Chapter 2

 

Diamond Growth and Characterisation

 

2.1. Background

 

2.1.1. Introduction

 

Diamond is one of the hardest natural materials, has one of the highest thermal conductivities at room temperature, is transparent over a very wide wavelength range, is one of the stiffest materials, one of the least compressible, and is inert to  most chemical reagents at room temperature. With these remarkable properties, diamond has sometimes been referred to as “the best natural material.”

 

Unfortunately, it has proved very difficult to exploit these properties, due to the cost, the lack of abundance of large natural diamonds and the fact that diamond until recently was only available in the form of stones or powder. It had been known for 200 years that diamond is composed only of carbon ¹.

 

 Many efforts have been made to synthesize diamond artificially using as a starting material another commonly occurring form of carbon, graphite (see figure 1.1). This has proved very difficult, mainly because at room temperature and pressure, graphite is the thermodynamically stable allotrope of carbon. Although the standard enthalpies of diamond and graphite only differ by 2.9 kJ mol^{-1 2}, a large activation barrier separates the two phases avoiding interconversion between them at room temperature and pressure. This large energy barrier is also responsible for its existence. When diamond is formed, it does not spontaneously convert to the graphite phase. For this reason, diamond is said to be metastable, that is kinetically stable but not thermodynamically stable.

 

 

2.1.2. High-pressure high-temperature technique

 

Knowledge of the conditions under which natural diamond is formed suggests that diamond can be formed by heating carbon under very high pressure. This process forms the basis of the high-pressure high-temperature (HPHT) growth technique ³. In this process graphite is compressed in a hydraulic press to tens of thousands of atmospheres, heated to over 2000 K in the presence of a suitable metal catalyst, and left until diamond crystallises.

 

The diamond crystals thus produced are used for a wide range of industrial processes, which use the hardness and wear resistance properties of diamond, such as cutting and machining components and the polishing and grinding of optics. However, the disadvantage of the HPHT method is that it still produces diamond in the form of single crystals ranging in size from nanometers to millimetres, limiting the range of the applications for which it can be used.

 

2.1.3. Chemical vapour deposition technique

 

The production of diamond via the addition of carbon atoms one-at-a-time to an initial template, was fist postulated in 1955 ⁴. It was suggested that if this could be accomplished, the much lower gas pressures that HPHT, there would be an advantage in terms of equipment and energy cost.

 

 These ideas promoted the experiments of Eversole ⁵ and Deryagin et al. ⁶, in which thermal decomposition of carbon-containing gases under reduced pressure was used to grow diamond on the surface of natural diamond crystals heated to 900 ^oC. However, the rate of growth in these early experiments was low, since graphite was co-deposited with the diamond leading to impure mixed phases. Diamond synthesis advanced in the late 1960s, when Angus’s group discovered that the presence of atomic hydrogen during the deposition process would lead to preferential etching of the graphite, rather than diamond ^{7, 8}.

 

Subsequent Russian work showed that such chemical vapour deposition (CVD) techniques could be used to grow diamond on non-diamond substrates ^{9, 10}.

 

In 1982 Japanese researchers at the National Institute for Research in Inorganic Materials (NIRIM) brought all these findings together to build a “hot filament reactor”, which grew good quality diamond films on non-diamond substrates at significant rates (ca. 1mm h^-1)^{11, 12}. The following year the same group reported another method for achieving diamond growth, using a microwave plasma reactor ^{13, 14}.

 

2.1.4. Hot Filament CVD

 

Hot Filament CVD (HFCVD) involves a gas phase chemical reaction occurring above a solid surface, which causes deposition onto that surface; the gas phase reaction activation requires a source of energy, in this case, a hot filament.

 

Normal conditions in this type of CVD chamber employ a precursor gas (usually CH₄) which is diluted in an excess of hydrogen in a typical mixing ratio of 1% by volume. The vacuum chamber is continually pumped using a rotary pump, while process gases are measured at carefully controlled rates (typically a total flow rate of a few hundred s.c.c.m.). Throttle valves keep the pressure in the chamber at typically 18-35 Torr, while a substrate heater is used to bring the substrate up to a temperature of 750-950^oC. The substrate to be coated e.g. a piece of silicon or molybdenum is placed on the heater, a few millimetres underneath the filament, which is electrically heated to temperatures of about 2300^oC. The filament is made from a metal that will be able to survive these conditions and not react significantly with the process gases. Metals like tungsten and tantalum are frequently used, although they slowly react with the carbon-containing gases to form the metal carbide. This changes their resistivity and makes them brittle, reducing their lifetime and hence limiting the maximum deposition time of a single run.

 

This HFCVD method is relatively cheap and easy to operate. It produces  polycrystalline diamond films at a rate about 1-12 mm h^-1, depending on the deposition conditions. However, it has some disadvantages. The hot filament is particularly sensitive to oxidizing or corrosive gases. This limits the different gas mixtures which could be employed. It is also very difficult to avoid contamination of the diamond film with filament material. For diamond to be used in mechanical applications, metallic impurities at the tens of ppm level are not a very significant problem. Diamond materials with this level of metal contamination are not acceptable for electronic applications.

 

2.1.5. The choice of substrates for growing CVD diamond

 

The choice of substrates depends on several criteria, some of them very simple. One of the requirements is that the melting point (at the process pressure) must be higher than the temperature required for diamond growth (normally greater than 750ºC). Next is that the substrate material should have a thermal expansion coefficient comparable with that of diamond. At the high growth temperatures currently used, a substrate will expand. Therefore the diamond coating will be grown upon and bonded directly to an expanded substrate. During the cooling stage, the substrate will contract back to its room temperature size. The diamond coating (with its very small expansion coefficient) will be relatively unmodified by the temperature change. Hence the diamond film will experience significant compressive stresses from the shrinking substrate, leading to bending of the sample, and/or cracking or even delamination of the whole film.

