EP1628908A1

EP1628908A1 - Formation of silicon nanostructures

Info

Publication number: EP1628908A1
Application number: EP04707340A
Authority: EP
Inventors: Andreas Markwitz
Original assignee: Institute Of Geological And Nuclear Sciences Ltd
Current assignee: Institute Of Geological And Nuclear Sciences Ltd
Priority date: 2003-01-31
Filing date: 2004-02-02
Publication date: 2006-03-01
Also published as: AU2004207125A1; WO2004067445A1; EP1628908A4

Abstract

The present invention comprises a method of forming nanostructures on a silicon substrate including the steps of in a chamber heating the substrate with an electron beam to a peak temperature, holding the peak temperature for a predetermined time, and decreasing the temperature of the substrate. Neon and carbon ions may be implanted into the substrate before the step of heating the substrate to produce different nanostructures. Implanting with nitrogen before heating prevents the formation of nanostructures. Plasma immersion ion implantation with nitrogen ions before heating also forms nanostructures.

Description

FORMATION OF SILICON NANOSTRUCUTRES

FIELD OF INVENTION

The invention comprises a new manufacturing procedure to fabricate silicon nanostructures using electron beam annealing.

BACKGROUND

The fabrication and properties of silicon nanostructures is the focus of much research due to their unique properties and potential applications. Silicon nanostructures may have applications in optoelectronics, in single electron devices and as cathodes for vacuum microelectronics. One key challenge facing the practical realisation of devices based on silicon nanostructure technology is the identification of suitable nanofabrication techniques. Conventional silicon nanostructure fabrication technology implements electron beam lithography for the high-resolution definition of nanostructure etch masks (for example as described in W. Chen and H. Ahmed, Appl. Phys. Lett., Vol 63, pp 1116, (1993)). Such a technique provides high uniformity, precise dimensional control and compatibility with conventional processing techniques. However the restrictively low throughput associated with electron-beam lithography is a serious constraint to its development as a practical industrial nanofabrication tool. Alternative nanofabrication tools which emphasise increased throughput have thus been investigated.

Research into alternative nanofabrication tools has principally focussed on the formation of nanostructure etch masks using faster and simpler techniques than electron-beam lithography. Self-assembled monolayers of colloidal particles, such as polystyrene beads have been used to fabricate such etch masks (for example as described in H. W. Deckman and J. H. Dunsmuir, Appl. Phys. Lett., Vol 41, pp 377, (1982), and C. Haginoya, M. Ishibashi, and K. Koike, Appl. Phys. Lett., Vol 72, pp 2934, (1997)). Although this natural lithography technique is compatible with high throughput, acceptable nanostructure uniformity over the sample surface has not been

127643 achieved. A second natural lithography approach, first used on GaAs substrates (see M. Green, M. Garcia-Parajo, F. Khaleque, and R. Murray, Appl. Phys. Lett., vol 62, 264 (1993) and B. W. Alphenaar, Z. A. K. Durrani, A. P. Heberle, and M. Wagner, Appl. Phys. Lett., vol 66, 1234, (1995)) and more recently adapted for silicon nanopillar formation (see K. Seeger, and R. E. Palmer, Appl. Phys. Lett., vol 74, 1627 (1999)) uses the deposition of a thin metal film which forms a granular layer consisting of isolated islands. High throughput is achieved with reasonable uniformity. The umformity may be improved using the deposition of size-selected Au clusters followed by electron cyclotron resonance plasma etching (ECR etching) with SF₆ at — 130 °C (as described in T. Tada, T. Kanayama, K. Koga, P. Weibel, S. J. Carroll, K. Seeger and R. E. Palmer, J. Phys. D., vol 31, L21 (1998)). High throughput and high uniformity were demonstrated however the cluster formation and etching processes rely on relatively uncommon fabrication equipment. Alternative silicon nanostructure fabrication techniques not founded on the definition of etch masks have been the focus of only limited research. Perhaps most significant is the use of scanning tunnelling microscopes for atom manipulation (see R. M. Ostrom, D. M. Taneribaum, and A. Gallagher, Appl. Phys. Lett., vol 61, 925, (1992)). Nanopillars of 3nm diameter have been realised with excellent spatial control, however this technique is not practical for high density fabrication. There is a need for a method of fabricating silicon nanowhiskers that has high throughput and produces nanowhiskers with acceptable uniformity.

SUMMARY OF INVENTION

It is the object of the invention to provide an improved method for forming silicon nanostructures or to at least provide the public with a useful choice.

In broad terms in one aspect the invention comprises a method of forming nanostructures on a silicon substrate including the steps of in a chamber heating the substrate with an electron beam to a peak temperature, holding the peak temperature for a predetermined time, and decreasing the temperature of the substrate.

127643 In broad terms in another aspect the invention comprises a method of preventing the formation of nanostructures on a silicon substrate including the steps of in a chamber implanting nitrogen ions with an implantation energy of between 50 and 150 keV into the silicon substrate, heating the substrate with an electron beam to a peak temperature, holding the peak temperature for a predetermined time, and decreasing the temperature of the substrate.

In broad terms in another aspect the invention comprises a nanostructure on a silicon substrate formed using the method of the invention.

In broad terms in another aspect the invention comprises a silicon substrate including a nanostructure formed using the method of the invention.

