Deposition Method

20210317565 · 2021-10-14

    Inventors

    Cpc classification

    International classification

    Abstract

    Sputter depositing a metallic layer on a substrate in the fabrication of a resonator device includes providing a magnetron sputtering apparatus comprising a chamber, a substrate support disposed within the chamber, a target made from a metallic material, and a plasma generating device, wherein the substrate support and the target are separated by a distance of 10 cm or less; supporting the substrate on the substrate support; performing a DC magnetron sputtering step that comprises sputtering the metallic material from the target onto the substrate so as to form a metallic layer on the substrate, wherein during the DC magnetron sputtering step the chamber has a pressure of at least 6 mTorr of a noble gas, the target is supplied with a power having a power density of at least 6 W/cm.sup.2, and the substrate has a temperature in the range of 200-600° C.

    Claims

    1. A method of sputter depositing a metallic layer on a substrate in the fabrication of a resonator device, the method comprising the steps of: providing a magnetron sputtering apparatus comprising a chamber, a substrate support disposed within the chamber, a target made from a metallic material, and a plasma generating device, wherein the substrate support and the target are separated by a distance of 10 cm or less; supporting the substrate on the substrate support; and performing a DC magnetron sputtering step that comprises sputtering the metallic material from the target onto the substrate so as to form a metallic layer on the substrate, wherein the metallic layer is composed of a metal selected from molybdenum (Mo), tungsten (W), titanium (Ti), tantalum (Ta), platinum (Pt) or ruthenium (Ru), and wherein during the DC magnetron sputtering step the chamber has a pressure of at least 6 mTorr of a noble gas, the target is supplied with a power having a power density of at least 6 W/cm.sup.2, and the substrate has a temperature in the range of 200-600° C.

    2. The method according to claim 1, wherein the substrate is a silicon substrate.

    3. The method according to claim 1, wherein the power density is at least 8 W/cm.sup.2.

    4. The method according to claim 1, wherein the power density is less than about 24 W/cm.sup.2.

    5. The method according to claim 1, wherein during the DC magnetron sputtering step the chamber has a pressure of at least 7 mTorr.

    6. The method according to claim 1, wherein during the DC magnetron sputtering step the chamber has a pressure of about 20 mTorr or less.

    7. The method according to claim 1, wherein the DC magnetron sputtering step is performed at a temperature in the range of 300-500° C.

    8. The method according to claim 1, wherein during the DC magnetron sputtering step, a bias power is supplied to the substrate support.

    9. The method according to claim 8, wherein the bias power is an RF bias power.

    10. The method according to claim 8, wherein the bias power has a power in the range of 10-600 W.

    11. The method according to claim 1, wherein the DC magnetron sputtering step comprises sputtering the metallic material from the target using a plasma formed from the noble gas, and wherein the noble gas is argon (Ar), krypton (Kr), xenon (Xe), or a mixture thereof.

    12. The method according to claim 1, wherein the metallic layer has a thickness of about 300 nm or less.

    13. The method according to claim 1, wherein the distance is about 75 mm or less.

    14. The method according to claim 1, wherein the resonator device is an acoustic wave device.

    15. A method of fabricating a resonator device comprising the steps of: (a) providing a substrate; (b) depositing a first metallic layer onto the substrate; (c) depositing a piezoelectric layer onto the first metallic layer; and (d) depositing a second metallic layer onto the piezoelectric layer, wherein at least step (b) or step (d) is performed using the method according to claim 1.

    16. A substrate comprising a molybdenum layer deposited using the method according to claim 1, wherein the molybdenum layer has a resistivity of less than about 10 μΩ.Math.cm.

    17. A resonator device comprising a molybdenum layer deposited using the method according to claim 1, wherein the molybdenum layer has a resistivity of less than about 10 μΩ.Math.cm.

    18. A resonator device according to claim 17, wherein the resonator device is an acoustic wave device.

    Description

    DESCRIPTION OF THE DRAWINGS

    [0038] Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

    [0039] FIG. 1 is a schematic illustration of a sputtering apparatus suitable for performing methods in accordance with embodiments of the present invention;

    [0040] FIG. 2 shows the thickness variation of a metallic layer deposited using an exemplary method of the present invention;

    [0041] FIG. 3 shows the radial sheet resistivity of a metallic layer deposited using an exemplary method of the present invention;

    [0042] FIG. 4 shows a plot of resistivity (in μΩ.Math.cm) as a function of pressure (in mTorr);

    [0043] FIG. 5 shows average stress and stress range in a metallic layer as a function of platen power; and

    [0044] FIG. 6 shows average stress in a metallic layer as a function of platen power.