 

In order to form adherent films, it is necessary that the substrate material be able to form a carbide layer. The reason is because diamond CVD on non-diamond substrates usually requires the formation of a thin carbide interfacial layers, on which the diamond then grows. The carbide has been described as the “glue” ¹⁵ which promotes growth of diamond and helps its adhesion by (partial) relief of the stresses at the interface (caused by lattice mismatch or substrate contraction).

 

Focusing on carbon substrate interactions, any substrate material can be divided into three categories, based upon their reactivity with carbon:

 

·        Little or no solubility or reaction with carbon.

 

These materials do not form a carbide layer. Therefore a diamond layer will  not adhere well to the surface. This can be used as a method to make free-standing diamond films, as the films will frequently delaminate after deposition. This category includes metals like copper, tin, lead, silver and gold, as well as non-metals like germanium, sapphire and alumina.

 

·        Substantial mutual solubility or reaction with carbon.

 

The substrate acts as a carbon sink, and deposited carbon dissolves into the surface, forming a solid solution. This can result in large quantities of carbon being transported into the bulk, rather than remaining at the surface (where it can promote diamond nucleation). Frequently diamond growth only begins after the substrate is saturated with carbon, and this can dramatically modify the physical properties of the resulting composite. Metals where this is significant include platinum, palladium, rhodium, nickel, titanium and iron (iron and stainless steel cannot be coated using simple CVD methods)

 

·        Carbide formation.

 

These include metals like titanium, zirconium, vanadium, hafnium, tantalum, chromium, molybdenum, cobalt, tungsten, nickel, iron, yttrium, aluminium, and certain other rare-earth metals. In some metals such as titanium, the carbide layer continues to grow during diamond deposition and can be hundreds of micrometers thick. Such thick interfacial carbide layers may dramatically affect the mechanical properties and therefore the utility of the diamond coatings on these materials. A non-metal, such as boron or silicon, and silicon-containing compounds such as silica (SiO₂), quartz and Si₃N₄, also form carbide layers. Substrates composed of carbides themselves, such as silicon carbide (SiC), tungsten carbide (WC) and titanium carbide (TiC) are often used for diamond deposition.

 

The difficulties associated with diamond growth on problematic materials have ensured the continuing popularity of silicon as a substrate material. It has a sufficiently high melting point (1956ºC), it forms only a localised carbide layer (a few nanometres thick), and it has a relatively low thermal expansion coefficient. Molybdenum and tungsten display similar properties. These two metals are also widely used as substrate materials.    

 

2.1.6. Nucleation

 

Growth of diamond begins when individual carbon atoms nucleate onto the surface to initiate the beginnings of an sp³ tetrahedral lattice.

 

There are two different types of diamond growth:

 

·        Homoepitaxial growth.

 

Using natural diamond substrates, the template for the required tetrahedral structure is already there, and the diamond lattice is just extended atom-by atom as deposition proceeds.

 

·        Heterojunction growth.

 

Using non-diamond substrates, there is no such template for the C atoms to follow, and those C atoms that deposit in non-diamond forms are immediately etched back into the gas phase by reaction with atomic H.

 

To deal with the problem of the initial induction period before which diamond starts to grow, the substrate surface often undergoes a pre-treatment prior to deposition in order to reduce the induction time for nucleation and to increase the density of the nucleation sites. There are two main methods to apply this pre-treatment:

 

(a)              Manual abrading. Abrasion of the substrate surface using diamond powder ranging in size from 10nm to 10 mm. It is believed that such polishing helps nucleation by either:

 

·        creating appropriately shaped scratches in the surface, which act as growth templates,

 

·        embedding nanometre-sized fragments of diamond into the surface, which then act as seed crystals,

 

·        a combination of both.

 

(b)        Ultrasonic agitation. A better-controlled version of abrasion of the surface is to use ultrasonic agitation to abrade the substrate immersed in a solution of diamond powder in water or 2-propanol (IPA).

 

2.1.7. The CVD diamond film

 

When individual diamond crystallites have nucleated on the surface, growth continues in three dimensions until the crystals coalesce. At this point a continuous film is formed and the only way growth can proceed is upwards. The resulting film is polycrystalline with many grain boundaries and defects, and exhibits a columnar structure extending upward from the substrate. Furthermore, as the film becomes thicker, the crystal size increases while the number of defects and grain boundaries decreases. This means that the outer layers of thicker films are frequently much better quality than the initial nucleating layers.

The surface morphology of the diamond film obtained during CVD depends dramatically on several process conditions, especially the gas mixing ratio. In CH₄/H₂ systems, depending upon the ratio of methane to hydrogen, the film can be randomly oriented (see figure 2.1) or have some degree of preferred orientation, such as (111) triangular or (100) square facets. With increasing methane concentrations, the crystal sizes decrease, until above ca. 3% CH₄ in H₂ the crystalline morphology disappears altogether (see figure 2.2). Such film is referred to as “nanocrystalline” or “ballas” diamond, and may be considered to be an aggregate of diamond nanocrystals and disordered graphite. Although this type of film might be considered inferior to the more crystalline and therefore better quality diamond films, it still possesses many of the desirable properties of diamond while being much smoother and considerably faster to deposit. Thus, by the simple tactic of changing the growth conditions, diamond films can be deposited with properties ranging from almost graphite to essentially those of natural diamond. This allows the quality, appearance and properties of a diamond film, as well as its growth rate and cost, to be easily designed to suit particular applications.

Figure 2.1. Randomly oriented surface of a HFCVD (sample B13)

 

Figure 2.2. HFCVD Diamond film grown in an atmosphere containing 2.8 % methane (“ballas” diamond) (sample B8)

 

2.1.8. The chemistry of CVD diamond growth

 

The complex chemical and physical processes which occur during diamond CVD compromise several different but interrelated features, and are illustrated in the figure 2.3. The process gasses first mix in the chamber before diffusion toward the substrate surface. They pass through an activation region, e.g. a hot filament, which provides energy to the gaseosus species. This activation causes molecules to fragment into reactive radicals and atoms, creates ions and electrons, and heats the gas up to temperatures reaching a few thousand kelvins. Beyond the activation region, these reactive fragments continue to mix and complete a complex set of chemical reactions until they strike the substrate surface. At this point the species either adsorb and react with the surface, desorb again back into the gas phase, or diffuse around close to the surface until an appropriate reaction site is found. If surface reaction occurs, one possible process, if all the conditions are suitable, is the growth of diamond.

