SPECIAL REPORT/ADVANCED CERAMICS: Nanodiamond: From Theory to Practice

Since the early 1900s, chemists, physicists and industrialists have had a need for a multifaceted material. Diamond is one such material, with applications ranging from jewelry and machinery to medicine. In the last five to seven years, diamond has evolved as its size has decreased from centimeters, millimeters and microns to nanometers.

Polycrystalline Diamond

Diamond can be found in two basic forms-monocrystalline and polycrystalline. We will focus here on the polycrystalline form. In the 1960s, DuPont was looking for new applications for its explosive products. The company calculated that the processing time required to make diamond could be decreased with an explosive shockwave.

One problem with this method was the question of how to control the reversible reaction due to the high energy required by the explosive mixture. To solve the problem, DuPont used a copper core within its explosive device to cool the reaction after the synthesis of the diamond.

Today, the process methodology to synthesize polycrystalline diamond involves the use of crystalline graphite and a catalyst packed around a copper tube. The copper tube is encased in a stainless steel tube that is in turn encased in a hardened steel tube packed with explosives. The tubes are then projected at a target. Upon striking the target, the resultant explosion rapidly converts the graphitic carbon to diamond. Because of the rapid cooling due to the copper core, conversion to monocrystalline diamond is stunted and a resultant polycrystalline crystal substructure is formed.

The poly-crystallites range in sizes of between 22 and 66 angstroms. Above  66 angstroms, the poly-diamond begins to lose its identity as a polycrystalline diamond. Above 108 angstroms, it acts much like monocrystalline diamond.

Ultra-Detonated Diamond

Ultra-detonated diamond (UDD) is produced by the detonation of an explosive mixture in an oxygen-deficient atmosphere. UDD undergoes a general shock loading, meaning a rapid rise in pressure and temperature from the explosive process. The general form of production involves a condition of 30 GPa at 1000 to 2300°C. The result is thermodynamically stable diamond (carbon) powder that contains agglomerates of 50 nm to 5 um and individual crystallites of 2-7 nm. A small fraction of these crystallites can be up to 10 to 50 nm crystals.

Three methods, all involving an oxygen-deficient process, can be used to produce UDD particles. The first process involves an explosive mixture (small charges) that is detonated in a reaction chamber in the presence of a blend of inert gases. (The chamber design and gas loading is a propriety mix.) The result is UDD diamond with a percentage of primary particle and secondary particles. The surface composition is affected by the gas composition and chamber design. In general, this process produces the highest concentration of complex surface carbons of different species.

The second process uses a modification of the inert gas process, where an explosive mixture of TNT and hexogen is detonated in a reactor chamber filled with inert gases and water. The main difference involves the secondary particle size and its surface composition. Water cools the explosive gases more rapidly, producing a more-ordered carbon surface structure that makes it possible to predict results in some applications.

A modification of the first two processes, the third method uses a different blend of explosives, a large reactor chamber and ice. The explosive charge is encapsulated in a solid block of ice and is detonated under water with an active water spray. The resultant product provides for a different surface group on the secondary particle. In terms of creating a more- ordered structure, this particular method provides a higher degree of ordered surface carbons and chemistry. Predictability is a key to reliability. For some processes that utilize nanodiamond, this predictability is essential to the desired outcome.

UDD powders contain both primary and secondary particles. The primary particles feature crystallite sizes that average 4 nm and range from 2-7 nm. The secondary particles are clusters of the primary particles with an outer coating of graphite, grapheme and other forms of carbons such as carbon onions and tubes, as well as some quasi-amorphous carbon (see Figure 1).

Applications

The recent development of nanodiamond (ND) production at a commercial level makes it possible to utilize ND's physicochemical properties in the field of nanotechnologies (see Figure 2). Nanodiamond particles synthesized by explosive detonation have drawn considerable attention in the past few years.

Nanodiamonds (unlike monocrystalline diamond) are hydrophilic (will hydrate with fluids including water), owing to their high surface-to-volume ratio. The ND core consists of the sp³ orbital crystal structure, and the surface has the sp² orbital structure along with hybrids of sp²/sp³. Therefore, the surface contains a complex arrangement of carbons throughout the cluster.