BRIEF DESCRIPTION OF DRAWINGS

The invention will be further described by way of example only and without intending to be limiting with reference to the following drawings, wherein:

Figure 1 is a flow chart of the steps of the annealing process for forming silicon nanowhiskers; Figure 2A is an atomic force microscope image of nanowhiskers formed on a

(100) silicon substrate;

Figure 2B is a chart showing the silicon substrate of Figure 2A and analysis of the silicon substrate;

Figure 3 A is an atomic force microscope image of nanowhiskers formed on a (111) silicon substrate;

Figure 3B is a chart showing the silicon substrate of Figure 3 A and analysis of the silicon substrate;

Figure 4A is an atomic force microscope image of nanowhiskers formed on a (100) silicon substrate; Figure 4B is an atomic force microscope image of nanowhiskers formed on a

(100) silicon substrate showing a cluster of nanowhiskers;

127643 Figure 5A is an atomic force microscope image of nanowhiskers formed on a (100) silicon substrate;

Figure 5B is an atomic force microscope image of nanowhiskers formed on a (100) silicon substrate; Figure 6 is a graph showing nanostructure density and nanostructure height as a function of annealing time;

Figure 7 is a graph showing nitrogen profiles for 50 keV and 100 keV ¹⁴N ions implanted into silicon;

Figure 8 is an atomic force microscope image of silicon after nitrogen ion implantation and annealing;

Figure 9 is a nuclear reaction analysis (NRA) microprobe line scan from the boundary between unimplanted and implanted areas.

Figure 10 is a flow chart showing of the steps used for forming nanoboulders;

Figure 11A is an atomic force microscope image of a silicon substrate surface implanted with carbon ions;

Figure 11B is a graph showing the variation in height on the surface of the silicon substrate of Figure 11 A;

Figure 12A is an atomic force microscope image of a silicon substrate surface implanted with carbon ions and annealed; Figure 12B is a second atomic force microscope image of the silicon substrate surface of Figure 12A implanted with carbon ions and annealed;

Figure 12C is a graph of the variation in height of the nanoboulders formed on the surface of the silicon substrate of Figure 12 A;

Figure 12D is a section of Figure 12 A showing the line on which the graph of Figure 12C has been taken;

Figure 13 shows an atomic force microscope image and scan analysis of nanoboulders formed on silicon (100);

Figure 14 shows an atomic force microscope image and scan analysis of nanoboulders formed on silicon (100); Figure 15 is a graph showing the relationship between implantation fluence and volume of nanoboulders and size of nanoboulders;

127643 Figure 16 is a graph showing the relationship between nanoboulder diameter and spacing between nanoboulders;

Figure 17 shows neon ion implantation depth profiles in a silicon substrate;

Figure 18 is an atomic force microscope image showing the surface of a silicon substrate after implantation with neon ions;

Figure 19 is an atomic force microscope image of the surface of a silicon substrate after implantation with neon ions and annealing;

Figure 20 is an atomic force microscope image of the surface of a silicon substrate after plasma immersion ion implantation with nitrogen and annealing; Figure 21 is an atomic force microscope image of the surface of a silicon substrate after plasma immersion ion implantation with nitrogen and annealing; and

Figure 22 is an atomic force microscope image of the surface of a silicon substrate after plasma immersion ion implantation with nitrogen and annealing.

DETAILED DESCRIPTION OF A PREFERRED FORM

Silicon nanostructures can have differing shapes include whiskers, boulder and cavities. Various embodiments of this invention cover each of these three nanostructures. The formation of nanostructures of the invention all include an annealing step with different nanostructures formed by implantation of particular ions before annealing. Figure 1 shows the steps used in the annealing process.

Silicon nanowhiskers are produced in a three step process as shown in Figure 1. In step 1 substrates are provided for processing. The substrates may be p-type or n-type and may be (100), (110), or (111) silicon substrates. If the substrate is p-type doping may be Boron doping. Alternatively any suitable dopant may be used.

In step 2 the silicon substrate on which the nanowhiskers are to be formed is cut to the desired target size and polished on that surface that will be annealed. For example a typical target size may be 1 ^χl cm. The silicon substrate is typically 0.5 mm thick and has resistivity of 1 - 10 Ωcm. However other thicknesses of silicon substrate may be used. In step 3 the targets are mechanically cleaned by spraying pressured air onto the

127643 surface of the silicon substrate that will be annealed. This process removes any dust or cutting products from the silicon but does not disturb the native oxide layer on the surface of the silicon substrate. This native oxide layer is generally 15 - 20 A thick and is formed by exposure to air and residual contaminants, typically carbon based compounds.

In step 4 the substrate is loaded into the annealing facility. The mechanically cleaned side of the substrate faces towards to electron beam direction. Once the residual gas pressure of less than 10^"5 mbar (10^'3 Pa) is reached, the annealing process may start. Ideally the residual gas pressure is as low as possible.

In step 5 the silicon substrate is heated by scanning the silicon surface with an electron beam. Typically the electron beam is raster scanned along the surface of the substrate. This process can be broken into three steps, heating, holding the peak temperature, and cooling. The process typically begins at room temperature. In the first step the substrate surface is gently heated with a temperature gradient of 5 - 10 °C/s to a peak temperature. In the second step the peak temperature, between 800 and 1200 °C, is held for between 3 and 120 seconds. In the third step the substrate surface is gently cooled with a temperature gradient of 5 - 10 °C/s, to room temperature. The electron beam is typically raster scanned over the substrate with a frequency between 1 and 10 kHz. The frequency of raster scanning is set to allow a homogeneous temperature across the substrate.

Typically the peak temperature is held for between 3 and 120 seconds. However longer peak temperature holding times (for example up to 1000 seconds) could be used. Processed targets can stay in the vacuum chamber for long time, such as a few hours to allow processing on many targets under identical conditions.

Figures 2 to 5 show examples of nanostructures formed using the annealing method described with reference to Figure 1.

127643 Figure 2A is a 4 μm x 4 μm top view atomic force microscope image of silicon nanowhiskers produced by the annealing process of the invention using rapid thermal electron beam annealing (EB-RTA) under high vacuum conditions as described with reference to Figure 1. The temperature gradient for both increasing and decreasing the temperature to produce the nanowhiskers of Figure 2A was 5°C/s. The substrate of Figure 2 A is (100) p-type silicon. As can be seen from this image a plurality of nanowhiskers has been formed.

