A wide variety of energetically assisted methods have been employed to grow diamond films at low pressures. Common features of the processes include the presence of atomic hydrogen, energetic carbon containing fragments and high surface mobilities. Some understanding of the molecular processes taking place during nucleation and growth of diamond has been achieved, but detailed molecular mechanisms are not known with certainty. Application of vapor grown diamond for abrasive grit, tool coatings and wear resistant surfaces can be expected shortly. However, the use of vapor grown, crystalline diamond in optical applications or as active semiconductor elements will require further control over surface roughness and crystalline quality. Related research has led to the discovery of a new class of materials, the so-called "diamondlike" phases. Two types of diamondlike materials may be distinguished, namely, the diamondlike hydrocarbons and the diamondlike carbons. These materials possess exceptional hardness, smoothness and chemical inertness. They show promise as combined anti-reflection and abrasion resistant coatings on optical elements, as protective coatings on magnetic and optical disks, as diffusion barriers and for photo-lithographic applications.
Chemically vapor deposited diamond films have a potential for optical applications but so far only translucent films have been produced. The reason for the low transparency is considered to be related to the polycrystalline nature of the films and to defects. To elucidate the problem and indicate ways of improving film quality, the growth mechanism of CVD diamond must be understood. In microwave plasma assisted CVD the plasma is induced by microwaves in a mixture of hydrogen and methane. A high concentration of atomic hydrogen, much above the thermal equilibrium concentration, helps to prevent codeposition of graphite. The growth of diamond occurs in a narrow temperature range (900-1000°C) as a result of the interplay of several phenomena.
Results of research on diamond films grown by the hot filament chemical vapor deposition process are discussed. The parameters for film deposition have been surveyed and the conditions for routine and reproducible film formation established for our deposition system. These were: 800°C substrate temperature, 52-78 sccm flow rate, 5x103 Pa deposition pressure and 99.5% H2, 0.5% CH4 gas composition. Characterization of the deposited films has been accomplished with scanning electron microscopy (SEM), x-ray diffraction (XRD), Auger electron spectroscopy (AES), electron energy loss spectroscopy (EELS) and Raman spectroscopy; and the presence of the diamond phase was verified. Initial depositions on Si and Al203 substrates resulted in individual diamond particles showing the distinct diamond morphology. These particles, when examined by Raman microprobe spectroscopy, displayed the diamond spectra. Subsequently, continuous diamond films were formed after pretreating the substrates by rubbing with 1 pm diamond abrasive before deposition. Films, all shown to be diamond, grown on fused silica, polycrystalline SiC and different orientations of single crystal Si (all pretreated), exhibit very similar surface topography and x-ray diffraction patterns. Additionally, x-ray diffraction shows no preferred orientation in the films. Mechanical surface measurements on a 5 pm thick film grown on Si show that the films possess an average surface roughness of 0.4 μm and a peak-to-valley roughness of 2.5 μm. Films deposited on optically clear fused silica substrates could be seen to be water white, suggesting no significant optical absorption. Considerable optical scatter, however, is present due to the roughness of the film surface. Deposition rates were of the order of 0.1 μm/hr, with the fastest apparent growth occurring on pretreated polycrystalline SiC substrates.
Diamond-like Carbon films synthesized by RF plasma-enhanced chemical vapour deposition were systematically characterized for their structural, optical and mechanical properties. The process parameters were optimized by the use of orthogonal tables to get the 'diamond-like' properties. DLC films were demonstrated to be efficient and durable antireflection coatings for IR devices. Plasma diagnostics were carried out by Langmuir probe method and Optical Emission Actionometry for Oxygen, Air and Water plasmas to give an understanding of the DLC etch mechanism.
Type IIa diamond crystals were implanted with boron ions with or without prior carbon ion implantation. The samples were kept at liquid nitrogen temperature during both implantation steps. A strong near-edge optical absorption band appeared after implantation, and partially recovered during annealing at 800 °C. For the highest B implantation fluence, optical absorption peaks at 2800 to 3000 cm-1 were observed that were in the same vicinity as the absorption peaks attributed to substitutional boron atoms in natural p-type diamond. Electrical measurements for three of the samples demonstrated well-defined activation energies that could be associated with hopping conduction and/or activation of B dopant atoms. This work shows that p-type doping in diamond by boron ion implantation is feasible, using a suitable combination of low temperature implantation and subsequent annealing.
