UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies on thin films of gold Chaurasia, Hari Krishna 1965

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

Item Metadata

Download

Media
831-UBC_1965_A7 C52.pdf [ 7.15MB ]
Metadata
JSON: 831-1.0104838.json
JSON-LD: 831-1.0104838-ld.json
RDF/XML (Pretty): 831-1.0104838-rdf.xml
RDF/JSON: 831-1.0104838-rdf.json
Turtle: 831-1.0104838-turtle.txt
N-Triples: 831-1.0104838-rdf-ntriples.txt
Original Record: 831-1.0104838-source.json
Full Text
831-1.0104838-fulltext.txt
Citation
831-1.0104838.ris

Full Text

STUDIES ON THIN FILMS OF GOLD by HARI KRISHNA CHAURASIA B.E.(Hons.), University of Saugar, 1956 M.E., University of Jabalpur^ 1958 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF ; MASTER OF APPLIED SCIENCE i n the Department of E l e c t r i c a l Engineering We accept th i s thesis as conforming to the standards required from candidates for the degree of Master of Applied Science Members of the Department of E l e c t r i c a l Engineering THE UNIVERSITY OF BRITISH COLUMBIA October 1964 > I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the r e q u i r e m e n t s f o r an advanced degree at the U n i v e r s i t y of • B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y , I f u r t h e r agree t h a t p e r -m i s s i o n f o r e x t e n s i v e c o p y i n g of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood t h a t ; c o p y i n g or p u b l i -c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n * Department of £ /ec^r i c A / Erxgfnzeri'tfjj The U n i v e r s i t y of B r i t i s h Columbia, Vancouver 8, Canada Date 13 • //. 19 £t+ ABSTRACT Investigations are made into the properties of 20 — 600 A gold films deposited onto freshly cleaved mica substrates by high vacuum evaporation. An accurate method of determining the equivalent thickness of a f i l m i s developed by combining a radio frequency technique employing quartz crystals with direct weighing. It i s found that thicknesses as low as 0*5 A can be monitored to within ± 5#. An attempt i s made to improve the structure of the gold films by using rapid deposition rates (15 to 20 A per second) and by depositing the gold films on nucleating s i l v e r layers of molecular thicknesses. The effect of increasing nucleating layer thicknesses (up to 25 A) i s studied by electron microscopy and d.c. resistance measurements. A c r i t i c a l nucleating layer thickness for s i l v e r i s found to exist between 5 to 7 A. On these c r i t i c a l layers, gold films down to 25 A are found conducting. A l l the films tested show an i r r e v e r s i b l e change i n conductivity with heating. Without a s i l v e r layer, continuous decrease i n the conductivity of gold films i s observed, the fi l m being destroyed at 450° C. On the other hand, increasingly better conductivities are observed by heating the gold-on-silver films, and an optimum annealing temperature of about 350° C i s indicated. Above this temperature the conductivity decreases; however, films are s t i l l conducting and continuous at 600° C. These films are, therefore, suitable as heat r e f l e c t i n g windows. i i D.C, and high frequency measurements (at 9 . 7 Gc) on gold films are given. In both cases, almost i d e n t i c a l values of surface resistance are obtained i n the range 1 0 to 1 , 0 0 0 ohms per square* A method for preparing gold f i l m bolometer elements using interrupted deposition of gold on a nucleating s i l v e r layer i s discussed. As very thin gold on s i l v e r films with high d.c» con-d u c t i v i t i e s are found to have low microwave transmission c o e f f i c i e n t s , the possible application of v i s u a l protection from microwave radiation i s discussed. i i i ACKNOWLEDGEMENT I wish to express my thanks to Dr. W.A.G, Voss, the supervisor of this project, for many hours of invaluable discussion and for his guidance throughout this study. To Drs.G.B. Walker, L. Young, M. Kharadly, CG. Englefield, E. Teghtsoonian and Mr. R.J. Stockwell I am grateful for their many useful suggestions. I am specially indebted to Mr. E.M. Edwards for his help with the electronic c i r c u i t s . Acknowledgement i s gra t e f u l l y given to Mr. T. Clark of the B r i t i s h Columbia Research Council for weighing the crystals and Dr. D. Tromans and Mr. E. von Tiesenhausen of the Department of Physical Metallurgy, U.B.C, for preparing the electron micrographs. I am grateful to Mr. N. Pentland for his technical help. To Messrs. D.R. McDiarmid, F.S. Chute and F.E. Vermeulen, I should l i k e to express my appreciation for many useful suggestions. This work was carried out under the auspices of the Colombo Plan with f i n a n c i a l assistance from the External Aid Office, Government of Canada; while the special leave of absence was granted by the Government of Madhya Pradesh, India. To these authorities, and to Dr. F. Noakes, Head of the Department of E l e c t r i c a l Engineering, U.B.C, I am grateful for the valuable f a c i l i t i e s provided. Special thanks and deep gratitude are owed to my wife, Meena, for her patience and s a c r i f i c e during my long sojourn abroad. x TABLE OF CONTENTS Page LIST OF TABLES . v i LIST OF ILLUSTRATIONS . v i i LIST OF PLATES ix ACKNOWLEDGEMENT • o 0 » » » * * » o * « » » * « » » * * » » o o * o « o o * o o » o o o * » x 1 o INTRODUCTION e > « > . » # » o © » « » » » » o * o * » « e « o f i » * o * < » * > o * * * * . > e * 1 2 • DEPOSITION METHODS » * « » » » © * » « * © » e © « * o » » © o © . » o o o o » » © T 2 * 1 I H " t I* 0 (111 C "t/X O H * ^ , f > & » « a o e o » o o » Q 9 < > o e © o © o o o o o o e o Q * T 2 . 2 Film Deposition Technique 1 1 3 . THE MEASUREMENT OF FILM THICKNESS .. 2 2 3 * 1 Ill"fcr O d l l C ~b X O H • • o « * » f t » o e a o e e » « » o » e « e o o D 9 e o * * * » 2 2 3 . 2 Change i n the Resonant Frequency of a Quartz Crystal with P a r a s i t i c Mass Deposition ...... 2 5 3 • 3 C X C U l "t • » * * i » » « o » o » © o o o » o o « 6 e o o c » * o » o o o o o o o c > © 2 6 3 o 4 Cl*y S t a l CoOllIlg * o « « o e o o f t a o o o < a o 6 o © * o o o o o o « a e » 2 6 3 © 5 Calibration of Crystals » © © © « o o b o 0 © o » © » © » o » © o 2 8 3 * 6 Discussion * » » » « » « » * e o « « i > 6 » o o » » o « o « o » e * o * o » » o 3 0 4 . ELECTRON MICROSCOPIC STUDIES ON THIN FILMS 3 6 4 . 1 Films for Electron Microscopy 3 6 4 . 2 Carbon Evaporation .......................... 3 6 4 . 3 Electron Microscopy Results ................. 3 8 5 . D.C. RESISTIVITY MEASUREMENTS ON THIN FILMS 4 5 5 . 1 D.C. R e s i s t i v i t y of Thin Films 4 5 5 . 2 Measurement of D.C. R e s i s t i v i t y 4 7 5 . 3 Types of Films Studied 4 8 5 . 4 Results . . . . * » • . • « » . . . » o « . » . . . » . . » O . » D O « . . . 5 0 5 . 5 Discussion . . . . . . . . . . . . . . o o . . . . . . . . . . . . . , . * . . 5 0 iv Page 6. MICROWAVE MEASUREMENTS • 62 6.1 Intro due t i on . . . . . . . . . a . . . . . . . . . . « . . » « > . . . . . * 62 6.2 Design of the Film Holder 63 6.3 Measurement of Microwave Surface Resistance . 66 6.4 Measurement of Microwave Transmission 6.5 Discussion 72 T o CONCLUS ION » e o » » « » t » * » » * » * o « » o * « o o « o o « » o « « « » « * e e « e * S3 REFERENCES • • • • » o « * » * » * * 0 * » * * » « f t « o o o o a o « » o o » « » » » a a i * » » o 33 v LIST OF TABLES I. Calibration of Crystals with Gold Films 31 I I . Calibration of Crystals with S i l v e r and Gold I I I . Affect of Annealing on Gold and Gold-on-Silver Films . . . . . . . . . . . . . . . . . . . , « « o . » « o » » o » o o . « . o » o b . . * » . 52 IV. Microwave and Light Transmission of Films ......... 80 V. Gold Film for Bolometer Element • • 81 v i LIST OP ILLUSTRATIONS Page 1. Film Configuration .»..«•»••••• • . 13 2. The Substrate Holder 0<>»».»»»»...»».....»..• 14 3. Details of the Base Plate Assembly ......... 16 4. Substrate Holder Assembly for Film Deposition • « » o * » o o o « * i » » * o » f r * a » * * « o » o o o * » « * f l 19 5» Substrate Holder Assembly for Contact F/vaporation « . . . • . . . » . » » . o « . . . . . . . . . » . • » » . . . 19 6(a) Block Schematic Diagram of the Thickness Measur ing System . . . , . . . . . * . . . » . . » . . . o o o o . o . 27 6(b) C i r c u i t Diagram for Thickness Measurement .. 27 7. Calibration Curve for the Crystals ......... 33 8. Apparatus for Carbon Evaporation . . . . . . . . . . i 37 9. The Surface Resistance of a Thin Film ...... 47 10. Relative R e s i s t i v i t y of Gold Films ......... 51 11. Eff e c t of t. on Relative R e s i s t i v i t y ...... 53 Ag 12(a) Film Deposition Over a Cleavage Step ....... 55 12(b) Cleavage Lines on Mica Substrate 55 13. Film Holder Details .»»»»..»•••............. 64 14. Transverse Film i n a Waveguide 68 15(a) High VSWR Me asurement with Minima as the 15(b) Vector Diagram for High VSWR Measurement ... 69 16. Apparatus for the Measurement of Microwave S U I * JL £L C e R© S 1 S ~fc ctri CG » » o » « « - * o « * « * » o o e » o o « o B i » e » 73 17» Apparatus for the Measurement of Microwave Transmission Coefficients ............ 73 18. Relative R e s i s t i v i t y at Microwaves vs Thick-19» Microwave Surface Resistancej R , vs D.C. Surface Resistance, R, . 0.................. 75 * d.c v i i 20* Transmission Coefficient vs Thickness ...... 76 21. Transmission Coefficient vs D,C. Surface v i i i LIST OF PLATES Pi £t"t© S I "t/O VI 6 » « » * 0 < * » » e » » * f r t > O O O O i ? * O Q O O » e D 0 0 0 0 9 0 » » 39 "t'O 44 (a) Electron D i f f r a c t i o n Patterns, (b) Transmission Electron Micrographs, Magnification 20,000. (c) Transmission Electron Micrographs, Magnification 100,000. i x 1. INTRODUCTION The deposition of metals i n the form of thin films was observed about a hundred years ago as sputtered metallic layers on the walls of glow discharge tubes and evaporated carbon coatings on the glass envelopes of Edison's incandescent carbon filament lamps. In 1857, Faraday^^ succeeded i n preparing films of a number of metals chemically and, also, by explosively evaporating metal wires i n an inert hydrogen atmosphere. How-ever, i t was not u n t i l the turn of the century that thin films (2) evoked any theoretical i n t e r e s t . Thomson ' i n 1901, was the f i r s t to suggest that the e l e c t r i c a l conductivity of metals i n the form th i n films, with, thicknesses comparable'to or less than the mean free path of the conduction electrons, must be less than that of the bulk metal. He attributed this lowering of conductivity to the decrease of the mean free path resulting from the diffuse ( i n e l a s t i c ) scattering of the electrons at the f i l m boundaries. This concept alone was a major topic of (3-9) research work on thin films for the next f i f t y years, ' and theoretical and experimental investigations on the e l e c t r i c a l conductivity of thin films continue even to date»(10-19) ^ number of recent publications have been concerned with nonmetallic (20-23) conduction i n thin f i l m s v ' when the deposit i s i n the form of descrete islands rather than a continuous layer. The con-duction mechanism i n such films i s due to the thermionic emission of electrons between the islands, which gives r i s e to a negative (21) temperature c o e f f i c i e n t of r e s i s t i v i t y , as shown by Hartman. ' There are several reasons for investigating the properties of metals, semiconductors and d i e l e c t r i c s i n the form of thin f i l m s . These are 1. Certain measurements can be made more conveniently using t h i n films, as i n (a) electron microscopic and electron d i f f r a c t i o n studies of structural order, dislocations, and migration phenomena i n metallurgy, (b) surface phenomena and surface reactions i n chemistry, and (c) superconductivity at millimeter and submillimeter wavelengths.^ 2 4^ 2. There i s a p o s s i b i l i t y of obtaining cold emission devices consisting of a metal-dielectric-metal sandwich (25) structure. Recent work by Scales indicates that the current-voltage characteristics of these junctions involve both a tunnelling e f f e c t and thermionic emission at normal temperatures, providing that the top emitting metal electrode and the d i e l e c t r i c layer are within a certain c r i t i c a l thickness range, 3. The electromagnetic properties of solids i n the form of t h i n films d i f f e r considerably from those of the bulk material• The effect of reduction of the f i l m thickness on magnetic properties i s also important because ferromagnetism i s a ( 9f>\ phenomenon characteristic to three dimensional crysta l s . Thin magnetic films have application as switching devices and memory elements i n computers. Another important aspect i s superconductivity. Unlike the t r a n s i t i o n from normal to superconducting state i n s t r a i n -free superconductors with decrease i n temperature, this trans-i t i o n i n thin films i s broad. This i s a consequence, not only of s t r a i n , but inhomogeneity, edge effects, and weak spots (27) inevitably associated with t h i n films. The effects which appear as a consequence of a reduction of the mean free path due to small dimensions may conveniently (16) be c a l l e d "mean free path e f f e c t s " or "size e f f e c t s " . These size effects show up i n many other phenomena besides e l e c t r i c a l conductivity* i . e . , thermal conductivity and thermoelectric (28 29 3l) (21 30) e f f e c t s , * ' temperature c o e f f i c i e n t of r e s i s t i v i t y , ' and the Hall E f f e c t . ^ 1 0 ^ The results of a large amount of theoretical research are now available, but comparatively l i t t l e experimental work has been (3—18) done. Many writers ' believe that the conductivity i n properly prepared thin films can be shown to obey the relations derived from the principles of the electron theory of metals as applied to the mean free path e f f e c t . Naturally* the essential requirement for getting useful experimental results i s that the conditions which are postulated during theoretical derivations must be s a t i s f i e d experimentally. According to Mayer, these conditions ares 1. Thin films must be coherent metal layers of high purity bounded by p a r a l l e l , smooth surfaces. 