    DETAILED DESCRIPTION OF THE DISCLOSURE

    [0045] Embodiments of the present invention relate to methods of sputter depositing a metallic layer onto a substrate in the fabrication of a resonator device, such as a bulk acoustic wave (BAW) or surface acoustic wave (SAW) device. Exemplary methods of the present invention can be performed using a suitable DC magnetron sputtering apparatus 10, as shown schematically in FIG. 1. A SPTS Sigma® deposition system with a DC Magnetron sputtering module, which is commercially available from SPTS Technologies Limited based in Newport, South Wales, UK, is suitable for performing DC magnetron sputtering methods according to embodiments of the present invention. Some well-known features of the DC magnetron sputtering apparatus 10 have been omitted from the figures for the sake of simplicity. The general operation of such a DC magnetron sputtering apparatus (e.g. generating a plasma) is well-known in the art and will not be described here other than where necessary for an understanding of the present invention.

    [0046] Referring to FIG. 1, the magnetron sputtering apparatus 10 comprises a substrate support (e.g. a platen) 12, which is disposed within a chamber 14. The grounded, metallic chamber 14 is suitable for containing a plasma. The substrate support 12 is adapted to hold a substrate 16 in a substantially horizontal orientation, with a front surface of the substrate facing upwards. The substrate 16 can be a wafer, such as a silicon wafer. The substrate 16 can be a patterned silicon wafer. The substrate 16 can be loaded and unloaded via a suitable loading/unloading slot 17.

    [0047] The apparatus 10 further comprises a target 18 disposed within the chamber 14. The target 18 and the substrate support 12 are arranged such that material sputtered from the target 18 can be deposited onto a substrate 16 being supported on the substrate support 12. Typically, the target 18 is held directly opposite and above the substrate support 12. The distance between the target 18 and the substrate 16 is about 10 cm or less, or optionally 75 mm or less. The target 18 is made from a sputtering material. The target 18 is made from a metallic material. The target can be made from molybdenum (Mo), tungsten (W), ruthenium (Ru), titanium (Ti), tantalum (Ta) or platinum (Pt). In exemplary methods of the present invention, the metallic material is typically a metal suitable for use as an electrode metal in a resonator device, such as a low thermo-elastic loss electrode metal.

    [0048] Gas can enter the chamber 14 via a gas inlet 20, and can be pumped from the chamber via a gas exhaust 22. The apparatus 10 further comprises a plasma generating means for generating a plasma within the chamber 14. The apparatus 10 further comprises a magnetron 24 for generating a magnetic field proximate to the target 18 that localises the plasma adjacent the target 18. The magnetron 24 is disposed outside of the chamber 14. A power supply 26 is configured to provide a DC power to the magnetron 24 during operation.

    [0049] A power supply 26 is configured to supply a power to the target 18. During a sputter deposition process, the target 18 acts as a cathode.

    [0050] A power supply 28 can supply a RF bias power to the substrate support 12. A controller with a graphical user interface (not shown) may be used to control the power supplies.

    [0051] In operation, a plasma is sustained by the plasma generating means and a power is applied to the target 18 so that species in the plasma (e.g. ions and neutral atoms) sputter material from the target 18. The sputtered material from the target 18 deposits onto the substrate 16 being supported on the substrate support 12 to form a metallic layer thereon. In a non-reactive sputtering process, the metallic layer is typically composed of the same material as the target 18.

    [0052] In a first exemplary method, metallic layers were sputter deposited onto a 200 mm diameter silicon wafer using a DC magnetron sputtering method. This deposition process is part of a process for manufacturing a resonator device such as an acoustic wave device (e.g. a BAW or SAW device). The apparatus 10 used to deposit the metallic layer was a SPTS Sigma® deposition system with a DC magnetron sputtering module. In this example, the target 18 was made from molybdenum, and the metallic layers were composed of molybdenum.

    [0053] The silicon substrate 16 was positioned onto the substrate support 12 within the deposition apparatus 10. The distance between the target 18 and the substrate 16 was set to about 10 cm.

    [0054] A DC magnetron sputtering process was then performed to sputter Mo from the target 18 onto the substrate 16. A noble gas was introduced into the chamber 14 via a gas inlet 20 and a plasma generated. The noble gas can be argon, krypton, xenon or any combination thereof. In the first exemplary method, the noble gas used to generate the plasma was argon. A power was applied to the target (cathode) 18 so as to encourage the species in the plasma to sputter material from the target 18.

