependent on the type of s c a t t e r i n g mechanism i n v o l v e d . For n e u t r a l impurity s c a t t e r i n g , Dumke [125] has estimated that free e l e c t r o n theory should not be i n e r r o r by more than 20% at photon energies up to hv=8EF , or X=460 nm. For i o n i z e d impurity s c a t t e r i n g , Dumke p r e d i c t s that the Drude theory w i l l s i g n i f i c a n t l y overestimate the a b s o r p t i o n at small wavelengths. Thus, f o r conducting ZnO f i l m s i t i s found that Drude theory approximates the behaviour of o p t i c a l constants at wavelengths X<Xp, with d i s c r e p a n c i e s that are q u a l i t a t i v e l y p r e d i c t e d f o r s c a t t e r c e n t r e s of n e u t r a l and i o n i z e d i m p u r i t i e s . The v a l i d i t y of the Drude model was t e s t e d for absorbing Zn excess f i l m s (ZnO-1 and ZnO-2), i n s p i t e of the g r e a t e r u n c e r t a i n t y i n f i l m m i c r o s t r u c t u r e and conduction mechanisms. Fi g u r e 5.15 shows the f r e e e l e c t r o n plasma p r o p e r t i e s p r e d i c t e d using n=1•0 cm2/Vs and N = 3 . 3 x l 02° cm"3 f o r ZnO-1, compared with experimental o p t i c a l c onstants and i n f r a r e d 1 06 ZnO-1 —I I I I I I L _ 0.6 1.2 1.8 2.4 WAVELENGTH (jjm) 0.8 >->0.6 U LU _J U_ UJ or 0.4 0.2 ZnO-1 o ° Experimental o o 0or° o o o0OoooOO o o o o o o Drude 8 12 WAVELENGTH (pm) 16 F i g u r e 5.15 Comparison of the measured o p t i c a l constants and i n f r a r e d r e f l e c t i v i t y of sample ZnO-1 with those c a l c u l a t e d u s i n g the Drude f r e e e l e c t r o n model. The measured e l e c t r i c a l parameters N = 3 . 3X 1 02° cm"3 and /a=1.0 cm2/Vs were used i n the Drude model. 1 07 r e f l e c t i v i t y . Due to the lower i n f r a r e d a b s o r p t i o n of dark ZnO f i l m s , the r e f l e c t i v i t y c a l c u l a t i o n had to account for the s u b s t r a t e r e f l e c t i v i t y . T h i s c a l c u l a t i o n i s d e t a i l e d in the Appendix, s e c t i o n 7.2. The Drude model i s found to provide an estimate of the a b s o r p t i o n c o e f f i c i e n t w i t h i n a f a c t o r of 2, and f o l l o w s the i n f r a r e d r e f l e c t i v i t y t r e n d , but at values lower than the experimental spectrum. It i s not s u r p r i s i n g that the Drude model f a i l s s i n c e only the f r e e e l e c t r o n c o n t r i b u t i o n to the d i e l e c t r i c constant has been c o n s i d e r e d , and the c o n t r i b u t i o n s from p o s s i b l e p r e c i p i t a t i o n s or i n c l u s i o n s of excess z i n c have been i g n o r e d . Attempts to d e f i n e the nature of the Zn excess were made by b i a s s p u t t e r i n g 100 nm t h i c k ZnO samples onto copper m i c r o g r i d s f o r t r a n s m i s s i o n e l e c t r o n microscope (TEM) s t u d i e s . F i g u r e 5.16 compares TEM photographs of t r a n s p a r e n t conducting ZnO ( b i a s -100V) and black ZnO (no s u b s t r a t e d i s c h a r g e ) . Within the TEM r e s o l u t i o n of about 3 nm, no obvious d i f f e r e n c e i s seen i n the s i m i l a r g r a n u l a r appearance of each f i l m . The expected g r a i n s i z e of about 40 nm i s comparable with the s t r u c t u r e s seen in these photographs. Other workers have r e p o r t e d TEM photographs of s u b - s t o i c h i o m e t r i c l n203 [127] and AlN [126] that are remarkably s i m i l a r in appearance to ZnO. A l s o , i t was observed that a l a r g e change i n AlN composition was r e q u i r e d before TEM photographs r e s o l v e d a s t r u c t u r e d i f f e r e n c e [126,130]. No c o n c l u s i o n s concerning the e x i s t e n c e of Zn p r e c i p i t a t e s may be r e l i a b l y drawn from my TEM d a t a . In s p i t e of the u n c e r t a i n t y i n the black ZnO f i l m F i g u r e 5.16 TEM photomicrographs comparing ZnO f i l m s d e p o s i t e d with and without a s u b s t r a t e d i s c h a r g e . 