2. They must be available i n exactly known thicknesses up to a few thousand angstrom units, 3» A l l the metal layers must have the same purity and structurej with minimum number of l a t t i c e defects; and f i n a l l y 4. During a set of measurements on films of continuously varying thicknesses, changes of structure such as impurity consents , number and extent of l a t t i c e defects* i r r e v e r s i b l e changes of resistance and possibly thickness, must be excluded. 4 The present work includes electron microscopic studies, d.c. and high frequency (9.7 Gcs) r e s i s t i v i t y measurements, and microwave transmission c o e f f i c i e n t measurements. It has been attempted to f u l f i l the above conditions as follows: (a) The choice of gold, a noble metal, ensured that on exposure to open atmosphere the films would not be contaminated by oxidation or chemical reactions from atmospheric impurities. In addition, the highest purity gold available (better than 99.999%) has been used to prepare the films. A new technique of depositing gold films with nucleating s i l v e r layers on mica has been investigated as a method of obtaining uniform f i l m s . An attempt was made to improve the uniformity of structure of these films by annealing them at a suitable temperature. (b) The method of thickness measurement consisted of the combination of a radio frequency system employing quartz crystals and the di r e c t weighing technique, resulting i n a high degree of s e n s i t i v i t y and accuracy. The smallest measurable thickness i s of the order of half an angstrom for gold and s i l v e r . (c) Mica cleavage faces were chosen for f i l m supports ( i . e . , the substrates). These are found to be smooth down to molecular dimensions. A good quality muscovite mica can cleave (32) true to a single l a t t i c e over several square centimeters, so that films deposited onto such a surface can be expected to have a uniformity of structure over large areas. Gold films have been found useful i n a number of p r a c t i c a l applications because of their s t a b i l i t y and good e l e c t r i c a l 5 conductivity. One of the e a r l i e s t investigations i n this regard (33) was for de-icing a i r c r a f t windows. Gold films sandwiched between two layers of bismuth oxide> about 100 A thick, were deposited on glass. A resulting multilayer with a 100 A thick gold f i l m was found to have a transmission of 82$ for white l i g h t , and the f i l m had adequate current carrying capacity to allow s u f f i c i e n t heating of glass windows for de-icing. (35) Holland and Siddal coated glass with a Bi02/Au/BiO2 multilayer of 450/130/450 A thicknesses which gave a transmission of 73$ for green l i g h t and a r e f l e c t i o n of 74$ i n the near infra-red region. These glass panels have been employed as observation windows of an enclosure exposed to infra-red radiation. Currently there i s a great interest i n the possible use of gold films as bolometers for low l e v e l microwave power (36) measurement. In a recent paper, Schiffman, Young and L a r r i c k v ; have noted that of a l l the conductive materials tested by them for t h i n f i l m bolometers, gold was found to be the most sensitive because of i t s r e l a t i v e l y high temperature c o e f f i c i e n t of resistance (0.003/°C). This value i s about the same as for platinum; however, platinum melts at a much higher temperature (1774°C) than gold (l063°C). Another very important application of gold films arises from the b i o l o g i c a l hazards of microwave radiation. It has been • • (37) pr o v i s i o n a l l y accepted that the maximum safe power flux value over the entire microwave frequency region i s 10 mW/cm . Experiments have shown that the effect of radiation on the eye, i n contrast with most of the other organs, i s cumulative and can 6 result i n t o t a l blindness. The formation of b i l a t e r a l cataracts (38) has been reported v ' i n a technician who was exposed intermittantly j during a period of three days, to an average 2 2 radiation l e v e l at 1.7 - 3.4 Gc of 5mV/cm. ; and to 120 mV/cm. for a t o t a l of two hours. It has, therefore, been recommended that personnel working i n the proximity of microwave radiators (39) should use face shields or goggles. Egan has shown that a 750 A thick gold f i l m on glass has a microwave power trans-mission factor ranging from 0.04$ at 5.9 Gc. to 0.004$ at 18.8 Gc. This result suggests that good e l e c t r i c a l conductivities of the coatings i s essential for microwave shielding, and also that the shielding improves as the frequency increases. It may, at the same time, be noted that a 750 A thick gold f i l m has a very poor transmission of l i g h t as well. Special attention i s given i n this work to an 84 A thick gold f i l m which has good optical and microwave properties. The work to be presented i n this thesis has been divided into three d i s t i n c t topics: 1. Development of a more sensitive method of thickness determination, 2. Electron microscopic studies on gold films deposited, with and without the nucleating s i l v e r layers, on freshly cleaved rock-salt c r y s t a l s , and 3. E l e c t r i c a l measurements on gold films deposited, with and without the s i l v e r layers, on freshly cleaved mica. These measurements include r e s i s t i v i t y at d.c. and 9.7 Gc and micro-wave transmission c o e f f i c i e n t s . The effect of annealing on the conductivity of some separately prepared films was also investigated. 7 2. DEPOSITION METHODS 2.1 Introduction; The structure of a f i l m formed after condensation from the vapour state on a s o l i d substrate may be stable either i n the form of (a) mono-layers, or (b) agglomerates of atoms called islands. In (a) the atoms are uniformly dispersed on the substrate while i n (b) they are gathered up into globules or agglomerates with small s u r f a c e s . W h e t h e r or not the isla n d i c formation of the films i s possible depends upon the mobility of the freshly (41) condensed atoms on the substrate. Lennard-Jones has dealt with the p r i n c i p a l factors influencing the mobility of the condensed atoms. Appleyard^ 4^ has summarised and commented on Lennard-Jones 1 r e s u l t s . It can be seen from th e i r discussion and subsequent investigations by other workers that the f i l m structure depends mainly on the following factors; i . The metal f i l m - substrate pair, i i . The substrate temperature, i i i . Melting point of the deposit, i v . Minimum annealing temperature for the bulk metal, and v. Rate of deposition. i and i i . : The substrate may be amorphous l i k e glass or s i l i c a , or the cleavage plate of ionic crystals l i k e mica or (42) (43) rock-salt. Rhodin and Benjamin and Weaverv have shown that the condensation energies are dependent on i and i i . If the forces binding the atoms to the substrate, as determined by the condensation energies, are greater than or comparable to those between the condensed atoms, a thin metal layer may possess 8 / A A \ a smooth, continuous texture. Chandra and Scott^ ' investigated the influence of i & i i on the condensation c o e f f i c i e n t , a, which they defined as the r a t i o of the number of atoms condensing to the t o t a l number incident on the substrate. The condensation c o e f f i c i e n t decreases with increase i n the substrate temperature. On the other hand, micropolycrystalline films with high densities ,of imperfections tend to be formed at low temperatures on both amorphous and c r y s t a l l i n e substrates, whereas condensation upon c r y s t a l l i n e substrates at high temperatures tends to produce oriented single crystal f i l m s . These e p i t a x i a l l y grown single c r y s t a l films are mechanically strained due to di f f e r e n t thermal expansions of the f i l m and the substrate which occurs when the f i l m cools down from the deposition to room temperature. i i i , L e v i n s t e i n ^ 4 ^ has shown that films formed from high melting point metals consist of microcrystals with no s p e c i f i c orientation, while the films of low melting point metals produce large crystals oriented p r e f e r e n t i a l l y to the substrate, i . e . , showing e p i t a x i a l growth. i v . The minimum annealing temperature i s a measure of the r e l a t i v e ease with which internal imperfections anneal out even at low temperatures and thus the r e s i s t i v i t y of the f i l m approaches that of the bulk metal. This i s never true for thin films of heavy metals, e.g., gold, platinum and tungsten. v. L e v i n s t e i n ^ 4 ^ also showed that the tendency to form islands decreases as the rate of deposition increases. Presumably, with higher rates, more nuclei are formed i n i t i a l l y and these act as nucleating centers to ensure a fine grain size, resulting i n a smooth texture. 9 I n i t i a l condensation of gold, at room temperatures and on a l l substrates, takes the form of agglomerates, so that very th i n films have very high e l e c t r i c a l r e s i s t i v i t y . Wilkinson's (47) results showed that below a thickness of 100 A, the r e s i s t i v i t y of gold films on glass was much greater than the bulk metal and increased rapidly as the thickness decreased. The colour of gold films on mica, i n transmission* varies from pink (at about 25 A), through blue to dark green and the trans-mission of l i g h t i s rather low. A number of attempts have therefore been made to modify the substrate suitably i n order to obtain better adhesion and consequently decreased mobility of the condensed metal atoms: (a) Pashley^ 4 8^ predeposited a 1000 A thick s i l v e r f i l m by vacuum evaporation onto heated mica as substrate. The res u l t i n g s i l v e r surface was atomically smooth and had the same cr y s t a l l i n e orientation as the underlying mica. Subsequent deposition of gold on this'surface at normal temperatures produced a single crystal f i l m exhibiting e p i t a x i a l growth. (33 ) (b) Gillham, Preston, and Williams ' prepared about 100 A thick bismuth oxide and indium oxide layers on glass by reactive sputtering. The gold films subsequently sputtered on these layers were transparent and highly conducting. Later, these workers showed that the e l e c t r i c a l conductivity of gold films sandwiched between two 100 A thick Bi02 layers and annealed at a temperature of 400°C, approached that of the bulk metal even (35) at a gold f i l m thickness of 60 A. Holland and Siddal ' also obtained similar results from Bi02/Au/Bi02 multilayers. A l l these workers made gold and B i 0 9 films by cathodic sputtering 10 (34) and Ennos confirmed their findings by depositing the same (l°) films by h i g h vacuum evaporation, Chopra and Bobb °' deposited gold films on niobium oxide layers, which were obtained by heat oxidation i n a i r of thick (5000 A) Nb films. These gold films had e l e c t r i c a l conductivities i d e n t i c a l to those on BiO^^ and I n2°3" (17) (c) Chopra, Bobb, and Francombe y obtained e p i t a x i a l l y grown single c r y s t a l films of gold by cathodic sputtering on to mica heated to a temperature of 300°C. The above methods have the following disadvantages, with respect to the present research: i . Thin films of gold on thicker s i l v e r layers cannot be used for e l e c t r i c a l conductivity measurements, i i . Producing metal oxide coatings on substrates i s rather complex, and i i i . The method used i n this thesis for the accurate determination of small thicknesses requires a quartz crystal and i t s semiconductor o s c i l l a t o r c i r c u i t to be placed- inside the vacuum chamber i n the proximity of the substrate. Heating the substrate to obtain e p i t a x i a l growth of the f i l m would therefore present d i f f i c u l t positioning problems and could result i n damage to the c i r c u i t components. Also, the resonant frequency of the thickness monitoring quartz cr y s t a l i s highly sensitive to temperature variations. In t h i s work, the gold films prepared by a rapid rate of deposition (15 to 20 A / s e c ) , and those deposited on very thin nucleating layers show good conductivity even at a thickness as low as 25 A, and have none of the above disadvantages. 11 2.2 Film Deposition Technique; Substrates; The substrate material used was a variety of muscovite mica, commercially known as B r a z i l i a n rum ruby mica. It was obtained i n the form of plates of random sizes from which 1" wide (or less by a maximum of 0.002") and 3 to 4" long strips were cut. The broad sides of these strips were knife-edged on a belt sander to f a c i l i t a t e cleaving. O p t i c a l l y f l a t sheets could then be obtained i n a thickness range of 0.005 to 0.01 cm. Further cleaving was done by cutting a li n e across the width of this s t r i p with a razor blade. Bending the mica sheet at this cutting l i n e yielded substrates between 0.001 to 0.003 cm. thick. These substrates were inspected by holding the two j o i n t l y cleaved sides facing each other against a window and matching defects i n the two which show up as very fine h a i r - l i n e scratches. If the defects i n the two sheets matched, p a r t i c u l a r l y over the required central area of more than 1" by 0.5" i n size, then both the sheets were rejected. The surfaces for f i l m deposition must be free from scratches, p a r t i a l layers and finger p r i n t s . Any of these defects, especially the grease marks of a finger print show up conspicuously on metal deposition, since the condensation Of the many groups of mica existing naturally, muscovite and phlogopite v a r i e t i e s are the most useful. Muscovite mica cleaves more eas i l y over large areas and i s attacked only by hydro-f l u o r i c acid which i s used to etch the mica surfaces. It i s found i n pegmatite deposits with f e l s p a r j quartz and semi-precious stones. Over 70$ of the world's supply i s mined i n India and B r a z i l . Chemically, muscovite mica i s hydrated potassium aluminium s i l i c a t e ; H^KAl^(SiO^)^, and by weight contains 45.5$ s i l i c a , 37.5$ alumina, 12$ potash and 5$ water of c r y s t a l l i z a t i o n . The structure i s c r y s t a l l i n e with rather complex orientation, and dehydration occurs on heating above 600°C. At 10 Gc, ruby mica has a d i e l e c t r i c constant of 5.4 and a loss tangent of 0.0003. 