 

Figure 2.3. Schematic diagram of the physical and chemical processes during diamond CVD.

 

There have been many studies of the gas phase chemistry ¹⁶ in the last 10 years aimed at obtaining a clear picture of the principles involved. The first clue was obtained from the “Bachmann triangle diagram” ¹⁷, which is a C-H-O composition diagram based upon over 70 deposition experiments in different reactors and using different process gases. Bachmann found that independent of the deposition system or gas mixture, diamond would only grow when the gas composition was close to and just above the CO tie-line (see figure 2.4). This means that diamond growth was independent of the nature of the gas phase precursors, and that the gas phase chemistry was so rapid it simply and effectively broke down the constituent gases to smaller, reactive components.

 

There have been many suggestions for the species involved in diamond growth;  including C, CH, C₂, C₂H, CH₃, C₂H₂, CH₃^-, and diamondoids, such as adamantine. However, since diamond can be grown in systems which have few ions present (e.g. HFCVD reactors), this suggests the growth species must be neutral moieties. Further numerical simulations have shown that diamond growth can be accounted for by a single growth species and a single surface mechanism. A number of studies have been performed to try to identify the growth species chemistry ¹⁶, and the general consensus is that CH₃ is the important radical.

 

The generally agreed mechanism for CVD diamond growth is as shown in figure 2.5. During growth, the diamond surface is nearly fully saturated with hydrogen. This coverage limits the number of sites where hydrocarbon species (probably CH₃) may adsorb, and also blocks migration sites once they are adsorbed. Atomic H abstracts a surface H to form H₂, leaving behind a reactive surface site. The most likely fate for this surface site is to react with another nearby H atom, returning the surface to its previous stable situation. However, occasionally a gas phase CH₃ radical can collide and react with the surface site, effectively adding a carbon to the lattice. This process of H abstraction and methyl addition may then occur on a site adjacent to the attached methyl.

Figure 2.4. Triangular diagram of Bachmann

 

A further H abstraction process on one of the chemisorbed groups creates a radical, which attacks the other nearby carbon group to complete the ring structure, locking the two carbons into the diamond lattice.

 

Diamond growth can be considered to be a stepwise addition of carbon atoms to the existing diamond lattice, catalysed by the presence of excess atomic H. In oxygen-containing systems, it is believed that the OH radical plays a similar role to atomic H, except that it is more effective at removing graphitic carbon, leading to higher growth rates and better quality films.

Figure 2.5. Schematic diagram of the diamond growth mechanism (in red colour the new bonds formed)

 

2.1.9. Role of atomic hydrogen in the CVD growth

 

It has been proposed that atomic hydrogen is the most critical component in the gas phase mixture, and indeed that it drives the whole chemical system. In a hot filament system, atomic hydrogen is produced heterogeneously by the thermal decomposition of H₂. A high concentration of atomic hydrogen is essential for a number of processes:

 

·        Although the bulk of diamond is fully sp³ bonded, at the surface there is effectively a “dangling bond”, which needs to be terminated in some way in order to prevent cross-linkage, and subsequent reconstruction of the surface to graphite. This surface termination is performed by hydrogen (or sometimes OH), that keeps the sp³ diamond lattice stable. During the diamond growth, some of these hydrogen atoms need to be removed and replaced by carbon-containing species. A large number of reactive hydrogen atoms close to the surface can quickly bond to any excess dangling bonds that may have been created by thermal desorption or abstraction of the surface hydrogen atoms, avoiding surface graphitization.

 

·        Atomic hydrogen is known to etch graphitic sp² carbon many times faster than diamond-like sp³ carbon. The H atoms serve to remove back to the gas phase any graphitic clusters that may form on the surface, while leaving the diamond cluster behind. Diamond growth could be considered as “five steps forward, but four steps back,” with the net result being a (slow) build-up of diamond.

 

·        H atoms are efficient scavengers of long-chained hydrocarbons, breaking them up into smaller pieces. This avoids the build-up of polymers or large ring structures in the gas phase, which might ultimately deposit onto the growing surface and inhibit diamond growth.

 

·        H atoms react with neutral species such as CH₄ to create reactive radicals, such as CH₃^·, which can attach to adequate surface sites

 

2.1.10. In situ doping

 

The effects caused by incorporation of dopants or impurities in CVD diamond films during the deposition have been studied extensively during the last 10 years. This was motivated first by the wish to make CVD diamond conducting in order to realise electronic applications. Secondly by the more fundamental objective to study the structural effects of dopants or impurities on the growth on CVD diamond films.

 

The most common impurities considered in diamond film technology are boron and nitrogen. Boron makes diamond p-type with an activation energy of 0.37 eV. Nitrogen forms a deep donor and is electrically inactive at room temperature. Its importance results from its strong influence on the structural properties and the growth velocity of the CVD diamond films ¹⁸. In order to realise n-type conductivity, several other dopants such as lithium, sodium and phosphorous have been proposed ¹⁹. Koizumi et al. ²⁰ reported on homoepitaxially grown CVD diamond films doped with phosphorous which exhibit n-type conductivity with an activation energy of 0.43 eV.

 

Normally, nitrogen can be simply introduced to gaseous environment as N₂ or NH₃. Boron, on the other hand, has to be added to the reactant gas in solid, liquid or gaseous form. One commonly used gaseous form of boron at room temperature is diborane. It has been used widely to dope CVD diamond films ^21-24. However, diborane is highly toxic and for this reason several less toxic and more easily handled sources in solid or liquid form have been investigated. Solid sources used for boron doping include boron powder ²⁵ and boron trioxide (B₂O₃) ²⁶. Boric acid (H₃BO₃) ²⁷, cyclic organic borinate ester ²⁶ and trimethylborate ((CH₃)₃BO₃) ²⁷ are liquid sources which have been applied successfully.