Metastable, amorphous carbon films, also referred to as diamond-like carbon (DLC), are prepared in a large variety of deposition conditions. Depending upon the conditions, DLC films may contain large amounts of hydrogen, and are referred to as hydrogenated amorphous carbon. The properties of DLC are strongly dependent on the preparation conditions and upon the amount of hydrogen incorporated in the film. Due to its extreme hardness, DLC can be used as a wear protective coating, while its chemical inertness to acids and alkali's make it suitable for protection against chemical attack. The optical transparency of the DLC over a large region of the spectrum makes it useful as a protective coating of optical components. However, hydrogenated DLC is usually also characterized by very high compressive stresses. Its application as a protective coating therefore requires strong adhesive bonds to the coated surface. The paper reviews structural and optical properties of DLC films and means of improving their adhesion to metallic surfaces.
Diamondlike Carbon (DLC), also known as amorphous hydrogenated carbon (or a:C-H), is a hard semitransparent material. Since it is amorphous and has no grain boundaries, it is proposed for use as a coating on infrared optical surfaces to protect them from the envi-ronmentally adverse chemicals such as the acids, salt water, etc., as well as particle impact. Moisture penetration studies of DLC, usicgthe Variable Angle Spectroscopic Ellipsometry (VASE), have been presented elsewhere 194'. In this paper, we first briefly mention sample preparation, done using various combinations of RF power and gas pressure in a plasma deposition chamber. Then, results of using an ultraviolet-visible (UV-VIS) absorbance spectrometer are presented. From the absorbance we calculated the band gap using the Tauc plots, and found values ranging from 0.2 to 1.25eV, depending on deposition condition. Secondly, a series of DLC samples were prepared using a Kaufman type ion beam deposition system. Using VASE, we have studied these samples after bombarding them with various fluences of Fl, and analyzed the data using two Lorentz oscillators to represent the optical properties. We have also shown the correlation between the optical properties and the amount of hydrogen present within the samples, as measured by proton recoil experiments.
Diamond particles have been produced by inductively coupled radio frequency plasma assisted chemical vapor deposition. Analysis indicates that particles having a thin graphitic surface, as well as diamond particles with no surface coatings have been deposited. Scanning electron microscopy analysis shows that particles are deposited on a pedestal which Auger spectroscopy indicates to be graphitic. This phenomenon has not been previously reported in the literature.
Remote plasma enhanced chemical vapor deposition has been used to investigate the nucleation and growth of diamond and diamond-like films on Ni surfaces. The primary results center around hydrogen dilution experiments. Hydrogen dilution when using the polycrystalline Ni substrates tends to reduce the growth rate and increase the electrical resistivity of the films (-107 Ω-cm), it is found that at even higher hydrogen dilutions (greater than 98% H2) the films become semicontinuous with sparse and sometimes no nucleation occurring. These films, like the ones grown at lower hydrogen dilution, do not show a 1332 cm-1 diamond Raman line but show graphitic and disordered carbon features. An attempt to grow heteroepitaxial diamond on Ni(111) surfaces under conditions of high hydrogen dilution (100:1) produced a sample with oriented hillocks which are heteroepitaxially in registration with the substrate. Raman analysis showed lines characteristic of graphite and disordered carbon with an additional line at 1398 cm-1. Transmission electron microscopy produced a diffraction pattern with the lattice spacing and symmetry of epitaxial graphite with some faint polycrystalline rings.
Raman spectroscopy has been used to examine diamond thin films produced by various CVD processes. The Raman spectra exhibit, features which suggest that the films are composites of diamond (sp3) and graphite-like (sp2) bonding. A brief outline of Raman scattering from composites is presented. A first attempt at modeling these types of films using composites of diamond and graphite powders is reported. It is found that the Raman features associated with sp2 bonding in the films do not correlate well to features exhibited by microcrystalline graphite.