From an intellectual perspective, nanoparticles open new frontiers in micromachines, medical applications and industries such as electronics, coatings and optics. For example, ND can be used in any application where superb finishes are required, such as in the polishing of optics that are used in telescopes, laser windows, acoustic wave guides and sapphire wafers. Next-generation diamond cutting tools are another application. With a more uniform and ordered microstructure, ND can be used to increase the wear resistance of drills so cutting tools can be sharper and produce better finishes.

Nanodiamond can also be used in the densification of other materials made from aluminum, boron nitride and magnesium oxide/niobium oxide combinations. Some examples would be use as a fully dense but lightweight armor or as filler for cermets in the structural ceramics industry. Fixed polishing applications are another opportunity. ND can be used in polishing films and pads for fiber optics applications, or to increase the useable area of silicon carbide wafers by removing defects.

Because of its nano-size and ability to "carry" medications into a cell without inflammation, ND is becoming a promising new material in medicine. A biocompatible material, ND has been used as a carrier (delivery system) of medications for treating tumors.

Another application involves advanced chemical mechanical planarization (CMP) for polishing sapphire, SiC, GaAs, GaN and other compound semiconductors to unprecedented levels of precision. ND also improves surface integrity (i.e., removing sub-surface damage, cracks or micro-cracks) and can achive surface finishes of less than 2 nm. Similar finishes can be achieved on silicon carbide wafers for IC memory devices.

Nanodiamond can also be used in electrolytic and electroless plating applications with very little modification to the process. It has been suggested that the effect of ultra-dispersed diamonds results in hardness, strength and fine structure with respect to current density and diamonds' concentration in the electrolyte. Potential applications of this coating include sliding bearings, insulators, medical appliances and aerospace parts.

Sapphire Polishing

Let's review a single application where ND can have an immediate impact-the final stage of sapphire polishing. It is beyond the scope of this article to address all the complexities of final polishing techniques, but it is necessary to note the variables in this process and see where ND can play a significant role. (It is important to note that the economic value of polishing is dependant not only on removal rate, but on a host of other process outputs such as finish, reject rate, production time required, and sub-surface damage.)

Variables in the sapphire polishing process include the speed of the machine (rotational) in revolutions per minute, the type of pad used on the platen (soft to hard), pressure (low to high), rotation (clockwise or counter-clockwise), back pressure on the workpiece, slurry composition, and drip rate. While some of these variables can be defined, the actual dynamic is in the interaction of the workpiece (in this case, sapphire), the ND-containing slurry, and the pad.

Physical properties of the fluids are important as they affect both fluid dynamics and material transport in polishing. These properties, including viscosity, density, pH, solids concentration and thermal stability, can also be varied by changes in the chemical composition of the slurry.

Particle size affects the number of particles in the slurry if weight percent concentrations are used. For small-diameter particles, a given slurry weight percent contains more particles than for large-diameter particles. The agglomeration of smaller particles can occur under certain conditions. If these agglomerations are due to van der Waals forces, they are considered "soft" agglomerates and may not be important.

If the agglomerations are a result of a chemical interaction of the particles in the makeup of the slurry, however, they become more important because this phenomenon affects average particle size and size distribution. In some cases, the agglomerates become hard and then fracture, thereby changing the average particle size and shape during the polishing process. This usually leads to defects and rejects.

The shapes of different particles may be significant. Some particles are characteristically spherical, but elliptical, blocky and sharp/fractured shapes are also used. Concentration effects may be negligible in some polishing operations or significant, as in CMP. Concentrations of solids may be given as weight percent or particles per volume.

As previously mentioned, in sapphire CMP, the type of slurry, particle and pad become the dominant variables. Many volumes of work have reported on the mechanisms of removal when the particles are harder than the workpiece. With a nanodiamond addition to colloidal silica in sapphire CMP, the mechanism is somewhat similar.

The applied load is born by particles embedded in the lap system in conventional polishing, while, in CMP, the applied load is born by the pad. Each abrasive particle removes material independently. At high concentrations, the removal rate tends to reach a peak limit as the pad becomes saturated. At low concentrations, each particle has an equal chance of embedding into the lap and the polishing rate increases with concentration.

As the particle embeds into the lap, material removal can be considered from the perspective of classical cutting mechanics. If the spacing of granules is sufficient, removal rates are controlled by the cross-sectional area of the material removed.