 

 The non-gaseous boron sources must be either heated or dissolved in a high-pressure liquid, e.g. alcohol, to increase their vapour pressure sufficiently. In order to control the boron concentration in the reactant gas, bubbler systems have been applied in the case of boron trioxide dissolved in alcohol, cyclic organic borinate ester and trimethylborate. Solid boron sources have also often been placed near the substrate and heated until they can diffuse to the growth surface.

 

The boron incorporation depends on the texture of the diamond film or the orientation of the single-crystal diamond. In single-crystal diamonds, Spytsin et al. ¹⁰ observed enhanced boron incorporation in (111) diamond crystals. This result was confirmed by Samlenski et al. ²⁸ who applied nuclear reaction analysis in order to quantify the concentration of boron in doped homoepitaxial films. They found the boron incorporation probability in (111) oriented films to be one order of magnitude higher than in the (100) oriented films. Similar results were obtained by Locher et al. ²⁷ applying secondary ion mass spectroscopy on B-doped polycrystalline CVD diamond with different textures. The total amount of boron incorporated into CVD diamond films can be varied over several orders of magnitude up to concentrations of about 10²¹ cm^-3 without significant deterioration of the structural quality under appropriate deposition conditions. Whereas the probability for boron incorporation can be higher than 10^-1, the nitrogen incorporation probability was found to be only about 5´10^-4, dependent also on the growth direction ²⁸. However, under specific deposition conditions nitrogen has a dramatic influence on the morphology and the structure of CVD diamond films ^{29, 30}.

 

2.1.11. Ex situ doping

 

Ex situ doping of diamond can be performed by diffusion or ion-implantation. Diffusion of impurities in diamond requires extremely high temperatures due to the low diffusion coefficient at moderate temperatures. Successful boron doping by diffusion was reported by Tsai et al. ³¹ for the fabrication of a diamond metal-semiconductor field-effect transitor (MESFET).

 

Ion implantation is the method of choice for modern microelectronics in silicon technology. However, implantation requires annealing to remove the damage and to electrically activate the implants. In this case diamond behaves quite differently than silicon. Misleading results for diamond can be obtained due to the electrical activity or damage in diamond caused by graphitization, amorphization or point-defect agglomerates.

 

For implantation energies from tens to hundreds of keV the penetration of energetic ions in the diamond lattice creates mainly vacancies and interstitial carbon atoms. Fontaine ³² characterized boron implanted CVD diamond by electron spin resonance and observed that below a dose threshold around 3´10¹⁵ B cm^-2 the spin number increased linearly with the ion dose and saturated above this threshold. Assuming that spins originate from dangling carbon bonds, which leave paramagnetic unpaired electrons, the defect formation rate was estimated to be 36 defects per implanted ion. For doses above the threshold the defect concentration was estimated to be in the order of 10²¹ cm^-3.

 

2.1.12. Summary

 

A review about the fundamental aspects of the diamond growth has been completed. In parallel the influence of the substrate, different gas mixtures, temperature and pressure on the growth process and growth mechanism have been discussed. In addition the role of the hydrogen in the growth of diamond films have been described. Finally some doping techniques to obtain low electrical resistance diamond have been explained.

 

2.2. Diamond Growth

 

The diamond samples of this thesis were provided by Dr. M. Latto and Dr. P. May. The information details the growth they have employed.

 

2.2.1. Introduction

 

Polycrystalline diamond possesses physical properties that suggest that electrodes fabricated from suitably doped samples of this material will exhibit advantageous attributes. For example, Compton and co-workers ^33-39 have shown that in sonoelectrochemical experiments diamond electrodes are resistive to surface damage. In addition it has been claimed that boron doped diamond electrodes possess advantageous electrochemical characteristics that include: a wide potential window, low background currents and excellent resistance to surface fouling (cross-references section 1.5).

 

Therefore it is necessary to design and build a system capable of growing diamond in the adequate conditions to obtain the desirable material with the requirements cited before.

 

A scheme of the elements and gas lines that constitute the system employed to grow diamond in this study is displayed in Figure 2.6.

 

The elements of the system will be described in the forthcoming sections.

 

2.2.2. Gas Flow System

 

Hydrogen (H₂), methane (CH₄) and diborane (B₂H₆) gases (BOC Speciality Gases) were used in the growth process.

 

Laboratory standard “high purity hydrogen” and methane were used. The purities of the hydrogen and methane were 99.995 (as “N4.5”) and 99.5% (as “N2.5”) respectively.

 

Diborane was supplied as a 4.75 % premix of diborane in hydrogen. Diborane is highly toxic, reactive and explosive gas. Therefore special care was taken to ensure that all gas lines were free of leaks (gas lines were leak tested several times before use). The gas cylinder, the dilution cylinder and the associated piping were kept in a specially designed fume cupboard to provide extra protection in case of an accident. The reactivity of diborane made it necessary to avoid the use of many materials including brass and nitrile rubber.

 

The source gases were stored in gas cylinders with nominal pressures up to     200 bar at 15ºC. Regulators reduced the cylinder pressures to the level of 1-2 bar above atmospheric pressure. Mass flow controllers fixed the flow rates into a gas manifold where the gases were mixed before they flowed along a common gas line into the reaction chamber.

 

The regulator on the diborane gas cylinder was a stainless steel single stage design which could be pumped down to vacuum and contained a purge line. This also allowed the dilution procedure to be performed. The hydrogen and methane gas cylinders used standard single brass regulators.

Figure 2.6. A schematic diagram of the gas lines which fed the diamond CVD chamber

 

2.2.3. Mass Flow Controllers

 

Three mass flow controllers (MFC) were used to adjust the flow rate of each gas. Each MFC fixed the flow of gas through a gas line and into a manifold where the gases were mixed (see table 2.1).