Raman microprobe studies of individual microcrystals of diamond and of thin diamond films deposited by the hot-filament chemical vapor deposition (CVD) method are focused on the determination of the purity of the diamond phase and on the extent and nature of defects of the diamond structure. The findings are discussed in relation to deposition parameters, growth mechanisms, and diamond morphology. The specimens consisted of single microparticles of sizes 3 to 40 µm, particle clusters, and continuous polycrystalline films of 3 to 8 μm thickness grown on silicon substrates. The interpretation of the results is based on the line shape, line width, and frequency position of the diamond line nominally at 1333 cm-1 Raman shift, as well as on other characteristic Raman bands in the region 1300 to 1600 cm-1 attributed to graphitic carbon components. Examined also are the relationship of the spectral background signal to the signal of the Raman features. Luminescence emissions arising from either structural imperfections or substitutional impurities in the diamond lattice are observed. A luminescence band centered around 738 nm (1.68 eV), attributed to either the neutral lattice vacancy in diamond, or possibly a silicon pair substitution in the diamond lattice, widely varied in intensity among the samples analyzed. The observation of this photoluminescence band is correlated with results from concurrent cathodoluminescence measurements.
Thin diamond-like and diamond films, grown by remote plasma-enhanced CVD (RPECVD) and by plasma CVD, were characterized using optical microscopy, elastic recoil detection spectroscopy (ERD), and Raman scattering. The H concentration of the films was measured by ERD and was related to the growth parameters and to the quality of the films as determined by Raman scattering. The H content of samples grown by RPECVD at the Research Triangle Institute (RTI) increased with decreasing growth temperature, varying from 6 at% H at growth temperatures from 500-720°C, to 25 at% H at a growth temperature of 20°C. The characteristic Raman frequency of natural diamond, 1332 cm-1, was observed for samples grown by Crystallume Corp. and by General Electric Corp. The samples obtained from Crystallume showed circular "bull's-eye" features by optical microscopy, and the width of the 1332 cm-1 peak was broad, indicating highly strained crystallites. For a sample from G.E., most of the film was good quality diamond, as indicated by a sharp 1332 cm-1 diamond frequency; the bottom (substrate) part of the film contained more H than the top part.
While the preponderance of the mechanical, optical, and electronic properties of natural diamond have been known for over a decade, only recently has artifact diamond in technologically useful form factors become an exciting possibility. The advent of sacrificial, lattice matched crystalline substrates provides the basis not only for semiconducting applications of diamond, but for optical mirrors, lenses, and windows as well. As a semiconductor, diamond has the highest resistivity, the highest saturated electron velocity, the highest thermal conductivity, the lowest dielectric constant, the highest dielectric strength, the greatest hardness, the largest bandgap and the smallest lattice constant of any material. It also has electron and hole mobilities greater than those of silicon. Its figure of merit as a microwave power amplifier is unexcelled and exceeds that of silicon by a multiplier of 8200. For integrated circuit potential, its thermal conductivity, saturated velocity, and dielectric constant also place it in the premier position (32 times that of silicon, 46 times that of GaAs). Although not verified, its radiation hardness should also be unmatched. Aside from its brilliant sparkle as a gemstone, there has been little use of diamond in the field of optics. Processing of the diamond surface now appears to be as simple as that of any other material --albeit with different techniques. In fact, it may be possible to etch diamond far more controllably (at economically viable rates) than any other material as the product of the etch is gaseous and the etched trough is self-cleaning. Other properties of diamond make it an ideal optical material. Among them are its unmatched thermal conductivity, its extremely low absorption loss above 228 nanometers, and unmatched Young's modulus, Poisson's ratio, tensile strength, hardness, thermal shock, and modulus of elasticity. If the recently-found mechanisms by which erbium impurities in III-V junctions can be made to "lase" are indeed applicable to indirect gap semiconductors, then even efficient diamond lasers at attractive portions of the spectrum become a distinct possibility.
The commercial availability of gem-quality synthetic diamonds, and the expanding research on various methods of diamond synthesis, have made it necessary for the gem diamond trade to find simple ways to identify synthetic diamonds. Using both magnification and various types of spectroscopy and luminescence techniques, a number of properties that are characteristic of synthetic diamonds have been found.