Figure 2B shows some analysis of the substrate of Figure 2A including a graph of a line across Figure 2A showing the height of some of the nanowhiskers. As can be seen from this graph the nanowhiskers in this example are up to 15 nm tall.

Figure 3 A is a 12 μm x 12 μm top view atomic force microscope image of silicon nanowhiskers produced by the annealing process of the invention using rapid thermal electron beam annealing under high vacuum conditions. The temperature gradient for both increasing and decreasing the temperature to produce the nanowhiskers of Figure 2A was 5°C/s. The substrate of Figure 3A is (111) p-type silicon. As can be seen from this image a plurality of nanowhiskers has been formed.

Figure 3B shows some analysis of the substrate of Figure 3 A including a graph of a line across Figure 3A showing the height of some of the nano-whiskers. As can be seen from this graph the nanowhiskers in this example are up to 50 nm tall.

As Figures 2 and 3 show, silicon nanowhiskers are produced by the manufacturing process using single sided p-type (100) and (111) silicon substrate materials (thickness 0.5 mm).

Atomic force microscope images on the virgin silicon surface show no significant surface structures as expected.

Figure 4 A is an 8μm x 8μm plan view atomic force microscope image of nanostructure growth following scanning of p-type silicon (100) at 1100°C for 15 seconds. The p-

127643 type silicon is B-doped, 1 - 10 Ωcm and was diced into a 10 x 10 mm² sample. This substrate was not cleaned and thus contained native oxide (15 - 20 A thick) formed by exposure to air and residual contaminants, typically carbon-based compounds. The annealing system used in the fabrication of the nanostructures of Figure 4A operates with electron energy of 20 keV and sample current up to 2mA. The electron beam is focussed on the sample surface to a spot of approximately 1 mm diameter which can be electromagnetically scanned over a 1 cm² sample area with X-Y sweep frequencies of 1 and 10 kHz. The beam current and hence sample temperature is controlled via a thermopile detector. True temperature measurements can be performed using a pair of 2-color pyrometers which face the front and back sample surfaces. Residual gas analysis facilities were provided that allow in-situ monitoring of gaseous species evolved during annealing. The annealing chamber was evacuated with a turbomolecular pump to a base pressure <1 x 10^"6 mbar.

Surface structure analysis of the nanowhiskers imaged in Figure 4A was performed using a Digital Instruments NanoScope Ilia AFM using a silicon probe (tip radius < 10 nm, half cone angle 10 ° at apex). Vertical and horizontal resolutions for this instrument are 1 nm and 5 nm respectively. Additional large area topographic information was provided by SEM.

It should be noted that the apex of each feature is expected to be sharper than the AFM tip apex and the AFM image will thus not be a true representation of the feature, rather a convolution of AFM tip and nanostructure. Such convolution effects, although preventing accurate imaging of the nanopillar apex, will not lead to misrepresentation of the nanowhisker density or height.

Figure 4B is a 3 μm 3 μm top view atomic force microscope image of the central region of Figure 4A. The region marked C in Figure 4B is an example of a nanostructure cluster region. Both images show evidence of pillar clustering, as indicated in region C of Figure 4B. Identical clustering was observed in SEM images and is thus not a facet of the AFM imaging. The size and number of clusters was independent of the annealing temperature within the range 800°C to 1100°C.

127643 Figure 5 A is a 4 μm x 4 μm top view atomic force microscope image of a nanowhiskers following annealing of p-type silicon (100) at 1100°C for 3 seconds. Figure 5B is a 4 μm x 4 μm top view atomic force microscope image of nanowhiskers following annealing of p-type silicon (100) at 1100° for 120 seconds. Surface structures of similar shape and crystallo graphic orientation were observed following annealing at 1100 °C for durations as little as 3 s as shown in the AFM topograph of Figure 5A. However, a large number (density) of nanostructures (up to 27.6 /μm ) was observed. The maximum pillar height was 11 nm with an average nanopillar height of only 6 nm. The formation of surface nanowhiskers persisted following annealing for durations up to 120 s as shown in Figure 5B. Here, the average pillar height observed was 12 nm, with a maximum height of 21 nm.

Cluster formation is evident in Figures 4A, 4B, 5A, 5B and the size and number of clusters was independent of annealing duration. Conversely, a clear linear dependence exists between the average pillar height and anneal time, as shown in Figure 6. Figure 6 also shows the relationship between the average nanowhisker density and the anneal duration which shows that density decays exponentially reaching a constant value of 17.5 /μm² after approximately 60 s. So as the peak temperature is held the nanowhisker height increases linearly and the nanowhisker density decreases exponentially to a constant value.

As can be seen from Figures 2 to 6 a range of temperature gradients, peak temperatures, peak temperature holding times and substrates can be used to produce silicon nanostructures using the method described with reference to Figure 1.

Implanting nitrogen ions into a silicon substrate before annealing (using the method described with reference to Figure 1) leads to areas free of nanostructures where the implantation takes place. This can be seen in Figure 8 where an area of the silicon substrate was implanted with nitrogen ions before annealing. To demonstrate the effect of nitrogen ion implantation substrates were p-type silicon (100) (B-doped, 1-10 Ωcm) diced into 10x10 mm² samples. No surface cleaning or other treatments were performed on the substrates and the surface thus contains the native oxide (1.5-2.0 nm

127643 thick) formed by exposure to air and residual contaminants, typically carbon-based compounds. The samples were implanted with 50 keV (fluence of 1.15xl0¹⁷ cm^"2) and 100 keV (fluence of 2.77x10¹⁷ cm^"2) 'V ions under high vacuum conditions and at room temperature. Annealing was performed using an EB-RTA system at peak temperature of 1100 °C. This system operates with electron energy of 20 keV and sample current up to 2 mA. The annealing chamber was evacuated with a turbomolecular pump to a base pressure <lxl0^"6 mbar. In preferred embodiments the implantation fluences range between 10¹⁷ and 5 x 10¹⁷ cm^"2.