 

The MFCs were controlled electronically and could be programmed to allow a wide range of flow rates through the gas lines. Two different MFC models were used to control diborane gas flow rates in order to achieve low or high doping levels.

 

A bypass line was fitted parallel to the MFCs to facilitate the pumping down of the diborane gas line.

 

There was the possibility of adding a fourth MFC for bringing a new gas into the mixture. H₂S, NH₄ were proposed as an alternative doping source but this facility was not explored in this study.

 

 

Table 2.1. Specifications of the mass flow controllers.

 

2.2.4. Dilution

 

Diborane was supplied as mixture of 4.75% B₂H₆ in H₂. This amount of diborane was too high for low doping and a dilution system was necessary to decrease the concentration. This process could be repeated as many times as necessary until the required concentration of B₂H₆ in H₂ was reached.

 

A lecture bottle (capacity 0.4 dm³) was used as the mixing vessel. When B₂H₆ was diluted, this lecture bottle was used as the reservoir of B₂H₆ allowing a fixed flow rate of diborane in hydrogen for several hours.

 

A picture of the dilution set-up could be seen in figure 2.7.

 

Figure 2.7. A picture of diborane dilution system

 

2.2.5. Deposition chamber

 

The deposition chamber consisted of a single-walled stainless steel six-way cross with welded joints and bolted flanges (see figure 2.8).

Figure 2.8. Hot filament CVD chamber

 

 The deposition chamber could be divided in the next different parts:

                       

-                     Top flange.

-                     Front flange.

-                     Bottom flange.

-                     Rear flange.

-                     Left flange.

-                     Right flange

 

2.2.5.1. Top flange

 

This top flange contains the following elements:

 

·          Three pairs of electrical contacts made from tungsten rods.

 

-           One pair was used to supply power to heat the substrate.

 

-           The other to supply power to the filaments.

 

-                     Earthing was achieved by a cable attached to this flange.

 

·            Filament holder and substrate heater were attached underneath the flange to allow new filaments and substrates to be loaded when this flange was removed.

 

Two different insulation processes were done in the top flange:

 

-           Electrical insulation was performed by glass coating.

 

-           Vacuum sealing was achieved by silicon rubber and epoxy resins.

 

2.2.5.2. Front flange

 

Front flange was built with a glass window to allow visual monitoring of the growth process.

 

2.2.5.3. Bottom flange

 

The bottom flange was connect to a two stage rotary pump via two stainless steel gas lines:

 

·          A narrow bore line fitted with a needle valve allowed an accurate control of the pressure inside the chamber during the growing process.

 

·          A broad bore line was used to reduce as much as possible the period of time to pump down the chamber.

 

Openings of these lines were protected by copper gauze.

 

The exhaust line from the rotary pump was guided to a vent.

 

2.2.5.4. Rear flange

 

Three gas lines fed into the CVD chamber through this flange (see figure 2.9):

 

·          Reaction gas inlet leading from the mixing manifold.

 

·          A connection to a pressure gauge (baratron) which was used to measure the pressure in the chamber during the reaction.

 

·          An air vent valve that allowed the chamber to reach atmospheric pressure before it was opened

Figure 2.9. Rear view of the hot filament CVD chamber.

 

2.2.5.5. Left flange

 

No elements are attached to this flange

 

2.2.5.6. Right flange

 

Right flange was provided with two additional electrical contacts.

 

2.2.6. Substrate heater

 

2.2.6.1. Construction of the substrate heater

 

A nickel-chromium wire (24 SWG, standard gauge) was protected with insulating ceramic beads and coiled. Next, everything was covered by fire cement and heated until cement was dried. A molybdenum plate was placed onto the top of the cement block to yield a flat surface where samples could be placed. Molybdenum was chosen because of its properties: high melting point, inert in the reaction environment and uniform distribution of the heat across the plate.

 

Electrical connection between the heating block and the tungsten feed-through were made with standard copper flex insulated by ceramic beads (see figure 2.10).

 

2.2.6.2. Operating with the substrate heater

 

An Iso-tech DC power supply was used for the substrate heater. This was protected from overheating by an external fan fitted behind the heat exchanger at the rear of the unit.

 

Substrate heaters were operated by passing a current of 4 A through the Ni-Cr wire. To reach this current it was necessary to apply a potential between 15 to 20 V. This wide range of potential was dependent on the length of the Ni-Cr wire. Using these parameters and comparing with the literature, it suggests a maximum heater temperature of approximately 2400ºC.

 

The temperature was limited by the cement which brokedown at approximately 1000ºC. Before the breakdown of the cement vapour occluded from the pores of the cement burning the filament and contaminating the chamber.

 

2.2.6.3. Maintenance

 

The substrate heater had an average lifetime around 800 hours, after that the wires were very rigid and brittle. At this stage they were very easily damaged.

 

The periodic replacement avoided one of the sources of boron contamination, the pores of the cement where boron was stored. The heater block could not be cleaned of this contamination the only way of “cleaning” was the replacement of the heater block at regular time periods.

Figure 2.10. After a deposition run could be observed the structure of the substrate heater and a filament

 

2.2.7. Filaments

 

2.2.7.1. Construction of the filament

 

Filaments were formed using tantalum wire (0.25 mm of diameter) coiled around a 4 mm rod. It was found that ten turns of wire spanning 2 cm was the optimal geometry for the growing process. Larger wires sagged producing a poor growth and in some cases damaging the sample surface due to contact.

 

The chamber was equipped with two filaments, connected in series, to allow large area growth or the possibility of using more than one substrate per run.

 

When operating the chamber in single filament mode there was the option of a simple rewiring to employ the free filament if the first filament failed during the process.