12 does not occur on o i l y surfaces. A fresh substrate^ about 0.03 mm. or less i n thickness, was cleaved for the preparation of each f i l m to avoid any surface contamination by dust or absorbed moisture. No chemical cleaning of the substrates was required. Ideally, a substrate must be as thin as possible since the f i l m has to be placed transversely i n a waveguide for micro-wave measurements and so the r e f l e c t i o n s due to the substrates must be a minimum. However, very thin sheets (below 0.02 mm. t h i c k ) , free from defects over the requisite area, are extremely d i f f i c u l t to cleave. Even for a 0.03 mm. sheet, i t took, on an average, twenty cleavings to get one suitable surface. A compromise was arrived at and the r e f l e c t i o n c o e f f i c i e n t due to the substrate alone was f i r s t measured. Film configuration: The shape and size of the f i l m was chosen to be suitable for simultaneous d.c. and microwave sur-face resistance measurements. The internal dimensions of a standard X—band rectangular waveguide are 0.4"by 0.9" so that a rectangular f i l m of sides 0.5" x 1.0" was chosen. The con-fi g u r a t i o n of the deposited f i l m with the required substrate size i s shown i n F i g . 1, with important dimensions given. A, the f i l m i s deposited f i r s t . B,B are the heavy (over 5000 A thick) gold-plated s i l v e r or gold-plated copper contacts, and give d.c. connections to the f i l m . About a 100 A thick gold layer on s i l v e r or copper contacts i s s u f f i c i e n t to prevent tarnishing or oxidation of the l a t t e r at normal temperatures. Both the f i l m (A) and the contacts (B,B) were deposited 13 F i g . 1. Film Configuration by high vacuum evaporation. Substrate holder: The mica sheet was positioned inside the vacuum coating unit above the filament source containing the evaporant charges - either s i l v e r or gold, together with a freshly cleaved plate of rock-salt to obtain films for electron microscopy. The substrate holder was therefore designed accordingly. The substrate holder assembly i s shown i n F i g . 2. The aluminium block A contains a channel for holding the mica sheets. Aluminium st r i p s B,B' are placed i n the transverse channels (marked 1 and 2) on top of the substrate. The inside edges of B.B1 are 1" + 0.001" apart. The remainder of the substrate i s shadowed by mica sheets G and G 1. A shadowing s t r i p , E, i s used during contacts deposition and f i t s i n the corresponding channel as shown. Clamps C and C 1 are screwed onto their respective sides 14 FIG. 2= THE SUBSTRATE HOLDER 15 and cast nearly l / l 6 inch shadows on the longitudinal edges of the substrate. The rock—salt plate holder, H, made of a thi n aluminium sheet, i s fixed at a corner below C for making the electron micro-scopy samples. The 5° t i l t towards the source i s to ensure normal vapour incidence on the NaCl plate and thus a deposition i d e n t i c a l to that obtained on the mica substrate. Details of the base plate assembly? F i g . 3 shows the base-plate assembly. The supporting aluminium plate i s 10" i n diameter and g- inch thick and can be fixed inside the. vacuum chamber so that the filament to substrate distance i s 7.5"* The substrate holder, i s placed s l i g h t l y off—center on the base plate so that i t i s exactly above filament* The 10° i n c l i n a t i o n of the crystal seat on the crystal holder (0) ensures that the crystal and the substrate are equidistant from the source. The c r y s t a l holder is e l e c t r i c a l l y insulated from the base plate by two mica sheets (K). Vashers (M,M), separated by an insulating porcelain stud (L), provide connections to the crystals which are taken out to the monitor o s c i l l a t o r through enameled copper wires. The monitor o s c i l l a t o r chamber (F) i s made vacuum tight by the threaded cap (E) pressing onto the greased "0" ri n g . Other d e t a i l s are shown on F i g . 3. FIG. 3: DETAILS OF THE BASE PLATE ASSEMBLY SCALE: DOUBLE FULL SIZE - 7 . 5 " F R O M S O U R C E 17 F i g . 3i Details of the Base Plate Assembly-Legend: A Glass to metal seal* B Nine pin tube socket* C Card containing the monitor o s c i l l a t o r c i r c u i t . The arrangement of the various components i s shown i n de t a i l s with e l e c t r i c a l connections indicated by dotted lines » D High vacuum "0" ring* E Brass screwed cap for the monitor o s c i l l a t o r chamber* F Brass o s c i l l a t o r chamber, G 10 inch diameter aluminium base plate, H Substrate holder (Fig. 2) K Mica insulating plates for crystal holder, L Insulating porcelain stud',/. M Washers and connections to the c r y s t a l , N Copper spring contact for the c r y s t a l , 0 Brass cr y s t a l seat* P AT-cut 5 Mc quartz cr y s t a l plate, Connections from the o s c i l l a t o r are from the f i r s t four pins reading clockwise, looking from top: 1, Ground, connected to the base plate, 2. D.C. Input supply and r . f . output, 3,4. to crystal 18 Film deposition procedure? The films were made from 99*999$ gold and s i l v e r and the contacts from 99.999$ copper* or 99.9$ s i l v e r plated with 99.9$ pure gold* Two types of filaments were used for evaporation sources. 1. Baskets made out of 0.02" diameter tungsten wires-these were used to evaporate the nucleating s i l v e r layers and very thin films of gold. Evaporation from the basket required a current of about 10 amp. at approximately 3 volts a.c. 2. Molybdenum boats- these were shaped out of a 0.003" thick sheet which was cut into 0.2" wide and 1.5" long s t r i p s . The central part of a s t r i p was pressed to form a receptacle for the charge. Boats are more e f f i c i e n t than the wires because the evaporation i s limited to the upper region only instead of being omnidirectional. Thicker gold films (t ^> 20A) and contacts were evaporated from these boats. The filament current required was between 40 to 55 amperes at 3 to 4 volts a.c. The procedure for f i l m deposition was as follows: i . Suitable filaments were mounted on the rotatable filament mounting platform i n the work chamber of the high vacuum coating # unit. The quantities of gold and s i l v e r estimated to give the desired thicknesses were placed on the appropriate filaments. A freshly cleaved mica sheet was placed i n the substrate holder. F i g . 4 shows the substrate holder assembled for the deposition of thin films and F i g . 5 that for the deposition of contacts. The various parts are denoted according to F i g . 2. 7S The high vacuum coating system was an Edward's unit Model 12E6 modified to hold the rotatable filament platform. 19 FIG. 4 SUBSTRATE HOLDER ASSEMBLY FOR FILM DEPOSITION' FIG. 5 SUBSTRATE HOLDER ASSEMBLY FOR CONTACT EVAPORATION 20 ii« The base plate assembly was mounted inside the work chamber with the crystal and substrate holder facing the source. The e l e c t r i c a l connections were made to ensure that the thickness measuring system was functioning properly. i i i . The glass b e l l jar and i t s metal shield were placed i n position and the work chamber was evacuated to a pressure of -4 less than 10 mm. of Hg* The evacuation time necessary was about half an hour* i v . The filament was heated by gradually increasing the supply voltage, through a variac transformer. The commencement of deposition was indicated by the continual change i n the output frequency as read on a d i g i t a l counter. The frequency d r i f t stopped when the charge was completely evaporated. The shortest possible time was taken for depositing the films i n order to avoid excessive heating of the thickness monitoring crystal by the heat radiated from the filament* An average deposition rate between 15 to 20 A per second was observed. v. The contacts were deposited by mounting the substrate with f i l m as shown i n Fig* 5* Requisite quantities of s i l v e r or copper and gold were placed^, i n that order, on the molybdenum boats. A shield was placed on the crystal holder to avoid deposition on the c r y s t a l * The assembled base-plate was mounted at a distance of about 3" from the filament. F i r s t s i l v e r and then gold were deposited by the procedure outlined i n ( i i i ) above. The o s c i l l a t o r connections were not made during contact deposition. The amount of metal for contact deposition was considerably larger than for the f i l m deposition, and i n this case the completion of evaporation was indicated by the decrease i n the filament current and by the filament glowing brighter. The time taken for f i l m and contact deposition was from less than 1 to 30 seconds. 3 . THE MEASUREMENT OF FILM THICKNESS 2 2 3 o l Introduction: To make an accurate determination of fi l m thickness requires an exact knowledge of either the surface density, P , or the average physical height of the deposit* Pg i s defined as the ratio of mass to area of the deposit. If P^ i s the bulk density of the material, and i f i t can be assumed that the f i l m i s a s l i c e off the bulk material* then the f i l m thickness i s PJP^* However, the above assumption i s far from correct. Blois and R i e s s e r ^ ^ have shown that since the vacuum deposited films are invariably porous i n structure with irregular surfaces, the apparent density of the material i n the form of a thin f i l m , Pp, decreases from the bulk value P^  as the f i l m gets thinner. Thus the ra t i o fg/Pfc does not give the average height of the deposit. It i s , nevertheless, often convenient to express the thickness of a porous f i l m as i f i t was a compact layer; mainly because Pp i s d i f f i c u l t to determine as i t i s highly dependent on the mode of deposition of the f i l m . Thus, t = Ps/p^ m a v D e called an equivalent thickness. The height of the deposit can, i n contrast, be defined as the "metric thickness"*, t • r T m In the past, a wide variety of methods have been used for thickness measurement and those i n use pri o r to 1 9 5 2 are reviewed ( 5 1 ) by Greenland. Some of the methods commonly used are d i s -cussed below? 1 . The multiple beam interferometry technique developed ( 5 2 3 2 ) ( 5 3 ) by Tolansky v * ' and recently modified by Stern v ; i s undoubtedly the most elegant method of measuring the average 23 value of t within + 5 A. There i s , however, some uncertainty when the f i l m surfaces are excessively i r r e g u l a r , 2. The methods of X-ray absorption and X—ray interference (54 55) (5 were developed quite early. ' 7 Sauro, Fankuchen and ¥ainfan x have investigated these methods recently. 3. The radio-active tracer technique has been used to determine the surface densities of the deposits of radioactive bismuth, g o l d ^ 8 ^ and antimony. 4* Computation of thickness from e l e c t r i c a l resistance measurement has also been attempted^^ but i s seldom used as i t requires a prior standardization. (42) 5. Rhodin v 7 devised a direct weighing method to determine P . This i s a very convenient procedure i f the area of the fi l m deposit i s large. Modern microbalances can weigh with an accuracy of + 1 microgram with good r e p r o d u c i b i l i t y . This ( 1 7 1 8 ) method has been employed recently by other workers. ' Calibration of crystals i n the radio frequency method used i n this thesis i s done by direct weighing. 6. Radio-frequency methods Dudding^"^ found that when the inductance c o i l of an r . f . o s c i l l a t o r tuned c i r c u i t i s placed near an aluminium f i l m , the change i n the o s c i l l a t o r frequency, Sf^, i s a function of the thickness of the aluminium f i l m deposit. He was the f i r s t to use this method to control the thickness of the aluminium deposits. An alternative r . f . method i s to use a quartz c r y s t a l . It i s a standard procedure to make a f i n a l adjustment of the resonant frequency of a quartz crystal by depositing a suitable quantity of gold on one of the planar faces by high vacuum 24 evaporation.^ 2^ Lins and Oberg^ 3^ made a deposition control system using an AT-cut quartz crystal with i t s planar dimensions much larger than the thickness. They used the multiple-^beam interferometry technique to calibrate the crystals for obtaining the t vs. Sf relationship. The method used i n this work i s e s s e n t i a l l y d i f f e r e n t ( 6 3 ) from that of Lins and Oberg v ' i n two respects: 1» It i s shown experimentally that a l i n e a r relationship exists between the change i n the resonant frequency of a quartz c r y s t a l * §f r» and the mass of the p a r a s i t i c deposit, m, on one of the planar faces, over a small frequency range. The slope of the §f /m curve i s determined by dir e c t weighing. Therefore i t i s a method for measuring the equivalent rather than the metric thickness of the f i l m . 2. The reference o s c i l l a t o r i s the fixed frequency one using another quartz c r y s t a l , hence the frequency s t a b i l i t y i s considerably improved, being of the order of 1 part i n 10 over the period of an hour. This performance i s s u f f i c i e n t for accurate determination of <8f and the use of more elaborate r * methods* e.g., a GT-cut quartz crystal and a temperature controlled c r y s t a l chamber i n the reference o s c i l l a t o r , i s not e s s e n t i a l . A d i g i t a l frequency counter i s used for the beat of Sf i ; r difference between the beat frequencies before and after the deposition. frequency measurement. The value O f i s obtained by the A GT-cut c r y s t a l , without thermostatic control* has a frequency constant within 1 ppm. from 0 to 100° C. 25 3.2 Change i n the Resonant Frequency of a Quartz Crystal with  P a r a s i t i c Mass Deposition; The equivalent c i r c u i t of a piezoelectric resonator i s a combination of RLC i n series shunted by another capacitance.^ 4^ The inductance i n the c i r c u i t i s proportional to the crystal mass, M, and hence, other parameters remaining constant, the resonant frequency i s proportional to the reciprocal of JMV The change i n the resonant frequency, §f r» with change, i n M w i l l therefore be a nonlinear function of the deposited mass, m» However, i f the value of Of i s not very large, i . e . , Sf < ^ f , the relationship between ' f and m can be approximated by straight li n e Sf = -K.m ... ... 3.2.1 r The value of K can be determined experimentally by measuring the values of ' )r"r a n a the difference i n weights of the crystals before and after deposition. If the area of the planar face of the cr y s t a l i s A, the surface density of the deposited f i l m w i l l be given by; J f O - SL - —£— 3 2 2 and hence the equivalent thickness of the film,. g / ^, i s Sf r "b — r\ -tr k *©•• 3#2*3 / b The value of the denominator w i l l be constant for a certain material^ so that the equivalent thickness of the f i l m can be calculated from this very simple relationship. 