109 m i c r o s t r u c t u r e , an i l l u s t r a t i v e c a l c u l a t i o n i s performed to show how the d i e l e c t r i c c o n s t a n t s may be m o d i f i e d by Zn i n c l u s i o n s . For t h i s c a l c u l a t i o n , I assume that i s l a n d s of pure Zn are randomly embedded in a matrix of ZnO. The simplest approach to the d e s c r i p t i o n of the d i e l e c t r i c p r o p e r t i e s of such a mixture was given by Maxwell-Garnett [128]. For s p h e r i c a l p a r t i c l e s of d i e l e c t r i c constant e,„, d i s p e r s e d in a medium of d i e l e c t r i c constant e, the C l a u s i u s - M o s s o t t i r e l a t i o n [134] for the p o l a r i z a b i l i t y of small spheres i s used to f i n d an average or e f f e c t i v e d i e l e c t r i c constant e; e = e + 3 f e ( e m - e ) (5.4) em+ 2 e - f ( e m - e ) where f i s the volume f r a c t i o n of embedded p a r t i c l e s . An expr e s s i o n f o r r d e t a i l i n g r e a l and imaginary components of the v a r i o u s d i e l e c t r i c c o n s t a n t s has been given by A f f i n i t o [131]. T h i s c a l c u l a t i o n assumes that the m e t a l l i c i s l a n d s are small with respect to the wavelength of l i g h t . Although there are many other s i m i l a r formalisms used to p r e d i c t the o p t i c a l p r o p e r t i e s of heterogeneous m a t e r i a l s , the Maxwell-Garnett theory has proved to be more s u c c e s s f u l i n t r e a t i n g the o p t i c a l p r o p e r t i e s of t h i n agglomerated metal f i l m s and of noble metal p a r t i c l e s imbedded i n g l a s s e s [128,129,130]. The d i e l e c t r i c constant of the ZnO m a t r i x , e, was c a l c u l a t e d f o r ZnO-1 from the Drude theory using M=1.0 cm2/Vs and N = 3 . 3 x l 02° cm"3. Tabulated values of the d i e l e c t r i c constant f o r bulk Zn [132,133] were used to d e s c r i b e the p r o p e r t i e s of Zn i s l a n d s , em . The volume f r a c t i o n , f=0.20 for 1 10 ZnO-1, was estimated from the RBS measurement of f i l m s t o i c h i o m e t r y and the known d e n s i t i e s of Zn and ZnO. The Maxwell-Garnett c a l c u l a t i o n was performed for sample ZnO-1 i n the i n f r a r e d , s i nce i t i s only i n t h i s wavelength region that the Drude theory for the o p t i c a l p r o p e r t i e s of the matrix i s expected to be v a l i d . F i g u r e 5.17 compares the measured i n f r a r e d r e f l e c t i v i t y of sample ZnO-1 with the Drude approximation and with the Maxwell-Garnett p r e d i c t i o n . I n c l u d i n g the e f f e c t s of excess Zn through the use of a simple e f f e c t i v e medium theory i s shown to improve the f i t to the experimental r e s u l t s . T h i s e x e r c i s e has c e r t a i n l y not proven that black ZnO i s composed of s p h e r i c a l p a r t i c l e s embedded in a ZnO m a t r i x , but i n d i c a t e s the need for the c o n s i d e r a t i o n of the e f f e c t s of f i l m m i c r o s t r u c t u r e ( p r e s e n t l y unknown) i n tandem with the e f f e c t s of conduction e l e c t r o n s i n order to adequately d e s c r i b e o p t i c a l c o n s t a n t s . 111 ZnO-1 0.8 >->0.6 U u _J u or 0.4 0.2 ^Experimental Maxwell-Garnett 1 NDrude 8 12 WAVELENGTH (jjm) 16 F i g u r e 5.17 Comparison of the measured i n f r a r e d r e f l e c t i v i t y of sample ZnO-1 with r e f l e c t i v i t i e s c a l c u l a t e d using 1) the Drude f r e e e l e c t r o n model and 2) the Maxwell-Garnett theory f o r a 0.20 volume f r a c t i o n of Zn i n c l u s i o n s . 