Large synthetic diamond single crystals, in sizes up to 1.4 ct, are produced on 4 commercial basis for some industrial application fields by Sumitomo Electric. The crystals are yellow colored type Ib stones which contain lower amounts of nitrogen (up to about 100 ppm) dispersed through the crystal structure in the form of singly substituting atoms. The impurity controlled type lb crystals have the highest thermal conductivity which is equivalent to that of pure type IIa crystals. Optical and thermal properties of diamond crystals are strongly affected by dispersed impurities. We studied the kinds of dispersed impurities and amounts of those impurity atoms in our synthesized crystals by SIMS. A relation of the thermal conductivities and the nitrogen concentrations of the crystals was examined. The state of nitrogen impurity in the crystals could be transformed by electron irradiation and subsequent high temperature annealing. The reaction rates for the transformation Ib nitrogen to type IaA aggregates and differences in crystal growth sectors have been studied. Vapor phase deposited diamond films are hopeful candidates for optical application of diamond. Preliminary spectroscopic analysis has been done for the free standing polycrystalline films.
When, in the course of advancing the state of the art, one slams into a material barrier to the construction of one's appointed gadget, it is customary, and at times mandatory, to drop to one's knees and pray to DARPA for deliverance. Deliverance in the form of the right stuff. Something superbly strong, something utterly transparent, something remarkably light, and with dielectric properties rivaling a perfect vacuum. Ideally, this something should also be bulletproof, better at conducting heat than a silver spoon, insoluble in boiling acid, radiation hard, non-toxic, and cheap. Well, historically, nine out of ten isn't bad for a start. Indeed, it's better than nothing. So let me begin with some history. What we today call solid state physics began not as science but as technology. Victorian low technology to be exact. The first practical solid state electronic devices, demonstrated by Ferdinand Braun* at Leipzig on November 14, 1876, were based neither on theory nor on synthesis, nor on crystal growth. For in those days these things existed not. They were instead dug up, mined as lead ore. The performance of the galena (PbS) cat's whisker diode was marginal; it was rapidly superseded by the first, worse vacuum tube . So also, early infrared optics of rock salt gave way to synthetic crystals. But those early artifacts' performance demanded a physical explanation, and after a brief hiatus, in order for Willard Gibbs to break ground by inventing thermodynamics, the modern theory of solids arose to provide it. It all stemmed from the enterprise of explaining first the optical properties and then the electronic bebavior of crystals found in rocks. Today diamond, along perhaps with Iceland spar, remains the last optical material to be technically exploited as it is found in nature. It is a barbarous relic, a throwback to high technology's dim Neolithic past. For nowadays we are used to thinking about synthetic optical materials, like zinc sulfide or selenide, as being mature, as being just so much up market optical glass. They're stock items--you pay the money and the stuff shows up in big transparent slabs. That wasn't true even in 1970. Then, as with diamond, there was essentially none to be had. The first measurements of the non-linear refractive index of ZnS had to be made on a sample of zinc ore from a Harvard museum--a pale green crystal of sphalerite whose optical quality then represented the state, not of the art, which was then non-existent, but of nature at her best. So what can we do today when we're interested in a material that we are still learning to make as well as nature does? There are two maxims of the earth sciences that could be of service in the near term to the area of materials science that we are gathered here to discuss. One is: "The best geologist is the one who has seen the most rocks"; the other is more germane still: "Rocks are just ceramics that happen to have been made by God." Diamond is no exception, for today more than ever before, ceramics aren't just common clay. So, naturally, there presently exist more shapes, forms, growth morphologies, surface textures, and degrees of optical quality in the diamonds of a good mineralogical collection than have, as of yet, been synthesized. Good, bad, and ugly, these varieties of natural diamond fall into two categories: the ones we have seen already synthesized and the ones that we will see. And the more we can find out about the natural history of the latter, the sooner we will see them synthetically reproduced.
Our theoretical results have demonstrated the usefulness of diamond films for improving the survivability of thin film metal mirrors against intense x-ray. Essentially, the low-Z, high thermal conductivity diamond film acts as an efficient heat sink, thus reducing the temperature in the metal film. Specifically, for an aluminum/fused silica mirror, our analysis has shown that interposing a 1-μm thick diamond film between the thin film aluminum and SiO2 substrate increases the aluminum melt fluence--from a 1 KeV black body, 10 ns x-ray pulse--by a factor of 6. In addition, we have found that for pulse widths on the order of about 10 ns, a diamond film thickness of about 1 μm appears to be most effective in reducing the aluminum temperature.