 

The tantalum wire (Aldrich) used had a purity of greater than 99.9%, the levels of impurity are listed in the table 2.2

 

Element	concentration (p.p.m.)
Aluminium, Al	120
Copper, Cu	85
Tin, Sn	80
Nickel, Ni	35
Chromium, Cr	20
Vanadium, V	15
Magnesium, Mg	1

 

Table 2.2. Trace elements in the tantalum wire

Values taken from Supplier’s Certificate of Analysis

 

2.2.7.2. Operating conditions

 

The filaments were attached to tungsten rods by stainless steel clamps and the power supply has provided by a Variac variable resistor with an external voltmeter and amperometer. The current was maintained at 6.75 A during the whole process.

 

Low-pressure conditions could not be reached as they could have resulted in tantalum sputtering.

 

Under growth conditions, tantalum filaments were carburised increasing their resistance and necessitating an increase in the potential from the initial 20 V to 30V at the end of the run (see figure 2.11).

 

After each run the filaments were very brittle and were replaced before a new growth process was started.

 

Figure 2.11. The picture was obtained using an infrared filter through the glass window during the growing process. It could be observed the filament, substrate and heat substrate.

 

2.2.8. Substrates

 

2.2.8.1. Main features of the substrates

 

Silicon (Si) was the main substrate for growing diamond for electrochemical characterisation. For studies using diamond heat sinks the material was grown on quartz substrates.

 

The surface area of the substrates was variable between 1 cm² (1 cm ´ 1 cm) to  4 cm² (2 cm ´ 2 cm). Typical was 2 cm² (2 cm ´ 1 cm) because this allowed two samples to be placed in the chamber at the same time giving a better yield in the synthesis process. As the samples have been grown under the same conditions, later electrochemical characterisation will allow better experimental comparison.

 

The samples were prepared by cutting pieces of silicon from silicon wafers supplied from the microchip fabrication industry. The thickness of these wafers varied from 0.5 mm to 3 mm. Quartz samples were already cut from the supplier in samples of 1 cm² (1 cm ´ 1 cm) and 1 mm thick.

 

2.2.8.2. Pre-treatment of silicon surface before growing

 

The pre-treatment of silicon surface before growing was as follows:

 

1.      Abrasion of silicon surface. Samples were rubbed with diamond powder (2-3 mm).

 

2.      Cleaning diamond waste. Diamond grit waste was removed with 2-propanol

(IPA) soaked cotton sticks.

 

3.        Ultrasonic bath. Samples were finally treated in IPA ultrasonic bath for 15 minutes to remove very small diamond powder or possible impurities of the silicon.

 

2.2.9. Electrical contacts

 

Different types of electrical contacts have been used on the boron doped polycrystalline diamond samples during this study:

·        Indium/Gallium eutectic contact

·        Silver loaded epoxy resin contact

·        Three layers (titanium/platinum/gold) metal contact

·        Titanium under layer contact

Further details about the electrical contacts can be found in chapter 3. 

 

2.2.10. Typical growth conditions

 

Table 2.3 summarises the typical growth conditions for the diamond films and appendix A gives full details of the growth conditions for the films used in these studies.

Table 2.3. Typical deposition conditions for the hot-filament CVD reactor

 

2.3. Diamond Characterization

 

2.3.1. Analytical techniques to characterize CVD diamond films

 

Material characterisation in CVD diamond has been vital in sustaining progress in its development, especially in recent years. Nowadays, there is easy access to highly efficient, fast and extremely sensitive analytical tools to distinguish diamond from any other forms of non-diamond carbon. Such tools include X-ray photoelectron spectroscopy (XPS), X-ray diffraction, low energy electron diffraction (LEED), electron energy loss spectroscopy (EELS), secondary ion mass spectrometry (SIMS), and transmission electron microscopy (TEM). Secondary ion mass spectroscopy (SIMS) cannot distinguish between the different forms of carbon but is useful as a means to obtain a spatial distribution of different elements that may be present within a few atomic layers of a surface. The two characterisation methods mainly used in the present work to characterise the samples provided by Dr. M. Latto and Dr. P. May are discussed in the next sections: scanning electron microscopy and laser Raman spectroscopy.

 

2.3.1.1. Scanning Electron Microscopy

 

Scanning Electron Microscopy (SEM) is a powerful tool for the examination of bulk specimens. Electrons are generated by thermionic emission from a tungsten filament heated to temperatures approaching 3000ºC. These are then focused onto the sample surface. A detector then collects secondary electrons emitted from the surface of the specimen. An image is then formed on a cathode-ray tube (CRT) display.

 

The SEM pictures were recorded on a Hitachi Model 3200 Scanning Electron Microscope operating with an accelerating potential of 25 keV. A Mamiya camera loaded with rolls of 6 ´ 7 cm film was used to capture images. Other SEM studies were made in a JEOL JSM 5600LV Scanning Electron Microscope. This second SEM machine had digital image rather than a conventional camera, allowing more flexible treatments for the SEM pictures.  

 

SEM requires samples that are electrically conductive. In order to increase the surface conductivity of insulating samples, a layer of gold (Au) was deposited with an Edwards S150A Sputter Coater. This treatment was not required for the boron doped diamond samples as the doping made them sufficiently conductive.

 

Inspection of the SEM micrographs can provide valuable information on the topology of the diamond films:

 

·        Top view. The average grain size as well as the surface morphology of the deposited films

Figure 2.12. Randomly oriented HFCVD diamond film (sample B141a)

 

·        Cross-sectional View. The film thickness and roughness

 

Figure 2.13.  Cross-section SEM picture of a thin diamond film

 

 

However, such images, in themselves, do not prove whether the films are diamond.

 

Additional SEM pictures can be seen in the next pages:

 

Figures 2.14 to 2.16 shows SEM images of a diamond sample using different resolutions. The size, quality and morphology of the diamond crystals in the film can be estimated.