1 1 2 CHAPTER SIX: SUMMARY This concluding chapter serves to summarize the new contributions of my work obtained from the f i r s t detailed study of react i v e l y bias sputtered conducting ZhO f i l m s . The reactive gas ba f f l e at the target was shown to enhance the oxidation of Zn sputtered from a metal target surface by increasing the p a r t i a l pressure of oxygen near the substrate and by decreasing the Zn flux to the substrate. When used in conjuction with an rf discharge at the substrate, the b a f f l e provided the further benefit of eliminating interaction between the dc target and rf substrate discharges. As suggested by other workers but not previously proven, the rf discharge at the substrate was shown to increase the re l a t i v e oxygen content of the f i l m . One mechanism of oxidation was found to be activation and ion platin g of oxygen species in the discharge. Other mechanisms active in promoting complete oxidation were pr e f e r e n t i a l resputtering and p r e f e r e n t i a l evaporation of excess z i n c . These re-emission processes were unexpected due to the low temperature substrates (380 K) and the low power substrate discharge (-100 V, 20 mW/cm2). However, resputter and evaporation rates were greatly enhanced above that expected for bulk Zn or bulk ZnO, attributed to the weak bonding of surface adatoms during f i l m growth. Nearly stoichiometric ZnO deposited at high oxygen p a r t i a l pressures was resistant to re-emission due to rapid oxidation and the strength of the Zn-0 bond. Zn excess films 113 d e p o s i t e d at low oxygen p a r t i a l p r e s s u r e s showed s i g n i f i c a n t m a t e r i a l re-emission (70%) due to slow o x i d a t i o n and the high vapour pr e s s u r e and weak bonding of Zn adatom aggregates. The enhanced re-emission phenomena found f o r ZnO might only be observed f o r a few other compounds, f o r instance C d2S n O « , where the m e t a l l i c c o n s t i t u e n t s e x h i b i t high vapour pressures and weak bonding. In c o n t r a s t , T i was chosen as a r e p r e s e n t a t i v e low vapour pr e s s u r e m e t a l . R e a c t i v e l y sputtered t i t a n i u m oxide f i l m s showed no m a t e r i a l re-emission f o r s u b s t r a t e b i a s e s up to -175 V. However, the oxygen ion p l a t i n g mechanism was s t i l l presumed to be a c t i v e , reducing the o p t i c a l a b s o r p t i o n a t t r i b u t e d to T i metal e x c e s s . ZnO was e s t a b l i s h e d as a new candidate m a t e r i a l for heat mi r r o r manufacture. Fi l m s s p u t t e r e d onto p o l y e s t e r sheet at a d e p o s i t i o n r a t e of 75 nm/min were found to have a t r a n s m i s s i o n to s o l a r energy of 75% and an 85% r e f l e c t i o n of 300 K blackbody r a d i a t i o n . In203:Sn f i l m s of r e s i s t i v i t y 9x10"" Gem were a l s o d e p o s i t e d onto p o l y e s t e r s h e e t , and at the time of p u b l i c a t i o n t h i s was the lowest recorded r e s i s t i v i t y f o r ITO on p l a s t i c s u b s t r a t e s . O p t i c a l , e l e c t r i c a l and s t r u c t u r a l p r o p e r t i e s were c h a r a c t e r i z e d f o r ZnO f i l m s as a f u n c t i o n of s u b s t r a t e b i a s from -7 V to -120 V. These p r o p e r t i e s are itemized below. 1. The r e l a t i v e oxygen content of the f i l m s i n c r e a s e d with b i a s from approximately ZnO<8 to ZnO. T h i s i n c r e a s e i s due to the o x i d a t i o n mechanisms j u s t d i s c u s s e d . 2. F i l m s d e p o s i t e d at h i g h b i a s had a much smoother surface 1 1 4 than the low bias f i l m s , attributed to a r e d i s t r i b u t i o n of deposited material by resputtering and evaporation processes. 