Surface structure analysis was performed using a Digital Instruments NanoScope Ilia AFM using a silicon probe (tip radius < 10 nm, half cone angle 10° at apex). Vertical and horizontal resolutions for this instrument are 1 nm and 5 nm respectively. Nuclear Reaction Analysis (NRA) measurements were carried using 0.920 MeV deuteron beam focused down to 20 μm and scanned over 5 x 5 mm . The beam current at the target was 2-3 nA. The nuclear reactions ¹⁶O(d,pι)¹⁷O (E_p = 1.39 MeV, dσ/dΩ = 4.6 mb sr^"1), ¹⁴N(d,α ¹²C (E_α = 5.63 MeV, dσ/dΩ = 0.9 mb sr^"1) and ¹²C(d,p₀)¹³C (E_p = 2.7 MeV, dσ/dΩ = 55 mb sr^"1) were used to determine the concentrations of C, N and O. Standard samples of Ta₂O₅ and TiN were used for calibration purposes.

Figure 7 shows the results of simulations of the ¹⁴N implantation profile as determined using DYNAMIC-TRIM for 50 keV and 100 keV implantations respectively. For both energies the ¹⁴N profile lies beneath the surface and corresponds to an under stoichiometric film. The mean projected range for 50 keV is 136 nm, surface concentration: 1.5-2%, peak concentration: 16.2 % and sputtering yield = 0.30. The mean projected range for 100 keV is 256 nm, surface concentration: 1 %, peak concentration: 25.6 % and sputtering yield = 0.21.

Figure 8 shows an atomic force microscope image of the silicon (100) surface following EB-RTA for a region ¹⁴N implanted at 50 keV. Examples of nanostructures formed without ¹⁴N implantation are shown in earlier figures. These characteristic images were taken from a single 10 x 10 mm silicon (100) substrate in which ¹⁴N implantation was restricted to a central 5 mm diameter spot. The implanted regions, for both 50 keV and

127643 100 keV implantations, show no evidence of surface nanostructuring with the measured surface roughness of the implanted surfaces being comparable to that of the virgin silicon (100). Nanostructures with an average height of 10 nm are seen throughout the non-implanted regions. These structures are identical to those formed on untreated silicon (100) surfaces following EB-RTA. It should be noted that these AFM image do not provide a true representation of the features, rather a convolution of AFM tip and nanostructure. Although preventing accurate imaging of the nanostructure apex, this convoluted image will not lead to misrepresentation of the pillar density or height.

To assess the role of ¹⁴N implantation into silicon (100) in the suppression of spontaneous nanostructuring NRA analysis was performed on the untreated, implanted and annealed silicon (100). The concentrations of carbon and oxygen in un-treated, implanted and annealed silicon (100) were also measured. Figure 9 shows the distribution of ¹⁴N from the centre of the 50 keV implanted region to the sample edge following EB-RTA. Identical behaviour was observed following 100 keV implantation. The ¹⁴N concentration at the centre of the sample is 1.15 x 10¹⁷ cm^"2, dropping abruptly to 1 10 cm^" , approximately 3.5 mm from the sample centre, corresponding to the boundary between the implanted and un-implanted region. The relatively broad ion probe used in these experiments does not allow us to provide quantitative information regarding the sharpness of the implantation profile. The concentration of ¹⁴N recorded in the un-implanted region following EB-RTA is identical to the concentration of ¹⁴N found in virgin silicon (100). Additionally, no implanted ¹⁴N is lost during annealing since the ¹⁴N concentration measured in the implanted region after EB-RTA is identical to that recorded prior to annealing. Low and comparable levels of oxygen (12 x 10¹⁵ cm^" ) and carbon (4 x 10 cm^" ) contamination were measured in the un-treated and un- implanted, annealed silicon (100) samples. However, the concentration of C (180 x 10¹⁵ cm^"2) and O (54 x 10¹⁵ cm^"2) contaminants in ¹⁴N implanted samples was observed to have increased slightly.

Annealing of these implanted regions resulted in a reduction of the C to 107 x 10 cm and O to 50 x 10 cm^" contamination although the final concentrations were still greater than that observed in virgin silicon (100). This increase in contamination concentration in implanted regions is believed to be a consequence of the implantation

127643 of residual gases during the implantation procedure. During annealing a fraction of these contaminants may desorb from the surface, as is evidenced by the reduced concentration. The presence of contamination does not suppress nanostructure growth. It is thus likely that the observed increase in contamination concentration does not explain the growth suppression. These results show that pre-implantation of ¹⁴N effectively suppresses spontaneous nanostructure growth on silicon (100) by EB-RTA.

The implantation of ¹⁴N or ¹⁵N ions into a silicon substrate before annealing the substrate using the method described with reference to Figure 1 suppresses the growth of nanostructures. By implanting ¹⁴N or ¹⁵N ions before annealing a silicon substrate areas of nanostructure and areas free of nanostructures can be created. This removes the need for alternative processing of an area of the substrate to prevent the formation of nanostructures before annealing using the method described with reference to Figure 1.

Nanoboulders are another nanostructure that can be formed using electron beam rapid thermal annealing (EB-RTA) as described with reference to Figure 1. Nanoboulders are boulder-like formations on the annealed surface of the silicon substrate and are formed by first implanting carbon ions into a surface of the silicon and then annealing that silicon surface.

Figure 10 is a flow chart showing the major steps performed when forming nanoboulders. In step 100 the target material is supplied. The target material may be (100), (110) or other p-type or n-type silicon.

