Plasma etching of diamond surfaces

Abstract

A polycrystalline CVD diamond material comprising a surface having a surface roughness R.sub.q of less than 5 nm, wherein said surface is damage free to the extent that if an anisotropic thermal revealing etch is applied thereto, a number density of defects revealed by the anisotropic thermal revealing etch is less than 100 per mm.sup.2.

Claims

1. A polycrystalline CVD diamond material comprising a surface having a surface roughness R.sub.q of less than 5 nm, wherein said polycrystalline CVD diamond material consists of a layer of polycrystalline diamond material that excludes any layers of single crystal diamond material and wherein said surface is damage free to the extent that the following criteria is fulfilled: if an anisotropic thermal revealing etch is applied thereto, a number density of defects revealed by the anisotropic thermal revealing etch is less than 100 per mm.sup.2, wherein the anisotropic thermal revealing etch is performed using the following procedure: (i) examining the surface at a magnification of 50 times using reflected light with a metallurgical microscope to ensure that there are no surface features present; (ii) exposing the surface to an air-butane flame thereby raising the surface to a temperature in a range 800 C. to 1000 C. for a period of 10 seconds; (iii) examining the surface at a magnification of 50 times using reflected light with a metallurgical microscope and counting defects revealed by the anisotropic thermal revealing etch to determine their number density; and (iv) repeating steps (ii) and (iii) and comparing the measured density of defects with that of the previous cycle until the following condition is met: if the number density of defects counted is less than or equal to 150% of the number density determined in the previous cycle, then all the defects are deemed to be revealed and the measurement recorded is the average of the measurements of the last two cycles, if not the cycle is repeated again, and wherein the number density of defects in step (iii) is measured by the following method: (i) the defects to be counted are those defects visible at a magnification of 50 times with a metallurgical microscope which fall totally or partially within a rectangular area 1 mm0.2 mm projected onto the surface being characterised; (ii) the area is selected at random over the surface or portion of the surface to be characterised and randomly oriented; (iii) the defects are counted in a minimum of 5 such areas; and (iv) the number density of defects is calculated by dividing the total number of defects counted by the total area examined to give a number density in terms of defects per mm.sup.2.

2. A polycrystalline CVD diamond material according to claim 1, wherein the surface roughness is less than 1 nm.

3. A polycrystalline CVD diamond material according to claim 1, wherein said surface is damage free to the extent that if the anisotropic thermal revealing etch is applied thereto, the number density of defects revealed by the anisotropic thermal revealing etch is less than 50 per mm.sup.2.

4. A polycrystalline CVD diamond material according to claim 1, wherein said surface is damage free to the extent that if a backscattering ion beam analysis is applied thereto, a backscattered ion yield is less than 5% of incident ions.

5. A polycrystalline CVD diamond material according to claim 1, wherein a maximum depth of an etched structural feature in a direction perpendicular to said surface is less than 20 m; and a smallest lateral dimension of the etched structural feature is at least 0.5 times the maximum depth of the etched structural feature.

6. A polycrystalline CVD diamond material according to claim 1, wherein said surface is a growth surface of the polycrystalline CVD diamond material.

7. A polycrystalline CVD diamond material according to claim 1, wherein said surface comprises topographical features.

8. A polycrystalline CVD diamond material according to claim 7, wherein said topographical features comprise structural features having a depth in a range 2 nm to 100 nm.

9. A polycrystalline CVD diamond material according to claim 7, wherein said topographical features form an optical surface structure.

10. A polycrystalline CVD diamond material according to claim 9, wherein said topographical features comprise structural features having a depth in a range 200 nm to 20 m.

11. A polycrystalline CVD diamond material according to claim 9, wherein said optical surface structure comprises a smooth curved surface.