3.3 Circuit t The block schematic diagram of the thickness monitoring arrangement i s shown i n Pig. 6(a) and the c i r c u i t i n F i g . 6(b). Some inconvenience was experienced i f the frequency of the cr y s t a l went much below 5 Mc/s, thereby going beyond the cut*-off frequency of the L-*C tuned c i r c u i t , so that the monitor c i r c u i t f a i l e d to o s c i l l a t e . Within reasonable l i m i t s , the o s c i l l a t i o n s could be i n i t i a t e d again by momentarily short-c i r c u i t i n g the c r y s t a l leads with a large (0.01 uF) capacitor. Otherwise a fresh monitoring crystal was required. In any case* the c i r c u i t was switched on before evacuating the chamber to ensure that the monitor o s c i l l a t o r was functioning. The used crystal was cleaned i n acqua—regia, to dissolve away a l l deposits, and coated with fresh contacts for further use. 3.4 Crystal Cooling; Piezoelectric crystals are temperature sensitive, though a properly mounted AT-cut quartz crystal has a zero temperature c o e f f i c i e n t of frequency i n the 0 to 20° C range. However, an unprotected c r y s t a l , as i n this case, i s often driven beyond this range. Inside the work chamber during evaporation, the crystal and any other surface receiving deposition i s automatically heated during the condensation process, both by the energy of the deposited atoms and by the thermal energy radiated from the (63) source. Lins and Oberg have studied these effects i n d e t a i l . The impact of molecules on the crystal face also affects the resonant frequency, but this effect i s only trans-i t o r y while the effect of temperature r i s e persists for a few minutes. It i s , therefore j. an important precaution to allow the cry s t a l to cool down and reach frequency s t a b i l i t y before taking 27 MONITOR OSCILLATOR* REFERENCE OSCILLATOR* BUFFER AMPLIFIER BUFFER AMPLIFIER MIXER AMPLIFIER DIGITAL FREQUENCY COUNTER** C.R.O. * Crystal controlled tunnel diode oscillators. * *Beckmen " E - P U T AND TIMER", Model 6146 (Accuracy of frequency measurement: 1 part in 10 7) FIG. 6 ( a ) : BLOCK - SCHEMATIC D I A G R A M OF THE THICKNESS MEASUR ING SYSTEM MONITOR OSC. CHAMBER TO CRYSTAL -01 Sl47K >1SK 30u.F S740K 3/1F.50V. > < -i^*S0V. 20H-f OUTPUT 2N44 > 4 - 7 K 5 0 M F > 4 7 , ( >4 7K FIG. 6(b): CIRCUIT DIAGRAM FOR THICKNESS MEASUREMENT 28 the f i n a l reading. The effect of v a r i a t i o n of room temperature during the deposition and crystal cooling periods has to be ignored. The crystal should be cooled i n vacuum because i t has been observed that the value of 8f increased on readmitting the a i r into the chamber. This was noticeably so when a small j thickness of metal was deposited, thereby indicating the absorption of gases i n the porous f i l m . The change i n §f due to absorbed gases was removed when the crystal was again placed i n high vacuum. The absorption of gases, as indicated by the above, was- much less for thicker films, thus confirming the general postulate that the f i l m structure becomes more compact and the density approaches that of the bulk material as the f i l m thickness i s increased, 3.5 Calibrations of Crystals; The 5 Mc/s quartz crystals were supplied with s i l v e r contacts deposited on one of the planar faces. Nine crystals i n t o t a l were used to obtain an average c a l i b r a t i o n . The s i l v e r contacts were dissolved away and replaced with gold contacts since s i l v e r i s susceptible to reaction from atmospheric impurities, especially sulphuric fumes; and this would affect the r e p r o d u c i b i l i t y i n weighing the c r y s t a l s . Before weighing, the crystals were cleaned i n repeatedly d i s t i l l e d acetone to remove any grease from them, and subsequently were always handled with tweezers. These crystals for c a l i b r a t i o n should not be placed i n p l a s t i c or glass containers since these become s t a t i c a l l y charged. The charge i s transferred to the crystals, thereby affecting their buoyancy and hence the accuracy of weighing. Another d i f f i c u l t y would arise due to fine dust p a r t i c l e s sticking to the c r y s t a l s . A 29 metal container was made to store the crystals to avoid these discrepencies. The crystals with gold contacts, and l a t e r with the d i f f e r e n t amounts of mass deposited, were weighed at the B r i t i s h Columbia Research Council. The weighing was done at least twice for each cry s t a l to check the r e p r o d u c i b i l i t y . Both gold and s i l v e r deposits were used for c a l i b r a t i o n . The relationship between S f r and m was recorded for each c r y s t a l . The ratio Sf /m was calculated i n each case and the average was taken to calculate the t vs. Sf relations for gold and s i l v e r . As expected from Eqn. 3«2*1> the values of Sf vs. m, plotted i n F i g , 7, l i e on a straight l i n e passing through the o r i g i n . Except for gold and other noble metals* the mass of the pure metal films w i l l change when they are exposed to the atmosphere. This i s due, mainly, to surface oxidation. To avoid this error> the s i l v e r deposits on crystals were protected with a thin layer of gold (Table I I ) . The following notations are used i n Tables I and I I . fjj = frequency of the reference c r y s t a l , f-^ = frequency of the thickness monitoring c r y s t a l , f^ = f - fj^ = difference frequency before deposition, f 2 = fjj - (fj^ - Sf ) = difference frequency after deposition, hence, -§f r = f^ - f ^ j the negative sign indicating the decrease i n f^ due to the p a r a s i t i c mass deposition, w-^  = weight of the crystal before metal deposition, w2 = weight of the cr y s t a l after metal deposition, m = w2 - w-^  = weight of the f i l m . In Table I i i f 3 = f R -30 _fM ~ S f r ( A g ) ~ Sf r(Au)] Sf r(Ag)= change i n the frequency with s i l v e r deposit, Sf r(Au)= change i n the frequency with gold deposit, m(Au)= change i n the weight due to gold deposit. according to the average values from Table I i m(Ag)= weight of the s i l v e r deposit, A = area of the crystal less the shadow area of the contact wire, = 2.54 2(0.75 2 - 0.0072) = 3.5826 sq. cm. 3.6 Discussions Ideally the reference o s c i l l a t o r should be a variable frequency one, sp that i t can be adjusted to a zero beat each time a fresh deposition i s commenced. Although t h i s f a c i l i t y i s provided i n commercial deposition thickness monitors, a considerable amount of frequency d r i f t i s present with a l l the L-C tuned variable frequency o s c i l l a t o r s . The use of another quartz cr y s t a l controlled o s c i l l a t o r as a reference avoids t h i s d i f f i c u l t y at the cost of some additional computations. The s t a b i l i t y obtained i n this manner i s of the order of 1 ppm over the period of an hour* with s c i n t i l l a t i o n s less than 0»5 cycles per ten seconds. This could probably be further improved by keeping the reference cr y s t a l i n a temperature controlled oven* but the additional trouble i s not considered necessary*' since the error i n obtaining §f r i s i n s i g n i f i c a n t . The performance of the c i r c u i t used and the resu l t i n g accuracy i n the measurement of §f r i s much better than any obtained previously ( The major inaccuracy i n the determination of the equi-valent thickness, t, by this method is due to the limitations of TABLE Ii Calibration of crystals with gold films f l f2 S f r w l v2 m S f r / m c/s c/s c/s gms. gms. a-gms• cycles/u-gm. 29,180 29,899 719 0.309422 0.309466 44 16.341 10,229 11*757 1528 0.314938 0.315039 101 15.129 -3,572 - 576 2996 0.313863 0.314062 199 15.055 18,936 22,126 3190 0.310133 0.310338 205 15.561 948 4,925 3997 0.311574 0.311826 252 15.861 13,602 21,691 8089 0.315323 0.315839 516 15.676 8,987 17,791 8804 0.315613 0.316193 580 15.179 34,041 46,075 12034 0.314989 0.315761 772 15.588 25,540 40,499 14959 0.315478 0.316434 956 15.468 . d f Average — = 15.67 cycles per microgram TABLE II % Calibration of crystals with s i l v e r and gold films f l f2 f3 Of r(Ag) b f r(Au) m(Au) v l W2 m(Ag) 6f r/m(Ag) c/s c/s c/s c/s c/s [i-gms • gms. gms. Li-gms. cycles/u-gm. 14,023 15,913 16,522 1,890 609 39 0.314948 0.315092 105 18.000* - 564 2,430 3,357 2,994 927 59 0,314446 0.314692 187 16.011 -3,097 • 1,144 2,073 4,241 929 59 0.308946 0.309276 271 15.649 25,222 31,727 32,150 6,505 423 27 0.314609 0.315057 421 15.451 -2,945 2,862 3,333 5,808 471 30 0.315086 0.315485 369 15.740 -10,930 -1,920 - 901 9,010 1,019 65 0.313466 0.314117 586 15.375 4,044 14,063 14,630 10,019 567 36 0.308052 0.308716 628 15.954 2,189 16,499 17,118 14,310 619 40 0.311803 0.312756 913 15.674 Average Of /m of a l l readings above except (*) = 15.6 cycles per microgram, t = 0.0925 x Sf for gold (pb = 19.32 gms/cc.) t = 0.1793 x Sf for s i l v e r (pb = 10.50 gms/cc.) 34 the weighing procedure. The raicrobalance used had an accuracy of + 1 ugm and the r e p r o d u c i b i l i t y i n the majority of the cases was within 1 ugm and never exceeded 4 ugm. Further, by making the c a l i b r a t i o n for comparably large masses of deposit, the over-a l l effects of the obtainable weighing accuracy and r e p r o d u c i b i l i t y become less c r i t i c a l . This also< eliminates the p o s s i b i l i t y of weighing error due to absorbed gases. From Tables I and II i t can be seen that the maximum deviation i n the Sf^./m ratio from the average value i s 4.5$, and i n most of the cases i s within 2$. The higher values of Sf^/m at small mass depositions i s possibly due to some deposits rubbing off from handling with the tweezers (very thin films have poor adhesion to the sub-strates)* Assuming that the procedure of extrapolating the S f r — m curve to the o r i g i n i s v a l i d , there i s t h e o r e t i c a l l y no lower l i m i t for the determination of the equivalent thicknesses of the films — keeping i n mind the basic differences i n defining the equivalent and the metric thicknesses. This system of determining t makes i t possible, for the f i r s t time, to work with known extremely small thicknesses. This i s obviously an i n d i r e c t method and i t i s assumed (44) that the condensation c o e f f i c i e n t s , of Chandra and Scott, are the same for various metals on the quartz cr y s t a l and other substrates* One essential precaution i s to plan the positioning of the monitoring crystal and the substrate during deposition such that both receive an equal amount of deposit. The geometrical scheme w i l l also be dependent upon the Evaporation character-(62) i s t i c s of the sources. Besides weighing, the accuracy of 3 5 the thickness values obtained depends upon these two conditions* The average of 2 to 5% accuracy of this method compares favourably with any other method available so far, e.g., the interferometric techniques which give an accuracy of + 5 A but are limited to a lower thickness of about 50 A at the best. An observation of second-by-second rate of deposition i s possible by this method using a recorder i n conjunction with the d i g i t a l counter. Some qualitative assessment of the porosity of a f i l m of a certain thickness may also be possible by noting the amount of absorbed gases. 36 4. ELECTRON MICROSCOPIC STUDIES ON THIN FILMS 4.1 Films for Electron Microscopys Mica substrates thicker than about 1000 A are unsuitable for transmission electron microscopy, as they are then opaque to the electron beam. The films were, therefore, deposited on freshly cleaved plates of rock-salt crystals (NaCl) from which they could be detached by dissolving away the s a l t . Thin films of the order of about 100 A are not self-supporting and there are (65 ) numerous standard procedures of preparing f i l m support f o i l s . In the present work, gold films on NaCl were supported by evaporated carbon f o i l s . On one face of the NaCl crystal of (OOl) orientation* l i n e p a r a l l e l to the sides were scribed with a needle point such that about 0.05" to 0.1" thick and at least 0.5"-square wafers could (66) be obtained. The procedure outlined by Gilman was used for cleaving. The films of gold and gold-on-silver were deposited by the procedure of Section 2.2, while the supporting carbon f o i l were deposited as described i n the following section. The composite f i l m on the NaCl plate was scored into small (about 2mm,) squares with a razor blade. On gradually dipping the NaCl plate i n water, the f i l m floated off due to surface tension. The specimen films were then mounted on fine mesh copper grids and dried out for microscopy. 4.2 Carbon Evaporations The carbon supporting f o i l s were (67) deposited by Bradley's method, ' with some sl i g h t modifications i n the filament arrangement. The apparatus for carbon evaporation and positioning of the substrate i s shown i n F i g . 8. The method of carbon evaporation i s to pass an alternating 37 ALUMINIUM DISC MICA SHEET WITH APERTURES TO SUPPORT ROCK SALT PLATES (WITH FILMS F A C I N G THE SOURCE) SCREW TO HOLD CARBON ROD-BRASS COLLAR-POINT C O N T A C T C A R B O N RODS /-BRASS COLLAR SPRING COPPER C O N T A C T PLATES' ' \ 9 0 o / ^ SCREW TO ADJUST-SPRING TENSION t a I B 5 B SECTION THROUGH 'A -A ' FIG.8: APPARATUS FOR CARBON EVAPORATION 38 current of between 20 to 30 amperes from a variable voltage supply^ through the pointed carbon rods which are pressed together l i g h t l y so that they are not parted during evaporation* Intense heating i n the region of the point contacts occurs, -4 thereby evaporating carbon* A vacuum better than 10 mm of Hg i s required. A glass dish containing phosphorous pentoxide i s placed iiiside the chamber to eliminate moisture. About 300 A carbon films were found to have adequate mechanical strength to support themselves and the gold films on the specimen holders. 