Figure 2.14. A continuous HFCVD diamond film. (Sample B117)

 

Figure 2.15. A continuous HFCVD diamond film. (Sample B117)

 

Figure 2.16. A continuous HFCVD diamond film. (Sample B117)

 

Figures 2.17 to 2.19 show SEM images of a sample where the diamond growth had been stopped after 5 hours and 35 minutes. Figure 2.17 shows the surface morphology at the centre of the sample where the diamond film was continuous and well defined facets. Figures 2.18 and 2.19 show the surface morphology at the edges of the sample where the growth had not progressed far beyond the nucleation stage, leaving individual crystals on the silicon substrate.

Figure 2.17. A continuous diamond film with well defined facets. (Sample B142a)

 

Figure 2.18. Incomplete surface coverage at the edge of a thin diamond film. The silicon substrate is visible behind the diamond crystals. (Sample B142a)

 

Figure 2.19. The reaction was stopped at the stage where the individual crystals have started to coalesce to form a continuous film. (Sample B142a)

 

The quality and morphology of the films was comparable to the same one in the industrial grown samples. Figures 2.20 to 2.22 show an industrial free-standing diamond film that was grown for an optical application in the aerospace industry using microwave plasma assisted CVD (MPACVD) ⁴⁰. The industrial film was grown continuously for ten days yields a much thicker sample (175 mm) than the HFCVD diamond films. It shows lager facet sizes (approximately 50 mm)

 

Figure 2.20. Industrial undoped MPACVD diamond

 

Figure 2.21. Industrial undoped MPACVD diamond.

 

 

Figure 2.22. A cross-section of a free-standing industrial undoped MPACVD diamond.

 

Figures 2.23 and 2.24 show two different samples grown on quartz substrate. The quality and morphology of the films was comparable to same one in the silicon substrate samples.

 

Figure 2.23. Randomly orientated undoped diamond sample on quartz substrate (sample Q1)

 

During the cooling stage of the diamond sample if the quartz substrate contracts in a non-homogeneous process the result could be the cracking of the diamond film (see figure 2.25). Other side effect could be the appearance of pinholes in the diamond structure (see figure 2.26)

 

Figure 2.24. Randomly orientated undoped diamond sample on quartz substrate (sample Q2)

 

Figure 2.25. Crack undoped diamond film on quartz substrate (sample Q3).

 

Figure 2.26. Pinhole in the undoped diamond structure grown on quartz substrate (sample Q3)

 

2.3.1.2. Laser Raman Spectroscopy

 

Laser Raman Spectroscopy (LRS) is the most widely used technique to identify the characteristic energies of the chemical bonds or to distinguish between different phases within the same material. For this reason it is a decisive tool for establishing that the film is indeed diamond and for providing some measure of the film quality. The Raman effect is an interaction between monochromatic light, such as from a laser source, and the chemical bonds within a specimen. Thus when the laser is irradiated on the specimen, a small number of photons may excite molecular vibrations in the specimen. Consequently, these photons will be scattered with a slightly lower energy.

 

Raman spectra were taken with a Renishaw Raman Imaging Microscope and Spectrometer with a spectral resolution of 1 cm^-1. An argon gas laser with a wavelength of 488 nm (green light) was used to provider the incident laser beam.

 

The Raman spectrum of natural diamond shows a sharp, single peak centred at a wavenumber of approximately 1332 cm^-1. This feature also dominates the Raman spectra of high quality, polycrystalline diamond film grown by CVD methods. However, additional peaks may be observed in the spectra which are characteristic of non-diamond contamination, depending on the deposition conditions (such as films grow with high methane concentration in gas phase). The Raman spectrum of graphite shows a broad feature centred on 1580 cm^-1. When characterising CVD diamond films, the observation of any broad resonance at higher wavenumber is generally taken to indicate the presence of graphite-like non-diamond phases containing sp²-bonded carbon atoms.

 

Figure 2.27. Raman spectrum of a type IIbª natural diamond.

 

The full width half maximum (FWHM) of the 1332 cm^-1 Raman line is another measure of film quality; good quality CVD diamond films normally produces peaks with a FWHM of 4-10 cm^-1, whilst linewidths close to those natural diamond if the films are non-continuous, giving rise to separate diamond crystals.

 

Some of the Raman spectra of the diamond samples used in this study are presented here:

 

Sample b140a presents a shoulder band about 1550 cm^-1 indicating the weak presence of graphite.

Figure 2.28. Raman spectrum of boron doped diamond film (B140a)

 

Sample b140b no presents a shoulder band about 1550 cm^-1 such as b140a sample, indicating that graphite is almost neglectled.

 

Figure 2.29. Raman spectrum of boron doped diamond film (B140b)

 

Sample Q1 (undoped diamond grown on quartz substrate) presents a shoulder band about 1550 cm^-1 indicating the weak presence of graphite when is compared against the sharp peak for approximately 1333 cm^-1 (see figure 2.30).

 

Figure 2.30. Raman spectrum of undoped diamond film grown on quartz substrate (Q1).

 

2.4. Summary

 

The system employed to grow the diamond samples used in these studies has been explained in detail. Furthermore analytical techniques used in the identification and characterisation of the diamond samples have been presented. Results were discussed. Good quality undoped and boron doped diamond samples have been achieved.

 

2.5. References

 

¹              S. Tennant, Phil. Trans. R. Soc. Lond. A, 1797, 87, 123.

²              F. P. Bundy, J. Geophys. Res., 1980, 85, 6930.

³              J. E. Field, 'The properties of natural and synthetic diamond', ed. Academic, 1992.

⁴              F. P. Bundy, H. T. Hall, H. M. Strong, and R. H. Wentforf, Nature, 1955, 176, 51.

⁵              W. G. Eversole, US Patent nos 3030187, 3030188, 1958.

⁶              B. V. Deryagin, D. V. Fedoseev, V. M. Lukyanovich, B. V. Spitsyn, A. V. Ryanov, and A. V. Lavrentyev, J. Cryst. Growth, 1968, 2, 380.

⁷              J. C. Angus, H. C. Will, and W. S. Stanko, J. Appl. Phys., 1968, 39, 2915.