3. The variation with bias of polycrystal orientation, size and stress was attributed to the competing e f f e c t s of c r y s t a l disruption by ion bombardment and c r y s t a l improvement due to a more stoichiometric composition. The best conducting films had a grain size of about 40 nm and a l a t t i c e s t r a i n of 0.8%. 4. Films deposited at high bias had electron m o b i l i t i e s near 20 cm2/Vs. It was proposed that t h i s mobility i s limited by ionized and neutral impurity s c a t t e r i n g . Low bias films had lower mo b i l i t i e s near 1 cm2/Vs, attributed to scatter centres created by the Zn excess. 5. Since the electron density was approximately constant with bias ( 2 X 1 02° cm"3), the f i l m r e s i s t i v i t y was minimized to 2X10"3 ficm at high bias. 6. Optical constants of the films were deduced for the wavelength range 0.4 to 2.5 Mm and r e f l e c t i v i t y from 2.5 to 20 nm wavelength. The best conducting films were v i s u a l l y transparent and had a measured infrared r e f l e c t i v i t y in good agreement with that calculated using the Drude theory of free electrons. The low mobility films were v i s u a l l y dark. Free electron absorption only partly accounted for this dark appearance. 1 15 CHAPTER SEVEN: APPENDIX 7.1 OPTICAL CONSTANTS OF TRANSPARENT FILMS To deduce the o p t i c a l constants of transparent ZnO films on quartz substrates, I used a modification of the technique proposed by Manifacier et a l . [121]. Figure 7.1 shows a ZnO film of complex r e f r a c t i v e index n-ik on a thick quartz substrate of r e f r a c t i v e index n, surrounded by a i r of r e f r a c t i v e index n0=1. Nominal thicknesses for the quartz and ZnO layers are given. For regions of n e g l i g i b l e absorption, t h i s thin f i l m structure exhibits interference effects in both r e f l e c t i o n and transmission spectra as seen previously in Figure 5.10. The exact expressions for transmitted and refl e c t e d l i g h t intensity from a 3 interface structure are cumbersome and, in general, d i f f i c u l t to solve. The method of Manifacier assumes a perfectly transparent substrate and weak absorption in the f i l m such that k2<(n-n0)2 and k2<(n-n1)2. Continuous functions T,(X) and T2(X) were constructed which are the envelopes of the transmission interference maxima and the transmission interference minima, respectively (Figure 5.10). The expressions for the o p t i c a l constants of the fil m then reduce to; n2 = N + (N2-n02n,2), / 2 (7.1) where N = (1/2)(n02+n,2) + 2n0n1(T1-T2)/T1T2 and exp(-47rkt/X) = C, [ 1 - (T,/T2 )'/2] (7.2) C2[1 + (T;/T2)'^] where C, = (n+n0)(n1+n), C2 = (n-n0)(n,-n) and t i s the f i l m thickness. Unfortunately the transmission, T(X), used by 116 d., = 700nm dp = 1mm _ Ai r ZnO Quartz n0=1 n + ik Ai r r 2 F i g u r e 7.1 A schematic of a ZnO f i l m on a quartz s u b s t r a t e l a b e l l i n g the t h i c k n e s s d, r e f r a c t i v e i n d i c e s n and i n t e r f a c e r e f l e c t i v i t i e s r . 1 17 Manifacier in this c a l i b r a t i o n was the light transmission into the substrate, not into the a i r as would be measured experimentally. To generate the transmission values T(X) required for these formulae from the experimental values Texp^)' 1 used a f i r s t order correction accounting for r e f l e c t i o n at the quartz-air interface; T(X) = Te x p(X)/[ 1-r(X) 3 (7.3) where r(X) i s the r e f l e c t i o n of a single quartz-air interface, and has a nominal value of 0.035. This modified technique was used to calculate n and k for the transparent, conducting samples ZnO-3, ZnO-4, ZnO-5, and ZnO-6 in the wavelength range 400 to 1400 nm and for the stoichiometric sample ZnO-7 in the range 400 to 2500 nm. Extinction c o e f f i c i e n t s were not greater than k=0.25, consistent with the assumption of weak absorption, k2<0.25. The r e f r a c t i v e indices n were determined to an accuracy of about 4% and k values to an accuracy of about 25%, due to experimental errors of 1.5% in transmission spectra and 7% in fi l m thickness. 