4*3 Electron Microscopy Resultss Electron d i f f r a c t i o n patterns and transmission electron micrographs (magnification 20K and 100K) are shown i n Plates I to VI. The parts numbered (a) are the d i f f r a c t i o n diagrams which are interpreted as follows 8 Plate Is The f i l m was gold without a s i l v e r layer. ii i s a Debye-Scherrer type of d i f f r a c t i o n pattern con-taining numerous sharp, concentric rings. The f i l m i s therefore p o l y c r y s t a l l i n e with random orientation of the micro-crystals* Plates Il-VIg The films were gold on s i l v e r layers. A l l the rings i n Plate I(a) are present, but the ring i n t e n s i t i e s vary around the ring circumference i n a symmetrical pattern. This indicates preferred orientation and a more texturized growth than that for the gold f i l m alone. None of the films examined had a single c r y s t a l l i n e structure. Other d e t a i l s for each f i l m are given with the corresponding plate. l ( a ) ; D i f f r a c t i o n Pattern 1(b); x 20,000 l(b) & ( c ) s Transmission micrographs (dark f i e l d ) . 1(c); x 100,000 40 PLATE II t A g = 3.95 A t. = 98.5 A Au ^ = 8 . 3 5 Average grain size = 800 A 11(a): D i f f r a c t i o n Pattern IP 11(a) & (b): Transmission micrographs. 1 K b ) x 20,000 1 1(c): x 100,000 41 PLATE III t. = 6.95 A t A u = 122.5 A Pf Average grain size = 1000 A Note: The value of p{/ph for the f i l m i n the next plate (IV) i s 7.22 I l l ( a ) : D i f f r a c t i o n Pattern. I I l ( c ) : x 100,000 if' PLATE IV t A g = 11.7 A t A u = 117.4 A EL -pb _ 7 , 2 2 Average grain size = 300 A. Note: The value of P f / P b foj a 108.4 A gold f i l m on t A = T,w \ TW^* -O ^ e Ag IV(a): D i f f r a c t i o n Pattern, 3.23 A i s 6.84 (Table I I I ) . IV(b): x 20,000 IV(b) & (c): Transmission micrographs. IV(c): x 100,000 43 2 . 7 5 ) The structure, although texturized (a) i s excessively-granular. The two p a r a l l e l lines starting from the bottom right hand corner i n (b) are probably the cleavage steps on the rock-salt c r y s t a l . V(b): x 20,000 V(b) & (c): Transmission micrographs. V(c): x 100,000 44 PLATE VI t . = 24.4 A Ag t A u = 109.2 A ft = 10.92 Average grain size = 330 A. Note: The value of Pf/Pl for a 108.4 A gold f i l m on t = 3.25 A (see Table III) i s 2.07, The structure i s s t i l l texturized (a), and some change i n the structure i s indicated as the fourth ring from the center (Plate I (a)) i s missing and the f i r s t ring i s faded. VI ( a ) : D i f f r a c t i o n Pattern. Vl(b) & (c): Transmission micrographs. VI (c): x 100,000 45 5. D.C. RESISTIVITY MEASUREMENTS ON THIN FILMS (2) 5,1 D.C. R e s i s t i v i t y of Thin Films; Thomson was the f i r s t to propose a formula expressing the r e s i s t i v i t y of a thin f i l m as a function of f i l m thickness. Since then, Planck* L o v e l l , ^ ^ Appleyard and L o v e l l , ^ ^ Fuchs,^^ and W e a l e ^ ^ have proposed other formulas. Of these, Fuchs' rigorous analysis and a subsequent review by Sondheimer(l^) a r e -taken as a basis for comparing experimental results by most of the workers. According to the modern theory of conduction i n metals i n i t i a t e d by B l o c h ^ ^ i n 1928. the e l e c t r i c a l resistance of a macroscopic sample i s separable into two parts; an ideal part of the r e s i s t i v i t y , p^t which i s strongly dependent upon the amplitude of the thermal motion of the ions; that i s , dependent i n a reversible way upon temperature, and p^, the so call e d residual resistance dependent upon the l a t t i c e defects but independent of tempjerature insofar as l a t t i c e defects are unaffected by temperature.' (A phenomenon occurring very often i n t h i n films i s the structure change with temperature resulting i n an i r r e v e r s i b l e change i n resistance). This gives the Matthiessen rule for bulk r e s i s t i v i t y , p ^ , as Pb = Pi + P r .. 5.1.1 In the above, the boundaries of the actual metal have been omitted which can be considered as a form of imperfection of a metal l a t t i c e . Another component i n the t o t a l r e s i s t i v i t y of metals i s therefore due to the scattering of the conduction electrons by the boundaries of a metal. This component becomes more s i g n i f i c a n t 46 compared with the other l a t t i c e defects as the thickness of a metal f i l m becomes small. This along with the impurities* i f any* present i n a th i n f i l m give the modified expression for the r e s i s t i v i t y of a t h i n f i l m * p^, as Pf = Pi + Pr + Pirn, + Pt *•* 5 - U 2 where p^m a n < i p^ a-re "the r e s i s t i v i t y components due to impurity and thickness? respectively. If the boundaries of the f i l m scatter the electrons d i f f u s e l y , i . e . * i n e l a s t i c a l l y * the mean free path of the con-duction electrons i s a r t i f i c i a l l y l i m i t e d . As a consequence, the increase i n r e s i s t i v i t y above the bulk value i s a function of the f i l m thickness* On the other hand, i f the electrons are specularly ( i . e . . e l a s t i c a l l y ) reflected from the f i l m boundaries, the r e s i s t i v i t y Of the f i l m remains the same as that of the (6) corresponding bulk metal, p^ i s given by a complicated function which i n the special case of f i l m thicknesses less than the mean free path of the conduction electrons, X, reduces to: p f,-= p b 1 + | | (1 - p)~j . . . ... 5.1.3 where p i s the scattering c o e f f i c i e n t , being unity for perfect specular and zero for perfect diffuse scattering. (47) Wilkinson^ has taken the value of X for gold as 970 A, (18) while Chopra and Bobb have obtained X as 400 A from the asymptotic value of conductivity for very thick f i l m s . This l a t t e r value of X also corresponds to p = 0.6 i n Eqn. 5.1.3, i . e . , a 60$ specular scattering. 47 5.2 Measurement of d*c. R e s i s t i v i t y : To determine the r e s i s t i v i t y of a thin f i l m , consider a square of the thin f i l m i n Pig. 9, with the measuring electrodes at the two opposite sides. If o~ i s the conductivity, A the area and b the side of the squaref the d.c. resistance of this f i l m , R, , i s F i g . 9. Surface Resistance of a Thin Film where t i s the f i l m thickness. For a square, this resistance value i s independent of the size of the square and f o r t h i s reason the resistance of such a f i l m , i . e . , the surface r e s i s t i v i t y , i s expressed as a certain number of ohms per square. The films used i n the present work were 1 inch wide with the evaporated d.c. contacts 0.5" apart, giving two squares i n p a r a l l e l . The surface r e s i s t i v i t y of the f i l m i s then equal to R 'dc ~ cr x A ~o~t 5.2.1 , 48 twice the resistance measured between two contact electrodes. Prom t h i s , P_p = t x B^ c *. • . . . 5.2*2 The bulk r e s i s t i v i t y of gold i s 2.44 x 10~^ ohm-cm. Hence, i pj- = 2^44 x 1 0 = ^ b ^ f *•* \ 5*3.3 The r e l a t i o n P f / p ^ ( ° r c j ^ f ^ ^s c a l ± e d * n e r e l a t i v e r e s i s t i v i t y of the th i n films, i . e . , the r e s i s t i v i t y r e l a t i v e to that pf the bulk metal. The d.c* resistance of the films was measured by a GR Impedance bridge type 1650-A using i t s internal 6 v o l t and an external 50 v o l t d*c* supply (the l a t t e r for increased n u l l detector s e n s i t i v i t y ) with an accuracy of + 1$. The f i l m resistance values obtained using both these supplies were found to be the same, indicating that a l l the films i n the conducting range obeyed Ohm's law and were, therefore, cohesive* 5.3 Types of Films Studied; The following three types of films were used for d.c. measure-ments: 1. Rapidly Deposited gold f i l m s : The average rate of deposition of the f i l m s , as indicated by the per second change i n the readout frequency on the d i g i t a l counter, (Section 3*6) was about 15 A/sec* 2* Gold films on nucleating s i l v e r layers: The thick-ness of the nucleating s i l v e r layers was between 1 to 25 A* and that of the gold films from 9 to 300 A* Gold films below 49 24 A were nonconducting and are therefore omitted* For the calcu l a t i o n of r e s i s t i v i t y from Eqn. 3.2.2* the thicknesses of the nucleating s i l v e r layers were ignored as these were seldom above 5$ of the t o t a l thickness* Also* s i l v e r layers (62) below about 50A are not conducting* ' The eff e c t of nucleating layer thicknesses on the conductivity was studied. # 3. Annealed gold films? The effect of annealing on the gold films with and without the nucleating s i l v e r layers was studied for a small number of fi l m s . I n i t i a l l y , an attempt was made to anneal the films with the contacts deposited and after making the i n i t i a l resistance measurements. This procedure invariably resulted i n the destruction of the films and the gold plated copper contacts were oxidised on the upper surface. The f i l m damage was as a consequence of the different thermal expansions of the f i l m * contacts, and the substrate. It was therefore necessary to make a fresh set of films which was done by taking a long s t r i p of mica for f i l m deposition. Each f i l m thus obtained was divided into two or three pieces. One of these pieces was kept unannealed while the rest were heated to di f f e r e n t temperatures. The contacts were deposited after annealing* Only one gold f i l m was tested for annealing since i t i s a well known fact that the conductivity of gold films alone decreases i r r e r v e r s i b l y with temperature r i s e . The effect of annealing was taken only as an indicat i o n of the general behaviour and to f i n d out an approximate value of the optimum annealing temperature. * The word annealing i s used as the equivalent of heat .treatment i n the broad sense* for brevity, as by Gillham et al,'33) 50 5*4 Re s u i t s ; The results of d.c. measurements are plotted i n F i g , 10* The l a t e s t available results on gold (17) films by Chopra, Bobb and Francombe ' are also reproduced i n F i g . 10, taking t h e i r two extreme res u l t s , v i z , the slowly evaporated gold films on mica at room temperature (rate of deposition about 0.3 to 1 A per second) and single c r y s t a l gold films obtained by sputtering onto heated mica (rate of deposition about 1 to 4 A per second). These two methods gave minimum and maximum re l a t i v e conductivities, respectively. These results are chosen because of two s i m i l a r i t i e s i n the films made by Chopra et a l . and those i n the present work. These are; (a) mica i s used as the substrate material and the f i l m size i s 0.5" x 1"; and (b) the d i r e c t weighing technique i s employed for thickness measurements* thus the same d e f i n i t i o n for thickness ( i . e . , the equivalent thickness) i s applicable. Table III shows the effect of annealing the films and also that of the increasing nucleating layer thicknesses on conductivity. In f i g . 11, r e l a t i v e r e s i s t i v i t y of a set of films i s plotted against the thickness of the nucleating s i l v e r layers. The dotted curves show the anticipated effect of t ^ on P f / p ^ ^ 0 T "the films with nearly i d e n t i c a l thicknesses of gold. 5.5 Discussion? In F i g . 10, r e l a t i v e r e s i s t i v i t y , p f/p-^i n a s been plotted against f i l m thickness, t, the curves indicating minimum r e s i s t i v i t y values. It i s believed that the general scatter of r e s i s t i v i t y points i s caused by the following factors: (52) (a) The cleavages faces of mica always have some defects and are i d e n t i c a l only i f they are from the same cleaving. H i l l s 51 52 TABLE III EFFECT OF ANNEALING ON Au and Au-on-Ag FILMS t(A) Heating R^c(ohms/sq) Pf Pb Temp« ° c Time min. Ag Au A | 13 (Annealed)i/ (Unannealed) A B 1.60 90.0 350 60 7.44 - 2.75 -3.95 98.5 300 600 15 60 10.26 47.60 20.70 tt 4.14 18.65 8.35 . tt 5.55 129.5 350 60 6.30 7.64 3.34 4.05 6.95 122.5 300 30 10.12 13.20 5.08 6*53 9.70 41.6 350 15 2.29K 2.52K 381.00 430.00 14.40 90.0 350 30 36.50 48.40 13.47 17.86 17.10 106.5 350 45 17.70 30.60 7.72 13*35 24.40 109.2 3 50 45 24.40 45.60 10.90 20.40 3.23 108.4 3 50 450* 60 60 4.66 7.83 6.84 » 2.06 3.47 3.03 .tt 0 132.0 350 450* 60 60 -44.40 comple 4.64 tely des 23.90 troyedi • 2.51 *These films were placed i n the oven together for annealing. 102 10 FIG. 11: EFFECT OF t A g ON RELATIVE RESISTIVITY \ I / v «\o / / •/ / / X \ ? / / / J I J L 53 0 2 4 6 8 10 20 26 A" »Ag > 54 and dales* and i n some cases small p i t s with a diameter of 1/200 mm. or less* make the surface resemble a wrinkled sheet of paper* These defects increase the overall area* and have a small e f f e c t on r e s i s t i v i t y * (b) The existence of cleavage steps on mica surface presents a more serious problem. These cleavage steps are i n f a i r l y straight lines of varying lengths. The areas between the cleavage lines have good optical uniformity. According to B r a g g , t h e height of the cleavage steps i s a simple multiple of 10 A ( i . e . , half the spacing between K-atoms)» This height i s often comparable i n magnitude to the f i l m thickness, A f i l m deposited on a step would appear as shown i n Pig. 12(a). No deposition can occur on the face BC which i s (62) p a r a l l e l to the di r e c t i o n of the metal vapour beam. ' The line running along BC, therefore, represents a weak spot, or, when the cleavage step i s higher than the f i l m thickness, a discontinuity. The d i r e c t i o n of the cleavage li n e affects the measured value of d«c. resistance as shown i n F i g . 12(b). (c) The very high rate of deposition cannot be con-t r o l l e d precisely because the entire process i s completed i n a few seconds. The deposition rate influences the structure of the films as well as the amount of occluded gases. (d) A high rate of deposition takes place at a high vapour beam intensity* The high beam int e n s i t y ensures a more uniform dispersion of the nuclei but c o n f l i c t s with the require-ment of obtaining a very th i n nucleating layer ( l to 10 A). In view of the effect of above mentioned uncontrollable factors on the measured r e s i s t i v i t y , films showing the least C L E A V A G E S T E P F I L M D I R E C T I O N O F M E T A L V A P O U R A T O M S FIG. 12(a) : FILM DEPOSITION OVER A CLEAVAGE STEP T H I C K C O N T A C T S ( 5 0 0 0 A ) CLEAVAGE LINES ON MICA SUSTRATE (X) R U N N I N G P A R A L L E L T O C O N T A C T S , C A U S I N G H I G H E R M E A S U R E D R E S I S T A N C E ( ? ) R U N N I N G N O R M A L T O C O N T A C T S , D O N O T M A T E R I A L I X A L T E R T H E M E A S U R E D R E S I S T A N C E V A L U E . FIG. 12(b) : A N D u i r e s i s t i v i t y " were chosen for interpretation. The deposition of metal atoms on the cleavage planes of ionic crystals d i f f e r s s i g n i f i c a n t l y from that on amorphous, f l a t surfaces l i k e glass. The l a t t e r serves only as a f l a t support and there i s no interaction between the atoms of the substrate and the neutral deposited atoms. In the cleavage surfaces of ionic c r y s t a l s , there i s , i n addition, a force of at t r a c t i o n between ions of the substrate and the deposited atoms, especially when the formation of the f i r s t atomic layer i s considered. ,7^~ ^  Benjamin and Weaver^"^ have proposed a map of the potential energy surfaces for : NaCl after the pattern given by Rhodin.