In step 101 the target silicon is prepared for processing. The silicon substrate on which the nanoboulders are to be formed is cut to the desired target size. For example a typical target size may be 1 x 1 cm. The silicon substrate is preferably (100) or (111) p- type silicon. The silicon substrate is typically 0.5 mm thick and has resistivity of 1 - 10 Ωcm. However other thicknesses of silicon substrate may be used. To remove any dust and silicon cutting products, in step 102 the targets are mechanically cleaned by spraying pressured air onto the surface of the silicon substrate that will be annealed. In step 103 the target is loaded into an ion implantation facility, a high vacuum chamber. The mechanically cleaned surface of the target faces towards to electron beam direction. The pressure in the vacuum chamber is then reduced. Once the residual gas pressure of 10^"6 mbar (10^"4 Pa) is reached, the implantation process may start. In preferred embodiments the residual gas pressure is less than 10^"3 Pa.

In step 104 carbon ions are implanted into the target surface of the silicon substrate. A focussed carbon ion beam, typically low energy 10 keV (tilt θ = 0°), is raster scanned (ion current density j < 10 μA scanned) over the surface to allow for a homogeneous implantation under high vacuum (Residual gas pressure Po < 1 x 10^" mbar (1 x 10^" Pa). Once the requested ion fluence is reached, e.g. 25 at.%, the low energy ion beam is shut off and the target can be removed from the implantation chamber. The implantation ion fluence ranges between 10¹⁶ and 5 x 10¹⁷ cm^"2. The carbon ions implanted into the surface may be ¹³C⁺ ions or ¹²C⁺ ions.

In step 105 the silicon substrates are transported from the ion implantation chamber of an electron beam annealing chamber. The transport of the targets from the implantation chamber to the electron beam annealing facility chamber can be undertaken at air pressure. A few minutes to a few hours can be used for the transportation procedure. The target is loaded into the annealing facility. The polished side of the target faces towards to electron beam direction. Once the residual gas pressure of 10^"6 mbar (10^"4 Pa) is reached, the annealing process may start. This process is the same as that described with reference to Figure 1.

In step 106 the silicon substrates are annealed. This process can be broken into three steps, heating, holding the peak temperature, and cooling. The process typically begins at room temperature. In the first step the substrate surface is gently heated with a temperature gradient of, 5 - 10 °C/s, to a peak temperature. In the second step the peak temperature, between 800 and 1200 °C, is held for between 3 and 120 seconds. In the third step the substrate surface is gently cooled with a temperature gradient of 5 - 10 °C/s, to room temperature.

127643 Processed targets can stay in the vacuum chamber for long time, such as hours to allow processing on many targets under identical conditions. The nanoboulders formed by this process are silicon carbide (SiC).

Figure 11A shows an atomic force microscope image of a section of the mechanically cleaned surface of a silicon substrate after implantation with carbon ions. Figure 11B shows the variation in surface roughness of the mechanically cleaned surface of the silicon substrate after ion implantation. The variation in the surface is other order of O.lnm. This roughness is similar to that of the mechanically cleaned surface before ion implantation.

Figures 12A and 12B show nanoboulders produced by the manufacturing process using single sided mechanically cleaned p-type (100) silicon substrate materials (thickness 0.5 mm). The substrates were implanted with ¹³C⁺ ions with ion energy of 20 keV and fluence of 7.8 x 10¹⁶ cm^"2. The ion current density used was 2 μA/cm². The ions were implanted in an 8 mm circular implantation area at a temperature close to room temperature. The ion implantation chamber was operated in near-vacuum conditions with residual gas pressure P₀ = 2 x 10^" mbar (2 x 10^" Pa). The projected carbon range was calculated as 33 nm and the carbon concentration was calculated as 25 at%.

Annealing of the substrate was performed in an annealing facility. The temperature of the substrate surface was raised at a rate of 5°C/s to a peak temperature of 1000°C. The peak temperature was held for 15 seconds before being lowered at a rate of 5°C/s. These figures are atomic force microscope images of a portion of the implanted and annealed silicon substrate. Figure 12B shows the nanoboulders clearly.

Figure 12C is a section of the silicon substrate of Figures 12A and 12B. Figure 12C clearly shows the profile of the nanoboulders along the section taken. Figure 12D shows the section used in Figure 12C. As can be seen in Figure 12C the height of the tallest nanoboulder in this section is between 100 and 200 nm. The height of nanoboulders produced using the method of the invention are typically of the order of

127643 100 to 200 nm depending on the implantation fluence and annealing parameters. Nanoboulders of different sizes can be produced by varying the process parameters.

The method of the invention which involves forming the nanoboulders by carbon implantation and rapid thermal annealing enables arrays of 100 - 300 nm high boulders with width similar to heights to be grown on (100) silicon substrate material which have potential for application as field emitters in flat panel displays.

The angle of implantation may not be specifically important for the manufacturing process as the implantation energy must be adjusted to the incident angle.

Implantation of the silicon substrate does not affect the average roughness of the silicon substrate. For example for a silicon substrate that is flat with an average roughness in the order of 0.1 nm, after implantation the surface remains flat on that order of magnitude. After annealing, nanoboulders are produced which are in the order of 100 — 300 nm depending on the implantation fluence and annealing parameters. Much smaller and larger nanoboulders may be produced with varying process parameters.

Figures 13 and 14 show nanoboulders formed using the method of the invention. The substrates used to form the nanoboulders of Figures 13 and 14 were p-type (100) (B- doped, 1 - 10 Ωcm) and diced into 10 10 mm² samples. No surface cleaning or other treatments were performed on the substrates and the surface thus contains the native oxide (15-20 A thick) formed by exposure to air and residual contaminants, typically carbon-based compounds. The silicon samples were implanted with ¹³C at 10 keV under high vacuum conditions. The ion current density of the raster scanned ion beam was maintained at 5 μA cm^"2 to avoid heating effects during the implantation process. Isotope ¹³C was implanted to aid subsequent analysis through elimination of interferences with the regular ¹²C isotope. Annealing was performed using an EB-RTA

127643 system with electron energy of 20 keV and sample current up to 2 mA. All samples were annealed at 1000 °C and at abase pressure <lxl0^"6mbar.