7.2 OPTICAL CONSTANTS OF ABSORBING FILMS To determine the op t i c a l constants of absorbing films showing no interference e f f e c t s , I developed an i t e r a t i v e procedure based on the standard formulae for t o t a l transmission and r e f l e c t i o n of li g h t intensity in the ZnO/quartz structure. Using the notation of Figure 7.1, these formulae are written as; 1 18 T = exp(-47rkd , /X) (1-r 0 ) (1-r ,) (1 - r 2 ) (7.4) R = r 0 + e x p ( - 8 7 r k d 1 / X ) [ ( 1 - r 0 ) 2 r 1 + ( 1 - r 0 ) 2 ( l - r l ) 2 r 2 ] (7.5) where higher order terms involv ing mul t iple r e f l e c t i o n s have been neglected. This approximation introduces a maximum absolute error of 0.003 in T or R. The interface r e f l e c t i v i t i e s r are obtained from the appropriate r e f r a c t i v e i n d i c e s . For example; r , = ( n - n , ) 2 + k 2 (7.6) (n+n,) 2 + k 2 The values used for the r e f r a c t i v e index of quar tz , n 1 f were those tabulated by Mali tson [135]. The i t e r a t i v e procedure used to f i n d n and k is as f o l l o w s : 1. R, T, and d were measured. 2. X and r 2 were known. 3. I n i t i a l i z e r0=R and r ^ O . 4. Evaluate k using equation (7 .4) . 5. Evaluate r 0 using equation' (7 .5) . 6. Evaluate n from r 0 . 7. Calculate r , using the new n and k values . 8. Repeat steps 4 through 7. Convergence to three f igure agreement of successive n and k values was obtained af te r about 5 i t e r a t i o n s . This technique was used to deduce n and k for samples ZnO-1 and ZnO-2 at a l l wavelength values , and for samples ZnO-3, ZnO-4, ZnO-5 and ZnO-6 at wavelengths X<400 nm and X>1400 nm. Uncer ta int ies in the f i n a l values of n and k due to experimental errors are s i m i l a r to those of the previous c a l c u l a t i o n ; An/n=4% and Ak/k=30%. The o p t i c a l constants c a l c u l a t e d in both the absorbing and transparent approximations at the boundary between the ' absorbing ' and ' t ransparent ' regions were found to agree within the 119 c a l c u l a t i o n a l u n c e r t a i n t y f o r a l l f i l m s . Equation (7.5) was a l s o used to c a l c u l a t e the t h e o r e t i c a l i n f r a r e d r e f l e c t i v i t y of the low m o b i l i t y (ZnO-1 and ZnO-2) ZnO f i l m s on q u a r t z . Values of n and k f o r ZnO were determined from the Drude theory and valu e s of the o p t i c a l c o nstants of quartz i n the i n f r a r e d are t a b u l a t e d by P h i l l i p [136]. Thus the i n f r a r e d r e f l e c t i v i t y c a l c u l a t e d from these f i l m s shows anomalous s t r u c t u r e ( F i g u r e 5.17) due to strong a b s o r p t i o n bands i n the q u a r t z s u b s t r a t e . High m o b i l i t y ZnO f i l m s are s t r o n g l y absorbing i n the i n f r a r e d and d i d not r e q u i r e c o n s i d e r a t i o n of s u b s t r a t e r e f l e c t i v i t y . 1 20 REFERENCES 1. R. McMahon, Solnote 3, Solar Energy S o c i e t y of Canada Inc., (1980). 2. J . Fan and F. Bachner, A p p l . Opt. Y5, 1012 (1976). 3. L. H o l l a n d and S. S i d d a l l , B r . J . A p p l . Phys. 9, 359 (1958). 4. H. G l a s e r , G l a s s Tech. 2J_, 254 (1980). 5. C M . Lampert, S o l . Energy Mat. 6, 1 (1981). 6. J . L . Vossen, Physics of Thin Films 9, 1 (1977). 7. K. Chopra, S. Major and D. Pandya, Thin S o l i d Films 102, 1 (1983). 8. Z.M. J a r z e b s k i , Phys. S t a t . S o l . (a) 7j_, 13 (1982). 9. F. S e i t z , Modern Theory of S o l i d s , (McGraw-Hill, New York, 1940) p.638. 