^ 2^ It i s suggested that there are regularly spaced "potentials wells", depending upon the orientation of atoms on the cleavage face, which act as the most l i k e l y sites for adsorption of an oncoming neutral atom, A deposited atom has a f i n i t e l i f e time i n a potential well depending on the substrate temperature. On the basis of this potential well concept* Benjamin and Weaver determined the adhesive energies for di f f e r e n t metal-substrate pairs* The adhesive energies were found to be far more for ionic c r y s t a l faces than for glass. This may be the main reason why the rapidly evaporated gold films on mica display a better con-d u c t i v i t y , as shown i n Pig. 10, than any obtained so far on other unmodified substrates. (72) The work of Basset v ' i s s i g n i f i c a n t for the insight i t gives into the l i k e l y nucleatipn centers for the deposited atoms* He showed that when gold films were deposited on a rock-salt cleavage surface* there was a preferential 57 nucleation on the edges of the steps on the substrate; even when the steps were monatomic i n height. He concluded that imperfections, impurities and cleavage steps, etc., on an otherwise smooth surface create nucleation centers which determine the further growth and hence the structure of the f i l m . In this regard, the comparison of data on atomically smooth mica and atomically rough oxide surfaces i s also i n s t r u c t i v e . The oncoming atoms have a larger adhesion to the l a t t e r and therefore less surface mobility, thus (73) preventing isla n d formation, Bannet v ' considers that there i s always some excess bismuth on BiC^ films and the gold atoms nucleate upon bismuth surface atoms instead of the larger nuclei being formed by surface d i f f u s i o n of gold atoms. One other possible explanation could be to consider the B i 0 2 surface f u l l of submicroscopic holes acting as nucleation centers for gold atoms* In either of the cases, the results are a smaller thickness at which a continuous layer w i l l form, and a greater nucleation density on oxide surfaces. The presence of excess bismuth on B i 0 2 films i s also indicated i n the papers of Gillham et a l , ^ " ^ and E n n o s . ^ ^ A technique of making zinc films also indicates that the atoms of a di f f e r e n t metal are preferred nucleation centers. Zn films do not form unless the substrate temperature i s of the order of -90°C or unless the vapour beam intensity i s very high. An alternative method i s to predeposit a molecular layer of a high melting point metal, e.g., s i l v e r , copper or t i n , on the substrate. This provides the necessary nuclei on which Zn films are obtained at normal temperatures. 58 The e f f e c t of the nucleating s i l v e r layer thickness can be explained by the theory pertaining to the mechanism of islan d formation,.(74) ^his i s as followss At thicknesses below about 1 A, the surface density of the s i l v e r atoms ( i . e . . the number of atoms per unit area) i s small. The atoms are uniformly dispersed a l l over the surface, presumably i n the potential wells on the ionic crystal surface* Further deposition increases the surface density. Then a stage i s reached when the neighbouring atoms or th e i r agglomerates are so close that the force of a t t r a c t i o n between them exceeds their adhesive energy to the substrate. This i n i t i a t e s the island formation due to surface migration of the atoms. The nuclei density i s then continuously lowered as more deposition takes place — the miclei getting larger i n s i z e . F i n a l l y a point i s reached at which a con-tinuous layer i s formed, the islands having merged together. For example, s i l v e r films are continuous and conducting only (62) above a thickness of about 50 A. The above discussion shows that there i s a c r i t i c a l thickness, that i s . an optimum value of nucleating layer thickness* when the nuclei density on the substrate i s the greatest. This i s just before the surface migration eommenees. Gold films when deposited on such layers should, then, show the maximum conductivity* Although either gold or s i l v e r films above the c r i t i c a l thickness w i l l show island formation* the presence of the d i f f e r e n t metal i n h i b i t s surface migration. F i g . 11 shows the e f f e c t of t. , the s i l v e r nucleating layer thickness, on P f / P ^ » The points are scattered about the expected curves since exactly i d e n t i c a l f i l m thicknesses are extremely d i f f i c u l t to obtain. A c r i t i c a l value of t ^ i s indicated to be between 5 to 7*5 A. Even a 25 A thick gold f i l m deposited on such a " c r i t i c a l s i l v e r layer" was found conducting, ( 7A \ Bassett. Menter and Pashley ' showed that an i n i t i a l formation of an a l l o y skin i s possible at low temperatures because the a r r i v i n g metal atoms have s u f f i c i e n t energy to diffuse a certain distance into the sub-layer. The a l l o y skin may have a better adhesion to the substrate so that this i s a competing factor to that of surface mobolity. Accepting t h i s premise, the f i l m structure w i l l depend on the nucleating-to-main metal p a i r s . Epitaxy may also be possible at normal temperatures i f the nucleating atoms are considered as uni*-formly dispersed i n the "potential wells" on an ionic c r y s t a l surface* The tendency towards texturised structure, as shown by the d i f f r a c t i o n patterns i n Plates II to VI, can be for this reason, (33) Gillham et a l * 7 showed that the transmission colour of thin, highly conducting gold films would be pale yellow* In the present work, this fact i s substantiated i n cases of films below about 80 A i n thickness, A rough estimation of the r e l a t i v e conductivity of gold films can be made from t h e i r transmission colour* A prominent b l u i s h or pink t a i n t , e*g** the ones obtained on heating a pure gold f i l m to various temp*Bratures, indicate poor conductivity* A pale yellow or yellowish-green colour: means a better conductivity. Thus the 60 l i g h t transmission of the films improves with conductivity, A comparison of annealing properties of gold films with "and without s i l v e r nucleating layers y i e l d s very encouraging r e s u l t s . One gold f i l m without a nucleating layer showed a five f o l d increase i n the Value of Pf/P^ when heated for one hour at 350" C. It was completely destroyed when heated to 450 0* On the other hand, the conductivity of the gold-on-silver films showed an improvement on heating. An optimum annealing temperature was found to be about 350° C. On heating beyond th i s temperature* the conductivity decreased. However* at 450° C, the value of Pf/P^ v a s o n ± y twice that at 350° C, After having been heated at 600° G f o r one hour* p^/p^ of one Ag-Au f i l m increased by a factor of f i v e over that of the same annealed at the optimum temperature. Nevertheless, the s i g n i -f i c a n t f a c t remains that the f i l m was s t i l l conducting. Heating beyond 600° C was obviously not possible as the substrate dehydration temperature i s then reached. It can therefore be concluded that the gold f i l m alone* even when very rapidly deposited* has poor cohesion and mechanical strength* because more structural defects appear as a result of d i f f e r e n t thermal expansions of the substrate and f i l m on heating. The improvement of the more adhesive Ag—Au films on annealing i s believed to b# for two reasons* (a) Removal of struetxiral defects* Gillham et a l * v , (18) Chopra et a l . * working with gold films on BiO,, and NbO respectively, showed that annealing produced some change i n surface topography* besides smoothing out some structural disorders; i n a way which resulted i n a marked drop i n the diffuse electron r e f l e c t i o n at the f i l m boundaries* •5fr (b) Expulsion of occluded and adsorbed gases s W i l k i n s o n ^ 7 ^ and W e i t z e n k a m p ^ f o u n d that adsorbed nitrogen had no e f f e c t on the conductivity of the films* In addition* Weitzenkamp made the following observationss ( i) The con-du c t i v i t y of gold films decreased on exposure to the atmosphere* as also to oxygen* After an equilibrium had been reached i n about 30 hours* replacing the f i l m i n a vacuum of -4 10 mm* of Hg did not improve the conductivity, indicating that the process of gas adsorption was i r r e v e r s i b l e i n the pressure ranges tested; and ( i i ) The adsorption of water vapour increased the conductivity by a factor as great as 100* It may be noted that Weitzenkamp purposely made his films with a discrete i s l a n d i c structure*, Also, one re s u l t of the present work (Section 2*3) indicates that the phenomenon of gas adsorption i s reversible^, at least to a great extents Stresses can, nevertheless, be generated by enclosed gas and impurity (65) atoms* It i s * therefore ? a p o s s i b i l i t y that i n i t i a l adsorption of gases produces certain structural changes i n the f i l m which are i r r e v e r s i b l e a f t e r the gases are driven out on replacing the f i l m i n vacuum* The reduction of porosity of the f i l m on annealing would reduce the amount of adsorbed gases considerably* ; Occlusion i s said to occur when residual gas molecules xn vacuum strike the substrate during deposition and are trapped within the f i l m by the oncoming metal atoms. Adsorption occurs due to the general porous structure of the f i l m . The amount of occluded gases i s . f a r greater i n sputtered films than i n the evaporated ones*(34) 6. MICROWAVE MEASUREMENTS 62 6.1 Introduction! Microwave surface r e s i s t i v i t y measurements have been concerned c h i e f l y with super-conductivity . ( 7 5 - 7 T ) At 9.4 Gc., Khaikin^ 7 5) was the f i r s t to make measurements on 400-1200 A t i n films on glass. The films formed one end wall of a resonant cavity made of lead. At millimeter (3 to 6 mm) and infrared (0.1 to 0.75 mm) wavelengths, Glover and Tinkham^ 2^ made superconductivity measurements on lead and t i n films by determining their transmission c o e f f i c i e n t s . To use either of the above methods at X-band frequencies requires large f i l m areas which are d i f f i c u l t to obtain. Microwave measurements presently available on gold films are confined to (39) those given by Egan, ' as reported i n Chapter 1, who determined transmission c o e f f i c i e n t s of a 750 A gold f i l m at various frequencies. In th i s section of the thesis, the microwave trans-mission c o e f f i c i e n t s of gold films of various thicknesses are investigated. As i n Egan's work, th i s i s with the view of a possible application of these films for microwave shielding. (78) Clark v ' was the f i r s t to investigate the waveguide method of surface r e s i s t i v i t y measurements. At 9.2 and 26.5 G c , he made measurements on Bi and Sb films on glass. The films were placed transversely i n a waveguide and terminated with a short c i r c u i t at a quarter wavelength. His results showed that the microwave surface resistance, R , of the Bi films was 0.8 times R, • while * u 1 dc* no such correlation existed for the Sb films. Clark's method has been used i n the present work for the measurement of R at 9.7 Gc. The only difference was i n the use a 63 of thin (0.01 to 0.03 mm.) mica sheets instead of glass as substrates. High VSWR measurements were made by Nunn's method, to be described. 6.2 Design of the Film Holder: The f i l m holder for waveguide F i g , 13. It i s designed so as to hold the 1" wide substrate i n channel A such that the f i l m i s positioned transversely i n the waveguide, and the contacts to the f i l m for d.c. measurements can be brought out through B. The f i l m can be insulated from the waveguide during i t s use as a bolometer element. There i s enough capacitive coupling between the contact s t r i p s and the broad waveguide walls for the f i l m to act e s s e n t i a l l y as a shunt con-ductance across the waveguide. Reflection c o e f f i c i e n t of the f i l m holder: A thin sheet of d i e l e c t r i c placed across the waveguide introduces a small r e f l e c t i o n that does not change rapidly with wavelength. The voltage standing wave r a t i o , s^, introduced into the matched (49) guide by this low loss d i e l e c t r i c sheet i s given by: d i e l e c t r i c constant, X i s the free space wavelength, and X i s c the cut-off wavelength of the a i r f i l l e d waveguide. Mica has d i e l e c t r i c constant e = 5.4 at 10 Gc, and the maximum measurements i s shown i n where d i s the thickness of the d i e l e c t r i c sheet, e i s i t s SECTION ALONG B-B SECTION A L O N G A - A FIG. 13: FILM HOLDER DETAILS 65 thickness d = 0.003 cm. From these values, at 9.7 G c , s =1.0124 However, the method of holding the d i e l e c t r i c sheet, as shown i n F i g . 13, introduces some discontinuity i n the walls of the waveguide, and this causes additional reflections from the plane of the d i e l e c t r i c sheet. The voltage standing wave ratio of (81) the f i l m holder can be measured by the s l i d i n g load technique ' which i s as follows: The moving short c i r c u i t i n F i g . 