Nanostructures were observed across the entire ¹³C implanted region while characteristic nanowhisker growth occurred in the un-implanted regions. A typical AFM image of nanostructures formed in (100) Si implanted at 10 keV with ¹³C⁺ ions and fluence, F = 3.6 x 10¹⁶ cm^"2 is shown in Figure 13. The ellipsoidal structures are randomly distributed throughout the region and AFM images show no commonality in the direction of nanostructure alignment. Under these implantation conditions the nanostructures exhibit an average diameter (along the major axis) of 330 nm and height of 87 nm. The structures cover the surface with a density of 0.54 μm^"2. The surface roughness in the region between nanostructures was found to be slightly greater than that exhibited by implanted silicon prior to annealing. Figure 14 shows a representative image of typical nanostructures formed following implantation with an increased fluence, F = 1.14 x 10¹⁷ cm^"2. The structures are identical in appearance to those observed following lower fluence implantation, however the number of structures has reduced considerably to a density of 0.07 μm^"2. Further, both the average diameter and height of the boulders have increased to 630 nm and 230 nm respectively. Again, the surface roughness between boulders is greater than that observed for untreated silicon.

Detailed structural analysis indicates that the modification in nanostructure geometry due to implanted fluence follow clear, linear relationships. This is shown in Figure 15 where the dependence between the volume and density of nanostructures, is plotted with respect to the implantation fluence. Here the boulders were treated as continuous, semi- ellipsoidal structures for calculation of the boulder volume. Manipulation of the implantation fluence can thus be exploited to accurately control the size and number of nanoboulders. These dependencies may be due to enhanced carbon diffusion, and hence crystal growth and coarsening, as a result of extended substrate damage due to increased implantation fluence. Additionally, a linear relationship between nanoboulder diameter and inter-boulder spacing was observed as displayed in Figure 16.

127643 Implanting neon ions in a silicon substrate before annealing leads to modification of the substrate surface in the form of cavities. Neon ion implantation occurs in an implantation chamber before annealing the silicon substrate in an annealing facility using the method described with reference to Figure 1. Figure 18 shows the surface of a silicon substrate after implantation with neon ions and Figure 19 shows the surface of a silicon substrate implanted with neon ions after annealing. In preferred embodiments neon ions are implanted with implantation energy between 5 and 30 keV. The preferred implantation current density is less than 20 μA/cm . The residual gas pressure during implantation is preferably less than 10^"3 Pa. In preferred embodiments the implantation fluence is between 10¹⁶ and 12 x 10¹⁷ cm^"2.

To produce the results shown in Figures 17 to 19 ²²Ne⁺ ions were implanted at 20 keV under normal incidence into (100) silicon substrate. The implantations were performed under high vacuum conditions at room temperature. Typically highly neon enriched amorphous surfaces containing up to 23 at% neon were produced with ion fluences ranging from 2.1 to 7.2 x 10 ions/cm in less than 15 minutes using an ion current density of 10μA/cm² that does not result in significant heating of the substrate surface. The ion beam was raster scanned over the silicon substrate.

Calculations predict that the neon implantation profiles are located close to the substrate surface. Figure 17 shows implantation profiles for neon ion implantation into (100) silicon with fluences of 3.6 x 10¹⁶ ions/cm² and 7.2 x 10¹⁶ ions/cm². These profiles were measured using resonant nuclear reaction analysis. As can be seen from this figure the neon ion concentration in the substrate is greater for the greater implantation fluence. Figure 17 also shows the neon ion profile after annealing the substrates. As shown there is a substantial decrease in neon within the substrate after annealing showing that the annealing process releases neon from the substrate. The annealing stage was performed in the manner described previously with a peak temperature of 1100°C which was maintained for 15 seconds. The temperature gradient for both the increasing and decreasing temperatures was 5°C/sec.

127643 Figure 18 is a 4μm x 4μm top view atomic force microscope image of a silicon substrate after implantation with neon ions with energy of 20 keV and fluence of 3.6 x 10¹⁶ ions/cm². This image shows a slight increase in the roughness of the silicon surface. Typically the average surface roughness for virgin silicon is 0.1 Onm and the roughness shown in Figure 18 is 0.13nm. Figure 18 also shows dust particles of up to 2.5nm. These dust particles are not unusual for non-chemically cleaned substrates.

Figure 19 is a 12μm x 12μm top view atomic force microscope image of a silicon substrate after implantation with neon ions and annealing. Annealing took place with a peak temperature of 1000°C held for 15 seconds and with an increasing and decreasing temperature gradient of 5°C/s. As can be seen in Figure 19 cavities have formed in the silicon surface. These cavities have an average diameter of 1.7 ± 0.2μm. The average depth of the cavities is 38 ± 2nm. This depth matches the projected range of the implanted neon ions. The average number of cavities is 0.07μm^" .

When neon ions were implanted into the silicon substrate with fluence of 7.2 x 10¹⁶ ions/cm the average number of cavities was 0.01 μm^" and the cavities had an average diameter of 1.0 ± 0.2μm and an average depth of 48.5 nm.

The large cavities formed using this process indicates that neon is released from the substrate as large bubbles during the annealing process. The change in cavity number and size with changing fluence suggests that the cavity dimensions can be controlled by changing the fluence of the implanted ions.

Implanting N ions using plasma immersion ion implantation into a silicon substrate before annealing leads to modification of the substrate surface in the form of cavities. Nitrogen ion implantation occurs in an implantation chamber before annealing the silicon substrate in an annealing facility. Plasma immersion ion implantation nitrogen implantation also implants oxygen into the substrate. The substrate surface then acts as an oxygen implanted surface which produces cavities when annealed using the method described with reference to Figure 1. Figures 20 to 22 show the surface of a silicon

127643 substrate after implantation with nitrogen ions using plasma immersion ion implantation and annealing.