10. The Southwall C o r p o r a t i o n , Palo A l t o , C a l i f o r n i a , U.S.A. 11. M. B r e t t , R. McMahon, J . A f f i n ' i t o and R. Parsons, J . Vac. S c i . T e c h n o l . AJ_, 352 (1983). 12. M.J. B r e t t and R.R. Parsons, S o l . State Comm. 54, 603 (1985). 13. M.J. B r e t t and R.R. Parsons, Can. J . Phys., i n p r e s s . 14. J . Morgan and D. B r o d i e , Can. J . Phys. 60, 1387 (1982). 15. J.B. Webb and D.F. W i l l i a m s , A p p l . Phys. L e t t . 39, 640 (1981). 16. T. Minami, H. Nanto and S. Takata, A p p l . Phys. L e t t . 41, 958 (1982). 17. L. Ryabova, V. Salun and I . Ser b i n o v , Thin S o l i d Films 92, 327 (1982). 18. R. Groth and E. Kauer, P h i l l . Tech. Rev. 26, 105 (1965). 19. P r i c e quoted by Canada Metal Inc., Vancouver, Canada. 20. F. Simonis, M. van der L e i j and C. Hoogendoorn, S o l . Energy Mat. J_, 221 (1979). 21. F. Smith and S. Lyu, J . Electrochem. Soc. 128, 1083 1 2 1 ( 1981). 22. N. M i y a t a , K. Miyake and S. Nao, Thin S o l i d F i l m s 5_8, 385 (1979). 23. S. Maniv, C. Miner and W.D. Westwood, J . Vac. S c i . T e c h n o l . j_8, 195 ( 1 981 ) . 24. D. H a l l , J . Electrochem. Soc. 124, 804 (1977). 25. G. Haacke, A p p l . Phys. L e t t . 28_, 622 (1976). 26. I. Hamberg, A. H j o r t s b e r g and C. G r a n q v i s t , A p p l . Phys. L e t t . 40, 362 (1982) . 27. J . Machet, J . G u i l l e , P. S a u l n i e r and S. Robert, Thin S o l i d F i l m s 80,149 (1981). 28. H. K o s t l i n , R. J o s t and W. Lems, Phys. S t a t . S o l . (a) 29, 87 (1975). 29. R.P. Howson, M. Ridge and C. Bishop, Thin S o l i d Films 80, 137 (1981 ) . 30. D. F r a s e r and H. Cook, J . Electrochem. Soc. 119, 1368 (1972). 31. M. Buchanan, J.B. Webb and D.F. W i l l i a m s , Thin S o l i d F i l m s 80, 373 (1981). 32. Y. Hsu and S. Ghandhi, J . Electrochem. Soc. 127, 1592 (1980). 33. J . M a n i f a c i e r , J . F i l l a r d and J . Bind, Thin S o l i d Films 77, 67 (1981). 34. P. P e t r o u , R. Singh and D.E. B r o d i e , A p p l . Phys. L e t t . 35, 930 (1979). 35. J . A r a n o v i c h , A. O r t i z and R. Bube, J . Vac. S c i . T e c h n o l . J_6, 994 (1979). 36. G. H e i l a n d , E. Mollwo and F. Stockmann, S o l i d S tate P h y s i c s 8, 191 (1959). 37. F. Kroger, The Chemistry of Imperfect C r y s t a l s , ( I n t e r s c i e n c e , New York, 1973) p.744. 38. T. Wada, S. K i k u t a , M. K i b a , K. Kiyozumi, T. Shimojo and M. Kakehi, J . C r y s t a l Growth 59, 363 (1982). 39. D. B l o c k , A. Herve and R. Cox, Phys. Rev. B25, 6049 (1982). 1 22 40. C. Gonzalez, D. Block, R.T. Cox and A. Herve, J . C r y s t a l Growth 59, 357 (1982). 41. W.S. Baer, Phys. Rev. J_5_4, 785 (1967). 42. E. P a r a d i s and A. Shuskus, Thin S o l i d F i l m s 38, 131 (1976). 43. S.B. Krupanidhi and M. Sayer, J . A p p l . Phys. 56, 3308 (1984). 44. M. Royer, J . Holmen, W. Wurm, 0. Aadland and M. Glenn, Sensors and Actuators 4, 357 (1983). 45. A.P. Roth and D.F. W i l l i a m s , J . Electrochem. Soc. 128, 2684 (1981). 46. J.C. M a n i f a c i e r , Thin S o l i d F i l m s 90, 297 (1982). 47. J . L . Vossen and W. Kern, Thin F i l m Processes, (Academic P r e s s , New York, 1978). 48. R.F. Bunshah, D e p o s i t i o n Technologies for Films and  C o a t i n g s , (Noyes P u b l i c a t i o n s , Park Ridge, 1982). 49. J . Rostworowski and R. Parsons, J . Vac. S c i . T e c h n o l . , i n p r e s s . 50. R. Howson, J . A v a r i t s i o t i s , M. Ridge and C. Bishop, Thin S o l i d F i l m s 58, 379 (1979). 51. R. Howson and M. Ridge, Thin S o l i d Films 119 (1981). 52. R. Brown, in Handbook of Thin F i l m Technology, e d i t e d by L. M a i s s e l and R. Glang, (McGraw-Hill, New York, 1970) p.6-39. 53. S. S c h i l l e r , U. H e i s i g , K. S t e i n f e l d e r , J . Strumpfel and W. S i e b e r , Vak. Tech. 30, 3 (1981). 