16 i s replaced by a s l i d i n g load. By varying the position of the load, a maximum and a minimum value of VSWR, s and s . , respectively, are ' max mm' * J 7 obtained. s and s . correspond, respectively, to the max mm r ' * J ' refl e c t i o n s from the discontinuity (in this case the f i l m holder) and the s l i d i n g load adding i n and out of phase. The maximum and minimum values of the r e f l e c t i o n c o e f f i c i e n t s . r and r . , respectively, are calculated from the general max m i n relationship: s - 1 r s + 1 As the r e f l e c t i o n from the discontinuity and the load add to give r and subtract to give r . , B max & mm' r = r T + max L D r . = r T - r_ or r~ - r, mm L D D L where r ^ and r ^ are the r e f l e c t i o n c o e f f i c i e n t s due to the load and the discontinuity, respectively. Therefore, 66 and rmax ~ rmin , 0 0 or rjj = — ••• •••• 6.2.2 The ambiguity i n the above equations i s removed by a second set of measurements on a diff e r e n t discontinuity and the same s l i d i n g load. Equations similar to 6.2.1 and 6.2.2 are obtained for r ^ and r ^ y "kne r e f l e c t i o n c o e f f i c i e n t of the second discontinuity. Thus r^, r ^ and r ^ a r e known. In the present case, T-Q corresponds to the r e f l e c t i o n c o e f f i c i e n t of the f i l m holder with a 0.03 mm thick mica sheet and TJJ2 "that to the f i l m holder without the mica sheet. From these measurements^ the following values were obtained: r ^ = r e f l e c t i o n c o e f f i c i e n t of the s l i d i n g load, = 0.0031 (or s L = 1.007) rjj = r e f l e c t i o n c o e f f i c i e n t of the f i l m holder with mica, = 0.0275 (or s D = 1.056) r^2 = r e f l e c t i o n c o e f f i c i e n t of the f i l m holder without mica, = 0.0155 (or s D 2 = 1.032) It should be noted that the above values also include the ref l e c t i o n s from the slotted l i n e , probe, and the attenuator pads. Typical values of VSWR for these are 1.01 and 1.15, respectively, while the VSWR of ty p i c a l thermistor and bolometer mounts at X-band i s of the order of 1.5. 6.3 Measurement of Microwave Surface Resistance: When a f i l m with a thickness much less than the skin depth i s placed as a r e s i s t i v e i r i s i n a waveguide, one quarter of a wavelength from a short c i r c u i t . 67 the nature of the standing wave pattern at the generator side of the f i l m gives d i r e c t l y the wave impedance, z , at the (78) surface of the f i l m nearer to the generator. Let z Q , y and % be the characteristic impedance* propagation constant and thickness, respectively, of the metal f i l m . Then, z = z cotlrVf, ... ... 6.3.1 s o / For the T E Q ^ mode of propagation i n a rectangular waveguide, o / .rr \2 o rr o» ... 6.3.2 y 2 = (a)2 - ^ 1 - J ^ and the characteristic impedance of the metal i s z ^  — •»» ...6.3,3 where CL i s the width of the waveguide, a/2% i s the frequency; and a, £ , and e r are, respectively, the permeability, d i e l e c t r i c constant, and the conductivity of the metal. For a metal with good conductivity, ^~/£> 1 and ttu^^Ti/a)2. With these assumptions, i t can be shown that y = Jj(0\iO" = a + j(3 ... . . , 6.3,4 and where z'Q = (1+j) / •» » 6.3.5 a = f3 = / 6>2<~r *** ... 6.3.6 J Since ^> 1, t a n h y f and Eqn. 6.3.1 becomes % Z Q z ^ — ^ jQ — ... ...6.3.7 68 Combining equations 6.3.3* 6.3.4, 6.3.5 and 6.3.7, z = ... ...6.3.8 s 0-1 giving the surface impedance. This expression i s i d e n t i c a l to that i n Eqn. 5.2.1 giving the d.c. surface resistance of a f i l m i n ohms per square* Fig* 14 shows the f i l m placed transversely i n a waveguide. Y^* are the characteristic admittances of the media on either side of the f i l m , and Y i s the characteristic ' o admittance of the metal. Thin Film Generator z = 0 ^ "^ -z — S 1 i >-Load F i g . 14* Transverse Film i n a Waveguide From the transmission lin e equations, the wave admittance (79) at the plane z = 0 iss 1 + Y~ c o t h y € Y = Y q tanh 7 1 + Y~ tanhy£ ... 6.3.9 With the assumptions made above. Y = Y q tanhy-£ 1 + =— cot o 1 - =— tanhy-o ' t. -... 6,3*10 Since Y^>^>Y^y and ky%)2 <^C1» "the second order terms may be omitted. Equation 6,2,10 combined with Eqn. 6.3.8 then gives 69 T = T oj1 + T3 t + To ,«• 6,3,11 The f i l m w i l l be at a point of zero admittance when backed by a perfect short c i r c u i t at a distance of odd multiples of X / 4 , the quarter guide wavelength. Eqn. 6.3.11 indicates that s effect of the thin f i l m i s to introduce an admittance i n shunt with the characteristic admittance of the waveguide. Substitution method for VSWR measurement: A general r e l a t i o n for measuring the V S W R is developed from the study of the incident and refl e c t e d waves on a lossless transmission l i n e , excited by a single frequency sinusoid. If the spatial variable z increases with the distance from the generator (Pig. 15a), then choosing the minima as the ori g i n , we haves S.W. Pattern Gen F i g . 15a. High VSWR Measurement with minima as the Reference F i g . 15b. Vector Diagram for the High VSWR Measure-ment 7 0 2 ^ V~~j = 1 + r 2 - 2 r c o s 2 P Q Z . . . . . . 6 . 3 . 1 2 w h e r e V = t o t a l v o l t a g e o n t h e l i n e a t p o i n t z , V \ = i n c i d e n t v o l t a g e o n t h e l i n e a t p o i n t z , r = t h e v o l t a g e r e f l e c t i o n c o e f f i c i e n t o f t h e l o a d , P q = p h a s e c o n s t a n t o f t h e t r a n s m i s s i o n l i n e i n r a d i a n s / m e t e r , T h u s , 0 = 0Z = . . . . . . 6 . 3 . 1 3 P r o m t h e a b o v e e q u a t i o n s a n d F i g . 1 5 b , t h e g e n e r a l r e l a t i o n i f o r t h e VSWR, s , c a n b e s h o w n t o b e : J l O k - c o s 2 © . . . 6 . 3 . 1 4 s i n 0 w h e r e X = g u i d e d w a v e l e n g t h i n c m s . , k = o c / l O , a n d V a = 2 0 l o g y • m i n = r a t i o , i n d b , o f t h e p o w e r a t a d i s t a n c e z f r o m t h e m i n i m a t o t h a t a t t h e m i n i m a . F o r a = 3 d b , i t c a n b e s h o w n , f r o m E q n , 6 . 3 . 1 4 , ^ ^ t h a t j . g . - 1 1 d = — ^ s i n it J 2 i s - 1 X , ^ ... 7 f o r s » l . . . . . . 6 . 3 . 1 5 TC S ^ X w h e r e d i s t h e d i s t a n c e b e t w e e n t h e t w o p o i n t s o n t h e w a v e , o n 71 either side of the standing wave minimum, at which the power i s twice i t s minimum value* Eqn. 6.3.15 shows that d decreases rapidly with the i n -creasing VSWR and a severs r e s t r i c t i o n i s imposed on the use of Eqn. 6.3.14 for very large values of VSVRs. The surface resistance measurements on the films were made at a constant power l e v e l * The standing wave meter was calibrated with a standard attenuator (maximum error below - 0.1 db). A probe diameter of 0.06 mm* and probe depths between 0.5 to 0.75 mm were used for the high VSWR measurements. The frequency ; measurement by the calibrated wavemeter was accurate to within 0.25 Mc/s. Fi g . 16 shows the arrangement for microwave surface re-sistance measurement. On replacing the r e s i s t i v e f i l m with a short c i r c u i t , the s h i f t i n the position of the minima was either zero on X /4» For a zero minima s h i f t , g R =sZ ........6.3.16 u o and for a X /4 minima s h i f t . g * Zo R — —- ........ 6.3.17 u s where Z q = the wave impedance, = 511.8.0, at 9.7 G c The curves showing R vs t and R vs R.. are plotted i n 6 a a dc * Fig . 18 and F i g . 19 respectively. 6.4 Measurement of Microwave Transmission Coefficients The arrangement for measuring the transmission c o e f f i c i e n t , T, i s shown i n F i g . 17. It i s as follows; 72 i« A mica sheet without the fi l m was placed i n the sample holder, and the calibrated attenuator was adjusted to 50 db. The reading on the VSWR amplifier was noted. i i . The f i l m under test was then placed i n the sample holder. The calibrated attenuator was adjusted to give the same reading on the VSWR amplifier as i n (i ) above. The difference between 50 db and the f i n a l reading of the attenuator gave T i n db. The use of a calibrated attenuator made these values independent of the detector and amplifier response laws. The curves showing T vs t and T vs R^c are plotted i n Fi g . 20 and F i g . 21 respectively. 6.5 Discussion: The results of diff e r e n t microwave measurements are discussed below: 1. Surface resistance at microwaves: Fi g . 19 shows that no fixed r e l a t i o n -ship i s apparent between R and R, . In general, R i s less jA & C JX than R, for high values of Rn and greater than R, for small dc dc dc values of R, . At lower thicknesses, the value of R, i s dc dc increased owing to the presence of i n v i s i b l e cleavage steps on mica. These cleavage steps would not affect the microwave measurements since they are much smaller i n height than the guide wavelength X • On the other hand, as the f i l m thickness increases, the assumption made i n deriving Eqn. 6.3.7, that I, X m ? where X m i s the guide wavelength inside metal, i s no (82 83") longer v a l i d . A more general methodv * ' must be applied for the thicker f i l m s . 73 KLYSTRON POWER SUPPLY X-13 KLYSTRON WAVE - METER ATTENUATOR PAD 10 db VSWR AMPLIFIER i CALIBRATED VARIABLE ATTENUATOR SLOTTED LINE PROBE AND DETECTOR FILM HOLDER MOVABLE SHORT CIRCUIT mm, 10 db) FIG. 16 APPARATUS FOR THE MEASUREMENT OF MICROWAVE SURFACE RESISTANCE KLYSTRON POWER SUPPLY WAVE- METER V. S.W.R. AMPLIFIER ) i i X- 13 ATTENUATOR ATTENUATOR FILM CALIBRATED TUNABLE w PAD 10 db PAD 10 db VARIABLE ATTENUATOR DETECTOR MOUNT KLYSTRON " 'J^ HOLDER FIG. 17 APPARATUS FOR THE MEASUREMENT OF TRANSMISSION COEFFICIENT 74 10^ 10 \ FIG. 18 RELATIVE RESISTIVITY AT MICROWAVES vs THICKNESS • GOLD FILMS ^ o GOLD ON S ILVER F ILMS \ ^ o Gold F i lms on t^j 5 t o 7 A \ \ \ \ ^ \ \ \ \ c \ """ — © o— V I I I I I I I I I I 2 0 3 0 4 5 6 7 8 9 10 2 2 3 4 5 6 0 0 A T — 75 PERCENTAGE OF POWER TRANSMISSION i *» o o ro O O O 2 cn O O O z m 5 cn r o 7) > Z </> 2 co cn O z o o rn < co co C JO Ti > n CO CO —I > Z n s o n Cb* o o* o to O Cn X PERCENTAGE OF POWER TRANSMITTED — • LL 78 Experimental errors arise from the d i f f i c u l t y i n measuring high VSWRs. the imperfect short c i r c u i t , the ref l e c t i o n s from the substrate holder, the f i n i t e losses i n waveguide walls, and the probe thickness. For high VSWRs, the signal at the minima i s below the noise l e v e l , thereby imposing a l i m i t on the VSWRs measurable by the substitution method. The effect of the remaining factors i s to lower the values of measured VSWR. When the measured value of s i s smaller than the actual one, i t can be seen from Eqns. 6.3.16 and 6.3.17 that R w i l l be lower than the true value for high r e s i s t i v i t y films (R = sZ ), and greater than the true value |X o for low r e s i s t i v i t y films (R^ = Z Q / s ) . Also, as s increases, the minima get sharper and the value of d from Eqn. 6.3.15, becomes smaller. At 9.7 Gc, d — 0.013 cm. for s = 100 and &— 0.0013 cm. for s = 1000; while the probe thickness i s 0*006 and thus comparable to d« An increasing amount of error w i l l therefore be introduced due to the difference i n phase and amplitudes of the voltages induced on the two sides of the probe nearer and away from the generator. Accurate measurement of s greater than about 100 i s only feasible with s t i l l thinner probes. There are, nevertheless, a good number of points between R =10 and R = 1,000 ohms/sq. which are f a i r l y near the LI R^ = R^c l i n e i n F i g . 19. Subject to the existence of the cleavage steps and th e i r effect on R^c (Section 5.5), the values of R can be determined from R, measurements with reasonable \i dc accuracy. 79 2* Microwave transmission c o e f f i c i e n t of thin films? Prom the occurrence of maxima and minima, i n F i g . 20, for higher thicknesses of gold f i l m s , a value of X of the order of 400 A ° ' m (49) i s indicated. By virtue of the r e l a t i o n ' this value of X corresponds to a d i e l e c t r i c constant, e , of m r • 1 m7 the order of 60 x 10"^. Khaikin observed that e for t i n m g increased i n the superconducting state by a factor of 5x10 from that of the same metal i n normal state. This suggests a possible order of magnitude of e m for high conductivity metals. In t h i n f i l m s , E m w i l l increase with t, the f i l m thickness, as the structure of the f i l m approaches that of the bulk metal F i g . 21 shows that the microwave transmission c o e f f i c i e n t i s dependent on the d.c. resistance of the films insofar as the effects of multiple r e f l e c t i o n s from the two boundary surfaces are not considered. In Table IV, some ty p i c a l results are shown for comparison. The value of T i s seen to decrease from less than 1$ for ( l ) to less than 0.1$ for (3). The corresponding decrease i n white l i g h t transmission i s from 85$ to 75$ i n the two cases. Films (l) and (2) are pale yellow i n transmission colour and (3) starts taking on a greenish hue. Pure gold films i n the same range of thicknesses are shown i n brackets. These films can be seen to have poor l i g h t trans-mission and microwave shielding properties. The s i g n i f i c a n t feature, thus, i s the fact that the l i g h t transmission and microwave shielding properties improve with increasing con-TABLE IV 80 Film No. Ag Au R dc Approximate -T White l i g h t Trans-mission. (A) (A) ohms/sq. db * 1. 7.52 50.0 17.80 -22.0 85 (0) (48.0) (3.38 K) (-0.4) (70) 2. 4.12 84.5 11.70 -27.9 80 * 3. 1.80 90.3 7.44 -30.9 75 (0) (99.0) (13.14) (-25.4) (65) * This f i l m was annealed at 350° C for one hour. du c t i v i t y for the same f i l m thickness. Improvement of con-du c t i v i t y with annealing and a high heat withstanding capacity are added advantages of the gold-on-silver f i l m s . The possible application i s i n glasses for shielding the eyes from microwave radiation; sandwiching a th i n f i l m of this type between two glass sheets would avoid mechanical abrasion problems. 3. Gold films for bolometers: A number of papers over the l a s t few years have dealt with the application of thin films, p a r t i c u l a r l y of gold and platinum, to bolometers for (36) microwave power measurements. Although Schiffman et a l . ' have reported experiments with gold f i l m bolometers, the method of obtaining these films with the appropriate surface resistance (equal to the characteristic impedance of the waveguide for matching), has not been published. A method for making suitable platinum films i s suggested by Lane which consists of heating the f i l m on a hot plate and to keep checking the 81 resistance t i l l a desired value i s obtained. This i s applicable to the gold films (without s i l v e r ) as well since their resistance i s i r r e v e r s i b l y increased on heating (Sec. 5.5). It can be seen from Figs. 10 and 18 that the t - Pf/p-^ curve r i s e s rapidly below t "= 100 A, so that the control of R^  by thickness v a r i a t i o n i s extremely d i f f i c u l t . An alternative method i s indicated by one result of the present work, which suggests the use of interrupted deposition of gold films on a nucleating s i l v e r layer. The R^ and R^c values of ( l ) gold on s i l v e r f i l m with interrupted deposition of gold, (2) gold—on-s i l v e r f i l m of nearly the same thickness continuously deposited and (3) a pure gold f i l m ; are shown i n Table V, Film 1 was exposed to a i r after the i n i t i a l deposition of 4.3 A Ag and 24.3 A Au layer, after which a 29.8 A Au layer i s coated, resulting i n the surface resistance much larger than that of f i l m no. 2. TABLE V Film No. Ag ^Au ^dc R (A) (A) ohms/sq. ohms/sq. 1. 