12. A method of fabricating the polycrystalline CVD diamond material according to claim 1, the method comprising: mechanically processing a surface of a polycrystalline CVD diamond material, wherein said polycrystalline CVD diamond material consists of a single layer of polycrystalline diamond material that excludes any layers of single crystal diamond material; and etching said surface to remove surface damage introduced during the mechanical processing, wherein said etching comprises inductively coupled plasma etching (ICP) using a gas mixture containing argon and chlorine, and wherein at least 0.5 m of polycrystalline CVD diamond material is removed from said surface during etching without unduly increasing a surface roughness R.sub.q of the surface, whereby after etching the surface has a surface roughness R.sub.q of less than 5 nm and is damage free to the extent that if an anisotropic thermal revealing etch is applied thereto, a number density of defects revealed by the anisotropic thermal revealing etch is less than 100 per mm.sup.2 as defined in claim 1.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) The following figures are referred to in examples 1 to 6.

(2) FIG. 1 is a scanning electron micrograph of an Ar/Cl ICP plasma etched trench approximately 16 nm deep.

(3) FIG. 2 is an atomic force microscopy image of a region of the trench in FIG. 1.

(4) FIG. 3 is a scanning electron micrograph of an Ar/CI ICP plasma etched trench approximately 53 nm deep.

(5) FIG. 4 is an atomic force microscopy image of a region of the trench in FIG. 3.

(6) FIG. 5 is line drawing based on an atomic force microscopy image of an MCI ICP plasma etched diamond micro-lens

(7) FIG. 6 is a measured line profile across the centre of the diamond micro-lens shown in FIG. 5, along with a fitted profile with ideal spherical curvature.

(8) FIG. 7 is a plot of etch rate and etch selectivity against ICP coil power.

(9) FIG. 8 is a plot of etch rate versus etch time.

(10) FIG. 9 is an atomic force microscopy image of a HPHT diamond surface polished using a resin-bonded wheel.

(11) FIG. 10 is an atomic force microscopy image of the HPHT diamond sample in FIG. 9 following Ar/CI ICP plasma etching.

DETAILED DESCRIPTION OF THE EMBODIMENTS

(12) The details of the invention will be described using six examples. The first example describes the use of the invention for making an essentially damage free surface on a series of single crystal CVD diamond plates and comparing the resultant surfaces with a sample that has not been etched. The second example describes the use of the invention for making fine structural features on a polycrystalline CVD diamond plate.

Example 1

(13) Four single crystal CVD diamond plates approximately 4 mm4 mm0.5 mm were laser sawn from larger as-grown blocks of single crystal CVD diamond made by the method disclosed in WO 01/96633. The large planar surfaces of the plates (hereinafter referred to as the major surfaces) were oriented within a few degrees of crystallographic surfaces. The smaller surfaces forming the edges of the plates (hereinafter referred to as the edge surfaces) were approximately parallel to <100> directions.

(14) The major surfaces of all four plates were initially lapped on a cast iron lapping wheel using progressively finer diamond grit suspended in a cellulose-based carrier. The final lapping stage used diamond grit in the size range 15 m to 25 m and the roughness R.sub.q measured using a stylus profilometer over a 0.8 mm length was in the range 100 to 150 nm for all three plates. Previous experiments have indicated that after this stage of lapping, the sub-surface damaged layer extends to a depth of approximately 10 m beneath the surface.

(15) All of the plates were chosen at random and their major surfaces were polished using a traditional diamond polishing scaife. Two cast iron scaife-wheels were used, the first prepared with diamond powder in the size range 2-4 m and the second with diamond powder in the size range 1-2 m. The depth of material removed during the scaife polishing process was determined to be approximately 15 m. After completion of the scaife processing, the R.sub.q values measured using a stylus profilometer over an 80 m length were less than 5 nm. More accurate measurements over areas (e.g. about 1 m by about 1 m) made with an atomic force microscope gave the R.sub.q listed in Table 1.

(16) One of the plates (referred to as Sample 1) was selected at random and subjected to the thermal revealing etch according to the method previously described (heating the surface in air to about 850 C. for 10 seconds using a small air-butane blow torch). The defects were counted according to the method of the invention. The plate showed considerable evidence of sub-surface damage, very much greater than 100 defects per mm.sup.2.