54. A . J . S t i r l i n g and W.D. Westwood, Thin S o l i d F i l m s 7, 1 (1971). 55. J . H e l l e r , Thin S o l i d F i l m s _T7, 163 (1973). 56. J . A f f i n i t o and R.R. Parsons, J . Vac. S c i . T e c h n o l . A2, 1275 (1984). 57. S. Maniv, C.J. Miner and W.D. Westwood, J . Vac. S c i . T e c h n o l . AJ_, 1370 (1983). 58. G. Kominiak, J . Vac. S c i . T e c h n o l . j_2, 689 (1975). 59. B r i a n Chapman, Glow Discharge P r o c e s s e s , (John W i l e y , New 1 23 York, 1980) chapter 5. 60. C. Horwitz, J . Vac. S c i . T e c h n o l . AJ_, 60 (1983). 61. K. K o h l e r , J.W. Coburn, D.E. Home, E. Kay and J.H. H e l l e r , J . A p p l . Phys. 57, 59 (1984). 62. J . Thornton and A. P e n f o l d , i n Thin F i l m P r o c e s s e s , e d i t e d by J . Vossen and W. Kern (Academic P r e s s , New York, 1978) p.99. 63. J.W. Coburn and E. Kay, J . A p p l . Phys. 43, 4965 (1972). 64. Y. Okamoto and H. Tamagawa, J . Phys. Soc. Japan 23, 187 ( 1970) . 65. W. Davis and T. V a n d e r s l i c e , Phys. Rev. 131, 219 (1963). 66. H.R. Koenig and L . I . M a i s s e l , IBM J . Res. Dev. _1_4, 168 (1970). 67. J . Machet, P. S a u l n i e r , J . Ezquera and J . G u i l l e , Vacuum 33, 279 (1983). 68. D.M. Mattox, Electrochem. Tech. 2, 295 (1964). 69. D.M. Mattox, i n Pr o c . Conf. Ion P l a t i n g and A l l i e d  Techniques, p.1 (1979). 70. D.M. Mattox, i n D e p o s i t i o n Technologies f o r Fil m s and  C o a t i n g s , e d i t e d by R.F. Bunshah (Noyes P u b l i c a t i o n s , Park Ridge, 1982) p.244. 71. J . E . Greene and S.A. B a r n e t t , J . Vac. S c i . T e c h n o l . 21, 285 (1982). 72. J.A. Thornton, paper presented at the I n t e r n a t i o n a l Conference on M e t a l l u r g i c a l C o a t i n g s , San Diego, C a l i f o r n i a , 1983. 73. J.P. Coad and R.A. Dugdale, i n Pr o c . Conf. Ion P l a t i n g  and A l l i e d Techniques, p.186 (1979). 74. H. Wi n t e r s , D. Raimondi and D. Home, J . A p p l . Phys. 40, 2996 (1969). 75. J . Cuomo and R. Gambino, J . Vac. S c i . T e c h n o l . J_2, 79 (1975). 76. D. Dove, R. Gambino, J . Cuomo and R. K o b l i s k a , J . Vac. S c i . T e c h n o l . j_3, 965 (1976). 77. R. Howson, K. S u z u k i , C. Bishop and M. Ridge, Vacuum 34, 291 (1984). 1 24 78. Y. Igasaki and H. M i t s u h a s h i , Thin S o l i d Films 7_0, 17 (1980) . 79. J . P o i t e v i n , G. Lemperiere and J . Tardy, Thin S o l i d F i l m s 97, 69 (1982). 80. H.F. Winters and E. Kay, J . A p p l . Phys. 43, 794 (1972). 81. P. Mazzoldi and G. D e l i a Mea, Thin S o l i d Films 7_7, 181 (1981) . 82. J . Sundgren, B. Johansson and S. K a r l s s o n , Thin S o l i d F ilms 80, 77 (1981 ) . 83. J.A. Thornton and J . L . Lamb, Thin S o l i d Films 119, 87 (1984). 84. S. Dushman, S c i e n t i f i c Foundations of Vacuum Technique, (John Wiley, New York, 1982) p.693. 85. B. Chapman, Glow Discharge P r o c e s s e s , (John Wiley, New York, 1980) appendix 5. 86. D.G. T e e r , J . Phys. D9, L187 (1976). 87. M. Hecq, A. Hecq and M. F o n t i g n i e s , Thin S o l i d Films 115, L45 (1984). 88. G. C a r t e r and J . C o l l i g o n , Ion Bombardment of S o l i d s , (Heineman, London, 1968) chapter 6. 89. S. Barnett and J . Greene, S u r f . S c i . 128, 401 (1983). 90. Y. S h i n t a n i , K. N a k a n i s h i , T. Takawaki and 0. Tada, Jap. J . A p p l . Phys. J_4, 1875 (1975). 91. J . Hanak and J . P e l l i c a n e , J . Vac. S c i . T e c h n o l . V3, 406 (1976). 92. J . Cuomo, R. Gambino, J . Harper and J . K u p t s i s , J . Vac. S c i . T e c h n o l . _1_5, 281 (1978). 93. R. Jones, C. Standley and L. M a i s s e l , J . A p p l . Phys. 38, 4656 (1967). 94. V. Nwoko and H. U h l i g , J . Eletrochem. Soc. n_2, 1181 (1965). 95. S. Kelemen, Y. G o l d s t e i n and B. A b e l e s , S u r f . S c i . 