4,3 24.3+ 400.0 341.0 29*8 = 54.1 2. 3.4 57.5 37.8 35.0 3. 0 78.2 40.4 35.5 This r e s u l t warrants a detailed investigation of the process since a definite p o s s i b i l i t y exists for obtaining suitable f i l m s . It may be noted that on having annealed the f i l m once, subsequent 82 reheating to temperatures below the annealing temperature w i l l affect the resistance only i n a reversible way. This property increases the tolerance of these films as bolometer elements. 7. CONCLUSION 83 The following conclusions are made from the work described i n this thesis: The method of determining the equivalent thickness gives an overall accuracy within 5$ and thicknesses down to about 0,5 A for gold or s i l v e r can be monitored. Electron micrographs and d i f f r a c t i o n patterns indicate that while the gold films are very granular and completely p o l y c r y a t a l l i n e , the gold-on-silver films have an improved texture and the c r y s t a l l i t e s appear to have preferred orientation, D.C, conductivity measurements show a much better conductivity for gold-on-silver films than that for the gold films alone. For a very thin f i l m , a c r i t i c a l nucleating layer thickness of s i l v e r from 5 to 7.5 A gives conducting gold films down to 25 A. The effect of nucleating layers on conductivity i s p a r t i c u l a r l y conspicuous for the gold f i l m thicknesses below 150 A where the two curves i n F i g . 10 merge. Annealing up to 350° C improves the conductivity of the gold-on-silver f i l m s . Further heating increases the resistance* though the films are s t i l l conducting and continuous on heating up to 600°C. The gold films show deteriorating conductivity on heating and become discontinuous and nonconducting at 450° C. The gold-silver composite films have, thus, a very high heat withstanding capacity. This, i n combination with good con-d u c t i v i t y makes them suitable for heat r e f l e c t i n g windows. Microwave and d»c, surface resistances for gold films are nearly the same in many cases, being between 10 to 1,000 ohms/sq. The highly conducting films between 50 to 100 A have good l i g h t transmission* and at 9.7 Gc, microwave transmission i s between -20 to -30 db ( i . e . , 1 to 0.1$). The additional high heat withstanding property of these films suggests a possible application i n making glasses to shield the eyes from microwave radiation* There i s a p o s s i b i l i t y of making very thin gold-silver films (about 50 to 60 A thick) by interrupted deposition of gold layers, suitable for bolometer elements. Further i n v e s t i -gation i n this d i r e c t i o n would provide useful r e s u l t s . REFERENCES 85 1. Faraday, M.s "Experimental Relations of Gold (and other Metals) to Light" - P h i l . Trans, Roy. Soc. (Lond.), 147, 145(1857). 2. Thomson, J.J. s "On the Theory of Metallic Conduction i n Thin Metallic Films" -Proc. Cambridge P h i l . S o c , 11, 120 (190.1). 3. Planck, V . J "Optical Constants of Copper Films" -Physik, Z., 15, 563(1914). 4. Appleyard, E.T.S . j and L o v e l l , E.C.B.; "The. E l e c t r i c a l Conductivity of Thin Metallic Films-IIs Caesium and Potassium on Pyrax Glass" -Proc. Roy, Soc. (Lond.), A-l58, 718(1937). 5. L o v e l l , A.C.B.g "The E l e c t r i c a l Conductivity of Thin Metallic F i l m s - I l l % A l k a l i Films with Properties of Normal Metals" -Proc. Roy. Soc. (Lond.), A-166, 270(1938). 6. Fuchs, K.% "The Conductivity of Thin Metallic Films according to the Electron Theory of Metals" -Proc. Cambridge P h i l ; S o c , 34, 100(1938). 7. London, H.: "The High Frequency Resistance of Super-conducting Tin" - P r o c Roy* S o c (Lond.), Ar176., 552(1940).. 8. Chambers, R.G.j "The Anomalous Skin E f f e c t " - P r oc Roy. Soc. (Lond.), A-215, 481(1952). 9. Dingle, R.B.% "The E l e c t r i c a l Conductivity of Thin Wires" -Proc. Roy. Soc. (Lond.), A-201, 545(1950). 10. Sondheimer, E.H.% "The Mean Free Path of Electrons i n Metals" -Advances i n Physics, v o l . 1 , 1 (.1.952). 11. Mayer, H., Nossek, R., and Thomas, H,% "The Mean Free Range of Conduction Electrons and Photo-Electrons s Measurement by means of Thin Films" - J . Phys. et le Radium, 17, 204(1956). .12. Englman, R., and Sondheimer, E.H. s "The E l e c t r i c a l Con-duc t i v i t y of Anisotropic Thin Films" -Proc. Phys. S o c (Lond.), B-69. 449(1956). 13. F r o h l i c h , H., Paranjpe, B.V., Kuper, C.G., and Nakajima y S.s "The Influence of Interelectronic C o l l i s i o n s and of Surfaces on E l e c t r i c a l Conductivity" -Proc. Phys. S o c (Lond.), B-69, 842(1956). 86 14. Kaganov, M.I., and Azbel, M. Ya. s "Conductivity of Thin Metallic Films" -Zh. Exptl. Teor. F i z . (SSSR), v o l . 27, no. 6(12), 762(1954). 15. Sondheimer, E,H»I " E l e c t r i c a l Conductivity of Anisotropic Thin Films" - J . Phys. Radium, v o l . 17. no. 3, 201(1956). 16. Mayer, H.: "Conduction Phenomena i n Thin Films" i n Structure and Properties of Thin films(eds. Neugebauer, C.A., Newkirk, J.B., and Vermilyea, D.A.), J . Wiley & Sons, 1959, p. 222. 17. Chopra, K.L., Bobb, L.C., and Francombe, M.H.% " E l e c t r i c a l R e s i s t i v i t y of Thin Single Crystal Gold Films" - J . appl. Phys., 34» 1899(1963). 18. Chopra, K.L., and Bobb, L.C.s " E l e c t r i c a l R e s i s t i v i t i e s Studies on P o l y c r y s t a l l i n e and E p i t a x i a l l y Grown Gold Films" -presented before the Philco sponsored International Conference on Thin Films, Blue B e l l , Pa., May, 1963, 19. Wei, L.Y.: "Conductivity of Metal Films of Thickness Comparable to the Electron Wavelength" -presented before the IEEE sponsored Canadian Electronics Conference, Toronto, Sept.-Oct., 1963. 20. Neugebauer, CA. and Webb, M.B. s " E l e c t r i c a l Conduction Phenomenon i n Ult r a — t h i n Evaporated Metal Films" - J . appl. Phys., 33, 74(1962). 21. Hartman, T.E.: " E l e c t r i c a l Conduction i n Discontinuous Thin Metal Films" - J . appl. Phys., 34, 943(1963). 22. Weitzenkamp, L.A.J "Conductivity of Thin Films of Gold" —M.S. Thesis, Dept. of E l e c t r i c a l Engineering, University of Nebraska, Lincoln, July, 1963. 23. Bashara, N.M.: "A Survey of Conduction Mechanism i n Very Thin Films" -IEEE Trans., v o l . CP-11. 4(1964). 24. Glover, R.E., and Tinkham, M.s "Conductivity of Super-conducting Films for Photon Energies 0.3 and 40 kT' " -Phys. Rev., 108, 246(1957) 25. Scales, J.L. I I I , . "Current-Voltage Relationship for Thin Film Tunnelling Structure" -IEEE Trans., v o l . CP-11. 9(1964). 87 26. Sato, H., Toth, R.S., and Astrue, R.W.: "Magnetic Domain Structures of E p i t a x i a l l y Grown Single Crystal Films" -presented before the Philco Sponsored International Conference on Thin Films, Blue B e l l , Pa., May, 1963. 27. Chaudhary, R.D.: "Superconductivity i n Thin Films" -Ph.D. Thesis, Dept. of Physics, Univ. of B r i t i s h Columbia, December, 1963. 28. J u s t i , E., Kohler, M., and Lautz, G.: "Pathlength Problems of Electrons i n Metals and Semiconductors" -Abh. Braunschweig Wiss. Gessel., 3_, 34(l95l). 29. J u s t i , E., Kohler, M., and Lautz, G.: "The Thermoelectric Power of Thin Films" -Z. Naturforsch., 6a, 456(1951). 30. Mayer, H.; Physik dunner Schichten, Stuttgart 1955. 31. Kaner, E.A .J "On the Theory of Galvanometric and Thermo-e l e c t r i c Effects i n Metallic Films" -Soviet Physics, J.E.T.P., 7, 454(1958), 32. Tolansky, S.: Surface Microtopography, Longmans, 1960. 33. Gillham, E.J., Preston, J.S., and Williams, D.W.: "Transparent, Conducting Films" - P h i l . Mag., 46, 1051(1955). 34. Ennos, A.E.s "Highly Conducting Gold Films prepared by Vacuum Evaporation" - B r i t . J-. appl. Phys., 8, 113(1957) 35. Holland, L., and Siddal, G.: "Heat Reflecting Windows using Gold and Bismuth Oxide Films" - B r i t . J . appl. Phys., 9, 359(1958). 36. Schiffman, B.M.* Young, L., and Lerrick, R.B.s "Thin Film Waveguide Bolometers for Multimode Power Measurements" -IEEE Trans., v o l . MTT-12, no. 2, 155(1964). 37. HarVey, A.F.: Microwave Engineering, Academic Press, 1963; Ch. 20. 38. Hirsch, F.G., and Parker, J.T.s " B i l a t e r a l Lenticular Opacities occurring i n a Technician Operating a Microwave Generator" -AMA Arch. Ind. Hyg., 6, 512(1952). 39. Egan, W.G.: "Eye Protection i n Radar F i e l d " -Elect. Engng., 76, 126(1957). 40. Appleyard, E.T.S.: "The E l e c t r i c a l Conductivity of Thin Films of A l k a l i Metals Spontaneously Deposited on Glass Surfaces" -Proc. Roy. Soc. (Lond.), extra part, 49, 89(1937). 88 41. Lennard-Jones, J.E.s "The Migration and Agglomeration of Metal Atoms on Solid Surfaces" -Proc. Roy. Soc. (Lond.), extra part, 49, 140(1937). 42. Rhodin, T.N.% "Oriented Arrangement of Thin Aluminium Films on Ionic Substrates" -Disc. Faraday S o c , 5, 215(1949). 43. Benjamin, P., and Weaver, C.: "Condensation Energies for Metals on Glass and other Substrates" - P r o c Roy. Soc (Lond.), A-252, 418(1959). 44. Chandra, S., and Scott, G.D.: "Condensation Coefficients of S i l v e r , Gold, and Copper i n Vacuum" -Canad. J . Phys., 36, 1148(1958). 45. M i t c h e l l , J.TfT. s Introduction to Structure and Properties of Thin Films, (eds. Neugebauer et a l . ) , J . Wiley & Sons, 1959, p. 3. 46. Levinstein, H.s "The Growth and Structure of Thin Metallic Films" - J . appl. Phys., 20, 306(1949). 47. Wilkinson, P.G.s "Lattice Distortion Spectrum of Evaporated Gold" - J . appl. Phys., 22, 419(1951). 48. Pashley, D.W.; "The Preparation of Smooth Single Crystal Surfaces of S i l v e r by an Evaporation Technique" - P h i l . Mag., 4, 316(1959). 49. Moreno, T.: Microwave Transmission Design Data, Dover, 1958. 50. B l o i s , M.S., J r . , and Riesser, L.M., J r . s "Apparent Density of Evaporated Thin Films" - J . appl. Phys., 25, 338(1954). 51. Greenland, K.M.s "Measurement and Control of the Thickness of Thin Films" -Vacuum, 2, 216(1952). 52. Tolansky,.S.s Multiple Beam Interferometry, Oxford Clarenden Press, 1948, 53. Stern, B.J. : "Thin Film Thickness Measurement using S i l v e r Modified Newton's Rings" -Rev. s c i . Instrum., 34-2, 152(1963). 54. Friedman, H., and Birks, L.S.s "Thickness Measurement of Thin Films by X-ray Absorption" -Rev. s c i . Instrum., 17, 874(1946). 55. Eisenstein, A.; "An X-Ray Method for Measuring the Thickness of Thin Crys t a l l i n e Films" - J . appl. Phys., 17, 874(1946). 89 56. Sauro, J . , Fankuchen, I», and Weinfan, N.s "X-Ray Interference Structure i n the Specularly fieflected Radiation from Thin Films" -Phys. Rev., 134-4, 1544(1963). 57. Antal,'J.J., and Weber, A.H.: "A Tracer Method for Thickness Measurement of Thin Bismuth Films" -Rev. s c i . Instrum., 23, 424(1952). 58. Preuss, L.E,: "A Study of the Propagation Mode for Metal Vapours i n Shadow Casting by Vacuum Evaporation of Au(l98) and C r ( 5 l ) " - J . appl. Phys., 24, 1401(1953). 59. Devienne, M.j "Determination of Thickness of a Thin Film Obtained by Condensation of A r t i f i c i a l Radioactive Element" -C.R. Acad. S c i . , Paris, 232, 1088(1951). 60. Holland, L . j "The Production, Properties and Uses of Thin Films Condensed i n Vacuo" -Vacuum, 1, 23(1951). 61. Dudding, R.W.: "Aluminium Backed Screens for Cathode Ray Tubes" - J . B r i t . Inst. Radio Engrs., 11-4. 455(1951). 62. Holland, L.: Vacuum Deposition of Thin Films, Chapman & Hal l , 1961. 63. Lins, S.J., and Oberg, P.E.: "Automatic Deposition Control" -Electronics, 36-13, 33(1963). 64. Cady, W.G.: P i e z o e l e c t r i c i t y , Dover, 1963; McGraw-Hill, 1948. 65. Thun, R.E.: "Structure of Thin Films" i n Physics of Thin Films»(ed. G. Hass), Academic Press, 1963; p. 187. 66. Gilman, J . J . : "Deformation and Fracture of Ionic Crystals" i n -Progress i n Cermaic Sciences, v o l . 1, (ed. J.E. Burke); Pergoman Press, 1961. 67. Bradley, D.E.: "Evaporated Carbon Films for use i n Electron Microscopy" - B r i t . J . appl. Phys., 5, 65(1954). 68. Weale, R.A.: "The R e s i s t i v i t y of Thin Metallic Films" -Proc. Phys. Soc. (Lond.), A-62. 135(1949). 69. Bloch, P. : "On the Quantum Mechanics of Electrons i n Crystal Structures" - Z e i t s . f . Physik, 52, 555(1928). 70. Bragg, W.L.: Atomic Structure of Minerals. Cornell Univ. Press, 1937, p. 210. 90 71. D i x i t , K.R.: "Epitaxial Growth of Thin Films" -Proc* Indian Acad. S c i . , A-48, 330(1958). 72. Basset, G.A.: "A New Technique for Decoration of Cleavage and S l i p Steps i n Ionic Crystal Surfaces" - P h i l * Mag., 3, 1042(1958). 73. Bannet, G.A.: "An Investigation into the Binding Energies of Thin Films" Structure and Properties of Thin Films, (eds. ;op Neugebauer, et a l . ) , J . Viley & Sons, 1959; p. 58* 74. Basset, G.A., Menter* J.V., and Pashley, D.¥.: "The Nucleation, Growth and Microstructure of Thin Films" i n Structure and Properties of Thin Films (eds. Neugebauer et a l . ) , J . Viley & Sons, 1959; p. 1. 75. Khaikin, M.S.: "Surface Resistance of Superconducting Tin at Frequency 9,400 Mc/s" -Doklady Acad. Nauk. SSSR, 56:3, 517(1952). 76. Pippard, A.B. : "Metallic Conduction at High Frequency and Low Temperatures" -Advances i n Electronics and Electron Physics, v o l . VI,~TTl954TT 77. Spiewak, M.: "Surface Impedance of Superconducting Tin" -Phys. Rev., v o l . 13, no. 6. 1479(1959). 78. Clark, D.E.: "The R e s i s t i v i t y of Thin Metallic Films" - B r i t . J . appl. Phys., 6, 158(1955). 79. Slater, J . C : Microwave Electronics, D, Van Nostrand Co*, 1963, p. 30. 80. Nunn, V.M.: "Substitution Method of Measuring Standing Vave Ratio" -Rev. s c i . Instrum., v o l . 32, no. 10, 1106(1961). 81. Hewlett Packard Application Note No. 38, Microwave Measurements for Calibration Laboratories, 1962, Section 4, p. l i s "SVR Measurement of Transmission Systems by Sl i d i n g Load Technique". 82. Speirsj L.s "Metal Optics at Centimeter Wavelengths—I" - P h i l . Mag., 39, 105(1948). 83. Schneider, M.V.s "Reflection and Transmission of Conductive Films" -Proc. IRE (Corresp.) 49, 1090(1961). 84. Lane, J.A.s "Transverse Film Bolo-meters for the Measure-ment of Power i n Rectangular Vaveguides" -Proc* IEE, 105-B, 77(1958). 

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.831.1-0104838/manifest

Comment

Related Items