This paper reviews existing and new applications of single crystal diamond, both natural and synthetic, in optical science. The traditional application is as transmissive components, making use of the very wide spectral transmission range, high thermal conductivity, and chemical inertness of diamond. Diamond windows for corrosive environments are well known; diamond surgical endoscope components are under development; and the use of sharpened diamonds as combined surgical cutting instruments and light pipes for internal illumination of the edge is commercial reality. The superb ability of diamond to conduct heat, combined with its very low thermal expansion coefficient makes it suitable for the transmission of high power laser energy, though there is a problem currently being addressed of a high surface reflection coefficient. It is very probable that CVD diamond-like films will form good anti-reflection coatings for diamond. In new applications, the technology of making diamond lenses is being developed. The use of diamond as a detector of ionising radiation is well known, but recent work shows its possibilities in thermoluminescent as well as conduction and pulse counting modes. There are further possibilities of using diamond for the detection and measurement of optical radiation. Examples are low, medium, and high intensity far ultraviolet (< 225 nm) and very high intensity near ultraviolet and visible light from excimer, dye, or argon lasers. Diamond is very radiation resistant! Sensitivities, response times and impurity trap levels have been measured and appropriate diamonds can be synthesised. The use of diamond as fast opto-electronic switches has been reported in the literature and the mechanical and thermal design of diamond "heat sink" substrates for semiconductor laser diodes is advancing rapidly.
This paper describes the development and testing of diamond-like carbon coatings using an RF glow discharge technique. Methods used to nucleate the diamond phase to reduce the growth of graphitic impurities and to improve the deposition process are discussed. Various optical, mechanical and medical applications of the coating are included.
Polycrystalline diamond films were deposited by the microwave-plasma chemical-vapor-deposition (CVD) on Si substrates using a mixture of methane and hydrogen for the source gas. In the morphology study of diamond films using a scanning electron microscope (SEM), it was found that upon increasing the methane concentration (hereafter denoted by c in units of vol%), the surface texture changed discontinuously from (111) to (100) at around c=0.4%, and gradually from (100) to microcrys-talline above c=1.2%. The diamond-Si interfaces and the defect struc-tures in the films were investigated by transmission electron micro-scopy (TEM). The film growth process was investigated by SEM, and it was found that the appearance of small grains and the formation of well-defined diamond faces took place repeatedly with time during the CVD synthesis. The film morphology of boron-doped diamond films on Si substrates and on non-doped diamond films were also presented.
Optics are an integral part of lasers, whose size and power output have grown greatly in the last several decades. As the power levels grow, the need for compatible high-power optical materials has also grown. Laser designers know that the limits on intracavity and extracted power are set by the damage to the laser optics rather than by what can be generated in the laser medium. The limits on laser power imposed by materials properties and their processing apply to high peak and average power systems and to reflecting and transmitting optics.
Two concepts have been proposed for using transmissive diamond optics in the laser cavity of an FEL in order to reduce the cavity length: (1) a resonant reflector stack and (2) an intracavity divergent lens. The limitations of both designs will be discussed. Diamond was selected as the optical material for its low coefficient of thermal expansion, low thermal optic coefficient, low optical absorption coefficient, and high thermal conductivity at liquid nitrogen temperatures.
The many unique physical properties of diamond make it useful as a thin film coating for laser optics. We have calculated the laser induced thermal stress resistance for diamond and other optical materials. The calculated stress resistance for diamond is orders of magnitude higher than any other material and, therefore, diamond films should have a higher laser damage threshold. Calculations also indicate that diamond film, because of its high thermal conductivity, exhibits tolerance for isolated impurity inclusions. Polycrystalline diamond films were deposited on silicon substrates using a d.c. plasma enhanced chemical vapor deposition process. The films were characterized by Raman and optical absorption spectroscopy and by ellipsometry. Laser induced damage thresholds of diamond film windows and films on silicon substrates were measured for single pulses of 532 nm and 1064 nm laser radiation. The measured damage thresholds for diamond windows are 6.0 J/sq.cm (300 MW/sq.cm) at 532 nm and 12.4 J/sq.cm (620 MW/sq.cm) at 1064 nm. For diamond on silicon, the damage thresholds are 3.65 J/sq.cm (182 MW/sq.cm) at 532 nm and 14.4 J/sq.cm (720 MW/sq.cm) at 1064 nm. These values compare favourably with those for other common materials used as optical coatings.