116, 488 (1982). 96. J . F . O'Hanlon, A User's Guide to Vacuum Technology, (John W i l e y , New York, 1980) p.371. 1 25 97. L. M a i s s e l , R. Jones and C. Standley, IBM J . Res. Dev. 14, 176 (1970). 98. H. Winters and E. Kay, J . A p p l . Phys. 3_8, 3928 (1967). 99. J . F . O'Hanlon, A User's Guide to Vacuum Technology, (John Wiley, New York, 1980) p.207. 100. American S o c i e t y f o r T e s t i n g and M a t e r i a l s , Document ASTM-E838, 1981. 101. F.S. H i c k e r n e l l , Proceedings of the 1979 U l t r a s o n i c s Symposium, (IEEE, New York, 1979) p. 932. 102. T. Minami, H. Nanto, S. Shooji and S. Takata, Thin S o l i d F i l m s 111, 167 (1984) . 103. J . A f f i n i t o , N. F o r t i e r and R.R. Parsons, J . Vac. S c i . T e c h n o l . A2, 316 (1984). 104. S. Maniv, W. Westwood and E. Columbini, J . Vac. S c i . T e c h n o l . 20, 162 (1982). 105. C. Gawlak and C. A i t a , J . Vac. S c i . Technol. AJ_, 415 (1983). 106. F. S a r o t t , Z. Iqbal and S. Veprek, S o l . State Comm. 42, 465 (1982). 107. D.M. Mattox and G. Kominiak, J . Vac. S c i . T e c h n o l . 9, 528 (1972). 108. J.A. Thornton, Thin S o l i d F i l m s 40, 335 (1977). 109. B. C u l l i t y , Elements of X-ray D i f f r a c t i o n , (Addisson-Wesley, London, 1959) p.99. 110. C. K i t t e l , I n t r o d u c t i o n to S o l i d State P h y s i c s , (John Wiley and Sons, New York, 1976) p.162. 111. J . Pankove and E. Annavedder, J . A p p l . Phys. 3_6, 3948 (1965). 112. O. C a p o r a l e t t i , S o l . Energy Mat. 7, 65 (1982). 113. R. M a n s f i e l d , P r o c . Phys. Soc. B69, 76 (1956). 114. E. Conwell and V. Weisskopf, Phys. Rev. 388 (1949). 115. C. E r g i n s o y , Phys. Rev. 79, 1013 (1950). 116. P. Grosse, F. Schmitte, G. Frank and H. K o s t l i n , Thin S o l i d F i l m s 90, 309 (1982). 1 26 117. J . Greene, CRC C r i t . Rev. S o l i d State Mater. S c i . J J _ , S u ppl. 3, 189 (1984). 118. A. Roth, J . Webb and D. W i l l i a m s , Phys. Rev. B25, 7836 (1982) . 119. K. I t o and T. Nakazawa, Jap. J . A p p l . Phys. 22, L245 (1983) . 120. P.S. Nayar, J . E l e c t . Mater. _M, 967 (1982). 121. J . M a n i f a c i e r , J . Gas i o t and J . F i l l a r d , J . Phys. E9, 1002 (1976). 122. E. Mollwo, Z. Agnew. Physik 6, 257 (1954). 123. W. P l i s k i n and S. Zanin, in Handbook of Thin F i l m  Technology, e d i t e d by L. M a i s s e l and R. Glang, (McGraw-Hill, New York, 1970) p.11-43. 124. J . Fan and F. Bachner, J . Electrochem. Soc. 123, 1719 ( 1975). 125. W. Dumke, Phys. Rev. j_24, 1813 (1961). 126. Normand F o r t i e r , unpublished data (1985). 127. A.F. Hebard and S. Nakahara, A p p l . Phys. L e t t . 4J_, 1130 ( 1982). 128. J . Maxwell-Garnett, P h i l . T r a n s . R. Soc. Lond. 205, 237 (1906). 129. S. Norrman, T. Andersson and C. G r a n q v i s t , Phys. Rev. B18, 674 (1978). 130. R.H. Doremus, J . A p p l . Phys. 3_7, 2775 (1966). 131. J . A f f i n i t o , PhD T h e s i s , U n i v e r s i t y of B r i t i s h Columbia, 1984. 132. A. Lenham and D. Treherne, P r o c . Phys. Soc. 8_3, 1059 (1964). 133. A. Lenham and D. Treherne, J . Opt. Soc. Am. 5_6, 752 (1966) . 134. J.D. Jackson, C l a s s i c a l E l e c t r o d y n a m i c s , (John W i l e y , New York, 1975) p.155. 135. I.H. M a l i t s o n , J . Opt. Soc. Am. 55, 1205 (1965). 136. H.R. P h i l l i p , J . A p p l . Phys. 50, 1053 (1979). 1 27 137. M.P. Seah, Thin S o l i d F i l m s 8J_, 279 (1981). 138. D.K. M u r t i , A p p l . S u r f . S c i . _M, 308 (1982). 139. K. Kim, W. B a i t i n g e r , J . Amy and N. Winograd, J . E l e c t r . S p e c t r . R e l . Phenom. 5, 351 (1974). 140. F.A. Grant, Rev. Mod. Phys. 3J_, 646 (1959). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0085519/manifest

Comment

Related Items