(17) The other three plates (referred to as Sample 2, Sample 3 and Sample 4) were subjected to inductively coupled plasma etches to remove material from the surface. The ICP etching conditions were: a platen power of approximately 300 W, a coil power of approximately 400 W, a chamber pressure of 5 mTorr (0.667 Pa), flow rates of 25 sccm for Ar and 40 sccm for Cl.sub.2. The etching times chosen for the three samples were 15, 30 and 60 minutes. The amount of material removed from the etched surfaces, the initial and final surface roughness, R.sub.q, values are in Table 1.

(18) TABLE-US-00001 TABLE 1 Characteristics of plates before and after plasma etching R.sub.q after Plasma R.sub.q after Approximate scaife etching plasma depth of polishing, time, etching, material removed, Sample nm minutes nm nm 1 0.9 N/A N/A N/A 2 1.2 15 1.1 900 3 0.9 30 0.7 2000 4 0.7 60 0.5 4000

(19) After the plasma etching process and after the measurement of the R.sub.q, all three plates were subjected to the thermal revealing etch to evaluate the extent of sub-surface damage. For each of samples 2 to 4 the number density of defects measured was in each case less than 10, unlike that revealed in sample 1, which was the plate not subjected to the plasma etch.

Example 2

(20) Two samples of polycrystalline CVD diamond about 10 mm10 mm in lateral extent and about 650 m thick were polished on their growth surfaces using conventional lapidary processes. The surface roughness values, R.sub.q, of the polished plates, measured by atomic force microscopy over a 2 m by 2 m area were 1.0 nm and 1.0 nm.

(21) Micro-trenches were etched into the two samples of polycrystalline diamond using an ArCl.sub.2 ICP plasma. Micro-trench photoresist features were mask patterned on the surface of the samples using standard lithographic methods. The samples were etched using the following ICP chamber conditions: a platen power of approximately 300 W, a coil power of approximately 400 W, a chamber pressure of 5 mTorr (0.667 Pa), flow rates of 25 sccm for Ar and 40 sccm for Cl.sub.2. In the first sample trenches of depth 16 nm were etched [FIG. 1 and FIG. 2] and in the second sample trenches of depth 53 nm were etched [FIG. 3 and FIG. 4]. It can be seen that the trench walls are well-defined, with no rough edges apparent. The etch rates were measured to be approximately 1 nm per second in both cases, which demonstrates the reproducibility of this method. It is also clear that the surfaces post-etch show no indication of the grain boundaries in these polycrystalline samples, which demonstrates that the etch is isotropic in nature. The R.sub.q, values (measured over 1 m1 m areas using an atomic force microscope) of the samples after etching the 16 and 53 nm deep trenches are 0.8 and 1.1 nm respectively, which should be compared to the pre-etched surface roughness of 1.0 nm. This indicates that under these conditions little or no surface roughening is caused using this etching method, and further supports the conclusion that the etch is isotropic.

Example 3

(22) Spherical micro-lens structures were fabricated in a sample of single crystal IIa natural diamond using Ar/Cl.sub.2 ICP etching in conjunction with the photoresist reflow method [H. W. Choi, E. Gu, C. Liu, J. M. Girkin, M. D. Dawson, J. of Appl. Phys. 97 (6), 063101 (2005); C. L. Lee, H. W. Choi, E. Gu, M. D. Dawson, Dia. Rel. Mat. 15, 725 (2006)]. A 7 m thick layer of Shipley SPR220 photoresist was spin-coated onto the diamond substrate first. After mask patterning, the photoresist pillars were formed. The sample was then placed on a hotplate at 125 C. for 2 minutes to allow the photoresist pillars to re-flow. Due to the surface tension, spherical lens structures were formed. These lens structures were then transferred onto the diamond substrate by ICP etching using Ar/Cl.sub.2 plasma. The ICP parameters used were: a platen power of 100 W, a coil power of 400 W, a chamber pressure of 5 mTorr (0.667 Pa), flow rates of 25 sccm for Ar and 40 sccm for Cl.sub.2, with an etching duration of 25 minutes.