To produce the results shown in Figures 20 to 22 N₂ ions were implanted using plasma immersion ion implantation at 10 keV under normal incidence into (100) and (111) silicon substrates. The implantations were performed under high vacuum conditions at room temperature in a clean implantation chamber. The gas pressure was lowered to 4.8 μbar. During implantation the substrate temperature was kept below 100°C. Different nitrogen to silicon ratios were achieved in the surface-near region by implanting nitrogen ions with various fluences.

Calculations were used to predict the mean projected range and peak concentration considering 5 keV N^4" ions. All of the substrates used in Figures 20 to 22 were annealed in the manner described previously with a peak temperature of 1000°C which was maintained for 60 seconds. The temperature gradient for both the increasing and decreasing temperatures was 5°C/sec.

Oxygen levels were measures in the substrates before and after annealing. Similar values of oxygen were measured on the substrates indicating that the annealing process does not promote the uptake of oxygen.

Figure 20 is a 3μm x 3μm top view atomic force microscope image of a (100) silicon substrate after implantation with nitrogen ions with energy of 10 keV and fluence of 5 x 10¹⁶ N ions/cm². After implantation the nitrogen areal density was 40 ± 2 x 10¹⁵ cm^"2. Oxygen was measured with an areal density of 50 ± 2 x 10¹⁵ cm^"2 resulting in an N/O ratio of approximately 1.0. After annealing the nitrogen areal density decreased to 35 x 10¹⁵ cm^"2 and the oxygen areal density decreased to 45 10¹⁵ cm^"2. The nitrogen fluence formed an understoichiometric (N/Si ratio < 1.0) silicon nitride layer. Small cavities less than 50nm in diameter and 3 nm in depth can be seen irregularly distributed over the substrate in Figure 20. These craters are formed from a loss of nitrogen and oxygen from the silicon surface.

127643 Figure 21 is a 4μm x 4μm top view atomic force microscope image of a (100) silicon substrate after implantation with nitrogen ions with energy of 10 keV and fluence of 1.6 x 10¹⁷ N ions/cm². After implantation the nitrogen areal density was 60 x 10¹⁵ cm^"2. Oxygen was measured with an areal density of 61 ± 2 x 10¹⁵ cm^"2 resulting in an N/O ratio of approximately 1.0. After annealing the nitrogen areal density decreased to 47 x 10¹⁵ cm^"2 and the oxygen areal density decreased to 31 x 10¹⁵ cm^"2. The nitrogen fluence formed an overstoichiometric (N/Si ratio > 1.0) silicon nitride layer. The surface of the substrate in Figure 21 is highly organised with large cavities with diameters of 1-2 μm and average depth of 80 nm. One such cavity can be seen in the bottom right of Figure 21. These craters are formed from a loss of nitrogen and oxygen from the silicon surface.

Figure 22 is a 4μm x 4μm top view atomic force microscope image of a (111) silicon substrate after implantation with nitrogen ions with energy of 10 keV and fluence of 1.0 10¹⁷ N ions/cm². After implantation the nitrogen areal density was 55 x 10¹⁵ cm^"2. Oxygen was measured with an areal density of 53 x 10 cm^" resulting in an N/O ratio of approximately 1.0. After annealing the nitrogen areal density decreased to 50 x 10¹⁵ cm^"2 and the oxygen areal density decreased to 31 x 10¹⁵ cm^"2. The nitrogen fluence formed a stoichiometric (N/Si ratio = 1.0) silicon nitride layer. Cavities with diameter about 0.5 μm and about 60 nm in depth can be seen distributed over the substrate in Figure 22. These craters are formed from a loss of nitrogen and oxygen from the silicon surface.

These results show that a change in fluence or stoichiometric ratio changes the size of the cavities formed.

The foregoing describes the invention including a preferred form thereof. Alterations and modifications as will be obvious to those skilled in the art are intended to be incorporated in the scope thereof.

127643

Claims

1. A method of forming nanostructures on a silicon substrate including the steps of: in a chamber heating the substrate with an electron beam to a peak temperature, holding the peak temperature for a predetermined time, and decreasing the temperature of the substrate.

2. A method of forming nanostructures on a silicon substrate as claimed in claim 1 wherein the substrate is heated at a constant rate.

3. A method of forming nanostructures on a silicon substrate as claimed in claim 2 wherein the rate of heating is in the range 5 - 10 °C/second.

4. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 3 wherein the temperature of the substrate is decreased at a constant rate.

5. A method of forming nanostructures on a silicon substrate as claimed in claim 4 wherein the rate of decrease is in the range 5 - 10 °C/second.

6. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 5 wherein the peak temperature is in the range 800 - 1200 °C.

7. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 6 wherein the peak temperature is held for between 3 and 120 seconds.

8. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 7 wherein the chamber is a high vacuum chamber.

9. A method of forming nanostructures on a silicon substrate as claimed in claim 8 wherein the pressure in the chamber is less than 10^"2 Pa while the nanostructures are being formed.

127643

10. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 9 wherein the electron beam is raster scanned over the silicon substrate.

11. A method of forming nanostructures on a silicon substrate as claimed in claim 10 wherein the scan rate of the electron beam is between 1 and 10 kHz.

12. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 11 wherein the substrate is (100) silicon.

13. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 11 wherein the substrate is (110) silicon.

14. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 11 wherein the substrate is (111) silicon.

15. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 14 wherein the substrate is p-type silicon.

16. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 14 wherein the substrate is n-type silicon.

17. A method of forming nanostructures on a silicon substrate as claimed in claim 15 wherein the substrate is Boron-doped.

18. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 17 wherein the substrate is implanted with carbon ions before the step of heating the substrate.

19. A method of forming nanostructures on a silicon substrate as claimed in claim 18 wherein the step of implanting the carbon ions uses a focussed beam of carbon ions.

127643

20. A method of forming nanostructures on a silicon substrate as claimed in claim 18 or claim 19 wherein the carbon ions are single charged.

21. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 20 wherein the carbon ions are implanted into the substrate with implantation energy between 5 and 20 keV.

22. A method of forming nanostructures on a silicon substrate as claimed in claim 21 wherein the implantation energy is about 10 keV.

23. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 22 wherein the carbon ions are implanted into the substrate with implantation current density less than 10 μA/cm^" .

24. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 23 wherein the carbon ions are implanted into the substrate with a gas pressure around the substrate of less than 10^"3 Pa.

25. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 24 wherein the carbon ions are implanted into the substrate with fluence between 10¹⁶ cm^"2 and 5 10¹⁷ cm^"2.

26. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 25 wherein the carbon ions are ¹³C⁺ ions.

27. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 18 to 25 wherein the carbon ions are ¹²C ions.

28. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 17 wherein the substrate is implanted with neon ions before the step of heating the substrate.

127643

29. A method of forming nanostructures on a silicon substrate as claimed in claim 28 wherein the step of implanting the neon ions uses a focussed beam of neon ions.

30. A method of forming nanostructures on a silicon substrate as claimed in claim 28 or claim 29 wherein the neon ions are single charged.

31. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 28 to 30 wherein the neon ions are implanted into the substrate with implantation energy between 5 and 30 keV.

32. A method of forming nanostructures on a silicon substrate as claimed in claim 31 wherein the implantation energy is about 20 keV.

33. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 28 to 32 wherein the neon ions are implanted into the subsfrate with implantation current density less than 20 μA/cm^"2.

34. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 28 to 33 wherein the neon ions are implanted into the substrate with a gas pressure around the substrate of less than 10^"3 Pa.

35. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 28 to 34 wherein the neon ions are implanted into the substrate with fluence between 10¹⁶ cm^"2 and 12 10¹⁷ cm^"2.

36. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 28 to 35 wherein the neon ions are ²²Ne⁺ ions.

37. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 1 to 17 wherein the substrate is implanted with nitrogen ions using plasma immersion ion implantation with implantation energy between 5 and 20 keV before the step of heating the substrate.

127643

38. A method of forming nanostructures on a silicon substrate as claimed in claim 37 wherein the step of implanting the nitrogen ions uses a focussed beam of nitrogen ions.

39. A method of forming nanostructures on a silicon substrate as claimed in claim 37 or claim 38 wherein the nitrogen ions are single charged.

40. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 37 to 39 wherein the implantation energy is about 10 keV.

41. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 37 to 40 wherein the nitrogen ions are implanted into the substrate with a gas pressure around the substrate of less than 10^" Pa.

42. A method of forming nanostructures on a silicon subsfrate as claimed in any one of claims 37 to 41 wherein the carbon ions are implanted into the substrate with fluence between 10¹⁶ cm^"2 and 5 x 10¹⁷ cm^"2.

43. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 37 to 42 wherein the carbon ions are ions.

44. A method of forming nanostructures on a silicon substrate as claimed in any one of claims 37 to 42 wherein the carbon ions are N⁺ ions.

45. A method of preventing the formation of nanostructures on a silicon substrate including the steps of: in a chamber implanting nitrogen ions with an implantation energy of between 50 and 150 keV into the silicon substrate, heating the substrate with an electron beam to a peak temperature, holding the peak temperature for a predetermined time, and decreasing the temperature of the substrate.

127643

46. A method of preventing the formation of nanostructures on a silicon substrate as claimed in claim 45 wherein the substrate is heated at a constant rate.

47. A method of preventing the formation of nanostructures on a silicon substrate as claimed in claim 46 wherein the rate of heating is in the range 5 - 10 °C/second.

48. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 47 wherein the temperature of the substrate is decreased at a constant rate.

49. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in claim 48 wherein the rate of decrease is in the range 5 - 10 °C/second.

50. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 49 wherein the peak temperature is in the range 800

- 1200 °C.

51. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 50 wherein the peak temperature is held for between 3 and 120 seconds.

52. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 51 wherein the chamber is a high vacuum chamber.

53. A method of preventing the formation of nanostructures on a silicon substrate as claimed in claim 52 wherein the pressure in the chamber is less than 10^"2 Pa.

54. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 53 wherein the electron beam is raster scanned over the silicon substrate.

127643

55. A method of preventing the formation of nanostructures on a silicon substrate as claimed in claim 54 wherein the scan rate of the electron beam is between 1 and 10 kHz.

56. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 55 wherein the substrate is (100) silicon.

57. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 55 wherein the substrate is (110) silicon.

58. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 55 wherein the subsfrate is (111) silicon.

59. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 58 wherein the subsfrate is p-type silicon.

60. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 59 wherein the substrate is n-type silicon.

61. A method of preventing the formation of nanostructures on a silicon substrate as claimed in claim 60 wherein the substrate is Boron-doped.

62. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 61 wherein the step of implanting the nitrogen ions uses a focussed beam of nitrogen ions.

63. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in claim 62 wherein the nitrogen ions are single charged.

64. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 63 wherein the implantation energy is about 50 keV.

65. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 63 wherein the implantation energy is about 100 keV.

66. A method of preventing the formation of nanostructures on a silicon subsfrate as claimed in any one of claims 45 to 65 wherein the nitrogen ions are implanted into the substrate with a gas pressure around the subsfrate of less than 10^"3 Pa.

67. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 66 wherein the carbon ions are implanted into the

17 0 17 0 substrate with fluence between 10 cm and 5 10 cm .

68. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 67 wherein the nitrqgen ions are ¹⁴N⁺ ions.

69. A method of preventing the formation of nanostructures on a silicon substrate as claimed in any one of claims 45 to 67 wherein the carbon ions are ¹⁵N ions.

70. A nanostructure on a silicon substrate formed using the method of any one of claims 1 to 44.

71. A silicon substrate including a nanostructure formed using the method of any one of claims 1 to 44.

127643