⁸              D. J. Poferl, N. C. Gadner, and J. C. Angus, J. Appl. Phys., 1973, 44, 1418.

⁹              B. V. Deryagin, B. V. Spytsyn, L. L. Builov, A. A. Klochov, A. E. Gorodetskii, and A. V. Smolyanimov, Dokl. Akad. Nauk. SSSR, 1976, 231, 333.

¹⁰             B. V. Spytsin, L. L. Bouilov, and B. V. Derjaguin, J. Cryst. Growth, 1981, 52, 219.

¹¹             S. Matsumoto, Y. Sato, M. Tsutsumi, and N. Setaka, J. Mater. Sci., 1982, 17, 3106.

¹²             S. Matsumoto, Y. Sato, M. Kamo, and N. Setaka, Jpn. J. Appl. Phys. Part 2 - Lett., 1982, 21, L183.

¹³             M. Kamo, Y. Sato, S. Matsumoto, and N. Setaka, J. Cryst. Growth, 1983, 62, 642.

¹⁴             Y. Saito, S. Matsuda, and S. Nogita, J. Mater. Sci. Lett., 1986, 5, 565.

¹⁵             P. W. May, N. M. Everitt, C. G. Trevor, M. N. R. Ashfold, and K. N. Rosser, Appl.Surf. Sci., 1993, 68, 299.

¹⁶             D. G. Goodwin and J. E. Butler, 'Handbook of industrial diamonds and diamond films', ed. M.A. Prelas, G. Popovici, and L. K. Bigelow, 1997.

¹⁷             P. K. Bachmann, H. J. Hagemann, H. Lade, D. Leers, F. Picht, D. U. Weichert, and H. Wilson, Mater. Res. Soc. Symp. Proc., 1994, 339, 267.

¹⁸             W. Muller-Serbet, E. Worner, F. Fuchs, C. Wild, and P. Koidl, App. Phys. Lett., 1996, 68, 759.

¹⁹             S. A. Kajihara, A. Antonelli, J. Bemhole, and R. Car, Phys. Rev. Lett., 1991, 66, 2010.

²⁰             S. Koizumi, M. Kamo, Y. Sato, H. Ozaki, and T. Inuzuka, App. Phys. Lett., 1997, 71, 1065.

²¹             N. Fujimori, H. Nakahata, and T. Imai, Japan J. Appl. Phys., 1990, 29, 824.

²²             K. Miyata, K. Kumagai, K. Nishimura, and K. Kobashi, J. Mat. Res., 1993, 8, 2845.

²³             J. Mort, D. Kuhman, M. Machonkin, M. Morgan, F. Jansen, K. Okumura, Y. M. LeGrice, and R. J. Nemanich, App Phys Lett, 1989, 55, 1121.

²⁴             H. Shiomi, Y. Nishibayashi, N. Fujimori, and K. Kobashi, Japan J. Appl. Phys., 1991, 30, 1363.

²⁵             S. A. Grot, C. W. Hartfield, G. S. Gildenblat, A. R. Badzian, and T. Badzian, App. Phys. Lett., 1991, 58, 1542.

²⁶             X. K. Zhang, J. G. Guo, Y. F. Yao, R. Wang, G. M. Chen, W. K. Zhou, and S. Yu, Appl. Phys. A, 1993, 56, 425.

²⁷             R. Locher, J. Wagner, F. Fuchs, M. Maier, P. Gonon, and P. Koidl, Diam. Relat. Mater., 1995, 4, 678.

²⁸             R. Samlenski, C. Haug, R. Brenn, C. Wild, R. Locher, and P. Koidl, Diam. Relat. Mater., 1996, 5, 947.

²⁹             R. Locher, C. Wild, N. Herres, D. Behr, and P. Koidl, App Phys Lett, 1994, 65, 34.

³⁰             S. Jin and T. D. Moustakas, Appl. Phys. Lett., 1994, 65, 403.

³¹             W. Tsai, M. Delfino, D. Hodul, M. Riazat, L. Y. Ching, G. Reynolds, and C. P. Cooper, IEEE Electron Device Lett., 1991, 13, 126.

³²             F. Fontaine, 'Doping of diamond by ion implantation', Unpublished review, 1997.

³³             R. G. Compton, F. Marken, C. H. Goeting, R. A. J. McKeown, J. S. Foord, G. Scarsbrook, R. S. Sussman, and A. J. Whitehead, Chem. Commun., 1998, 1961.

³⁴             C. H. Goeting, F. Marken, R. G. Compton and J. S. Foord, Chem. Commun., 1999, 17, 1697.

³⁵             C. H. Goeting, J. S. Foord, F. Marken, and R. G. Compton, Diam. Relat. Mater., 1999, 9, 824.

³⁶             K. B. Holt, J. Del Campo, J. S. Foord, R. G. Compton, and F. Marken, J. Electroanal. Chem. , 2001, 513, 94.

³⁷             A. J. Saterlay, S. J. Wilkins, C. H. Goeting, J. S. Foord, R. G. Compton, and F. Marken, J. Solid State Electrochem., 2000, 4, 383.

³⁸             A. J. Saterlay, S. J. Wilkins, K. B. Holt, J. S. Foord, R. G. Compton, and F. Marken, J. Electrochem. Soc. , 2001, 148, E66.

³⁹             J. D. Wadhawan, F. J. Del Campo, R. G. Compton, J. S. Foord, F. Marken, S. D. Bull, S. G. Davies, D. J. Walton, and S. Riley, J. Electroanal. Chem., 2001, 507, 135.

⁴⁰             D. A. Tossell, M. C. Costello, A. P. Webb and K. C. Vanner, Pure and Appl. Chem., 1994, 66, 1335.

⁴¹          W. J. P. Van Enckevort, 'Physical, chemical and microstructural characterisation and properties of diamond in synthetic diamond: emerging CVD science and technology', ed. K. P. Spear and J. P. Dismukes, J. Wiley and sons Inc., 1994.