(23) A line drawing based on an atomic force microscopy image of a representative diamond micro-lens is shown in FIG. 5. The surface diameter and height of the lens were measured to be 50 m and 0.75 m respectively. The profile of the diamond micro-lenses was studied by examining the cross-sectional scan of the atomic force microscopy images. The measured profile was then compared to the curvature of a circle. The measured cross-sectional profile of the diamond micro-lens and the fitted spherical profile are plotted in FIG. 6. It can be seen that the deviation from the ideal profile is very small, indicating the cross-sectional profile of the micro-lens is very close to spherical in shape.

Example 4

(24) The etch rate and etch selectivity (as defined earlier) of the Ar/Cl.sub.2 ICP plasma was investigated as a function of the ICP coil power using the following conditions: ICP coil powers of 100 W. 400 W and 900 W; a constant ICP platen power of 300 W; a chamber pressure of 5 mTorr (0.667 Pa); flow rates of 25 sccm for Ar and 40 sccm for Cl.sub.2; an etching duration of 5 minutes. Single crystal natural diamond samples were used for these experiments and photoresist was used as the mask material. The results are plotted in FIG. 7. It is observed that the etch rate increases linearly with the ICP coil power and etch selectivity is approximately constant (at around 0.09). In a similar experiment the etch selectivity on type Ib synthetic single crystal diamond was also found to be around 0.09. This should be compared to the etch selectivity using an Ar/O.sub.2 ICP plasma, which is around 0.20 [C. L. Lee, H. W. Choi, E. Gu, M. D. Dawson, Dia. Rel. Mat. 15, 725 (2006)]. Hence, using photoresist masks of identical thickness, the height or depth of microstructures fabricated in diamond (such as micro-lenses and micro-trenches) using ICP Ar/Cl.sub.2 etching would be approximately half of those fabricated using ICP Ar/O.sub.2 etching. Thus the lower etch selectivity obtained using Ar/Cl.sub.2 is more suitable for etching shallower structural features into diamond in which precise control on the depth or height of the microstructure is required.

Example 5

(25) Etch rate as a function of time was investigated by etching micro-trenches in type Ib synthetic single crystal diamond samples. Trenches were etched in three samples using identical photoresist masks and Under/\r/Cl.sub.2 ICP plasma conditions identical to those described in Example 2. The three samples were etched respectively for 9, 20 and 40 seconds. The depths of the micro-trenches formed were measured by atomic force microscopy and are plotted in FIG. 8 as a function of the etch time. The mean etch rate was extracted by performing a simple linear curve fit to the data, and found to be 1.290.03 nm s.sup.1. The small uncertainty in this number indicates that the method yields a reproducible etch rate, which is an important requirement of a manufacturing process.

Example 6

(26) A type Ib HPHT single crystal sample, mechanically polished using a resin-bonded polishing wheel, was etched for 10 minutes using the following ICP chamber conditions: a platen power of approximately 300 W, a coil power of approximately 100 W, a chamber pressure of 5 mTorr (0.667 Pa), flow rates of 25 sccm for Ar and 40 sccm for Cl.sub.2. FIG. 9 shows a representative atomic force microscope scan over a 1 m1 m surface area of the polished sample prior to ICP etching. The surface is characterised by nanometer-scale linear trenches or grooves resulting from the polishing and the roughness R.sub.q over this area was measured to be 0.53 nm. FIG. 10 shows a representative AFM scan over a 1 m1 m surface area of the sample after the ICP etch. It can be seen that the linear polishing grooves are shallower than before and the roughness R.sub.n over this area was measured to be 0.19 nm. This example demonstrates the ability of this method to reduce the roughness of mechanically pre-polished diamond surfaces.