For soft laps (or for the soft polyurethane foam pads used on laps in CMP), sub-micron diameter abrasives can be pushed deep into the lap. Flexing and rebounding of the lap surface make direct contact with the workpiece and carry most of the applied load. The number of active particles is the product of the contact area and the number of particles per area of lap surface.

In addition to the variables discussed above, pad roughness and abrasive size distribution also have an effect. Increasing the width of the particle distribution brings two countervailing effects. A reduction in the number of particles in the mean of the distribution where a considerable amount of work is done can either reduce material removal rate for small mean particle sizes or increase the removal rate with a potential loss of surface quality with a larger proportion of large particles. ND is in the same size regime as colloidal silica and ceria slurries, with sizes ranging from 20 to 100 nanometers. Any potential surface damage is mitigated by ND's unique clustering.

As noted, the lap typically comprises a stiff plate that may then be faced with solid films. The transitions in material removal mechanisms associated with the workpiece/particle-lap interactions can be understood, in part, in the context of the load-bearing characteristics of the system. One extreme case occurs when the entire load is supported by the particles interposed between the lap and the work (e.g., diamond polishing of electroless nickel optics or silicon). At the other extreme, the entire load may be carried on the lap (pad) surface. Frequently, polishing processes operate at some intermediate state between these extremes.

The use of ND in planarization or fine polishing to the angstrom level eliminates many of the associated problems of sub-surface damage and heat buildup that can be encountered when using conventional single-crystal poly-particles processed to nanometer size ranges (see Figure 3).

Case Study

A trial was conducted at a watch glass manufacturer in Switzerland using 100 nm ND particles in colloidal silica for the final CMP of R-plane sapphire watch glass. The purpose of the study was to identify either increased production rates or quality improvements using ND in a percent replacement for colloidal ceria or silica.

Kostosol 3550 silica was used, along with a 5w% blend of nanodiamond in a chemically compatible fluid. The polishing pad was manufactured by Rodel and had a Shore A rating of 90, and the machine manufacturer was COMES.

For the initial test, the cycle time was 4.5 minutes at 3.5 bar. The R-plane sapphire pieces were 35 mm in diameter, and four pieces were used per run. (Results are shown in Table 1.) The test was then duplicated on 42 mm diameter wafers with two pieces per run (see Table 2), and again using the same size of sapphire wafers with four pieces per run (see Table 3). The next test was a re-polishing to remove scratches (rejects) from pre-CMP polish defects. For this test, the sapphire pieces were 36 mm diameter with four pieces per polish (see Table 4).

In all cases, the use of ND showed a beneficial effect on both quality and the potential to reduce manufacturing time. In addition, the ND suspension was stable for one month. These results have been duplicated at another sapphire watch maker as well.

Collodial silica and ceria are soft materials, and part of the polishing action using these agents occurs through a surface chemical process. The tribo-chemical wear process in CMP and CMP-like operations, which is difficult to control, takes place in the action interstice between the pad and workpiece. When polishing operations are conducted on the functional surfaces of brittle media, the surface of the medium to be polished, the polishing agent and the carrier of the polishing agent are all factors.

The suspending fluid also exerts considerable influence on the level of wear sustained by the pad/piece. To date, it has not been possible to develop a hypothesis that combines all of these aspects, though it is known that the polishing process is determined by the reactions that take place at the contact surface between the workpiece and the polishing medium in aqueous solutions.

The addition of ND, even in small quantities, provides particles with a hard core surrounded by a functionalized cluster that, in effect, provides a "burnishing" to the workpiece. The end result is that sapphire or other substrates are pseudo-mechanically polished to a high degree. In addition, the chemical process is muted and time reductions are possible using a tailored ND.

Viable Alternative

ND should be considered in the final polish of any substrate where the tribology of the substrate is critical. In addition to the reduction of subsurface damage and increased surface finish, it is also possible to increase productivity and reduce costs through the introduction of ND.

For more information regarding ND, contact NanoDiamond Products Limited, Unit 4 Beechpark House, Smithstown Industrial Estate, Shannon, County Clare, Ireland; call (353) 61-360333; e-mail ron@nanodiaproducts.com; or visit www.nanodiaproducts.com.

Links

• NanoDiamond Products Ltd.