    [0055] During the DC magnetron sputtering process, the chamber 14 has a pressure of at least 6 mTorr, and the power applied to the target 18 has a power density of at least 6 W/cm.sup.2. Typically, the chamber pressure is in the range 6-20 mTorr. Typically, the target power density is in the range 8-24 W/cm.sup.2.

    [0056] The thickness, resistivity and stress of the deposited metallic layers were measured. The thickness of the metallic layer was measured using a non-destructive metrology measurement using a MetaPULSE™ instrument. FIG. 2 shows a contour map of the thickness of the as deposited Mo metallic layer. The average film thickness was 189.36 nm and the film showed excellent thickness uniformity with a 1σ value of 0.458% (see Table 1).

    TABLE-US-00001 TABLE 1 Thickness Thickness Rs Rs Bulk resistivity (nm) 1sigma % (Ω/sq) 1sigma % (μΩ .Math. cm) 189.36 0.458 0.4437 0.721 8.40

    [0057] The sheet resistivity (R.sub.s) of the metallic layer was measured using a four-point probe technique, and the resistivity determined using the following equation.


    R.sub.s=resistivity/film thickness

    [0058] FIG. 3 shows a contour map of the sheet resistivity of the as deposited film.

    [0059] The average bulk resistivity was 8.4 μΩ.Math.cm, as represented by the (thickest) lines 32. Values at one standard deviation are represented by (intermediate thickness) lines 34. The resistivity values achieved using the present method are far superior to (i.e. lower than) those known to have been previously reported. Additionally, the present method enables excellent thickness uniformity to be achieved.

    [0060] The stress of the as deposited film can be controlled by applying a bias power, such as an RF bias power, to the substrate support 12 during the DC magnetron sputtering process. Typically, the RF bias power has a frequency of 13.56 Hz. However, other RF frequencies could be used. Table 2 shows how the bias power supplied to the substrate support affects the resistivity and stress in the as deposited Mo films under the same high power (i.e. at least 6 W/cm.sup.2) and high pressure (i.e. at least 6 mTorr) conditions.

    TABLE-US-00002 TABLE 2 Platen power Resistivity Resistivity Avg. Stress Stress Range (W) (μΩ .Math. cm) 1σ% (MPa) (MPa) 400 8.83 0.62 −198 183 275 9.00 0.68 16 118 215 8.92 1.10 185 91

    [0061] By applying an appropriate bias power to the substrate support (e.g. platen) 12, the average stress in the deposited metallic layer can be tuned. In general, a higher bias power applied to the substrate support 12 during the DC magnetron sputtering step causes the stress in the as deposited metallic layer to be more compressive (less tensile). The bias power supplied to the substrate support 12 can be tuned so that the deposited metallic layer can have an average stress of approximately 0 MPa. Furthermore, the high power and high pressure deposition conditions according to the present invention provide a resultant metallic layer exhibiting a stress variation across the film of <200 MPa. Further still, all films exhibited a low resistivity of μΩ.Math.cm. Such metallic layers with uniform thickness, uniform stress, and low resistivity exhibit desirable properties that can be used to fabricate higher quality, smaller resonator devices, such as BAW and SAW devices with fewer losses.

    [0062] These methods and results go against the currently received wisdom in the art. The currently received wisdom suggests that increasing the chamber pressure during the DC magnetron sputtering step, causes the sputtered material to be scattered to a greater degree and the energy of the sputtered material landing on the substrate 16 is decreased. This is believed to cause an increase in gas incorporation in the as deposited film, resulting in a less dense and less conductive metallic layer. However, the present inventors have found that contrary to this currently received wisdom, the combination of a high pressure (i.e. at least 6 mTorr) together with a high target power density (i.e. at least 6 W/cm.sup.2) can result in deposited metallic layers having an improved (lower) resistivity, whilst also affording excellent thickness uniformity and stress uniformity. The excellent (low) resistivity values of the metallic layer deposited using the present method suggests that the films are dense with minimal gas incorporation in the as deposited films. These metallic layers have improved characteristics and allow resonator devices with an improve quality factor to be produced.

    [0063] FIG. 4 shows a plot of how the resistivity of a 300 nm thick Mo metallic layer deposited using DC Magnetron sputtering varied as a function of chamber pressure (without any bias applied to the substrate support 12). Plot A used a power density applied to the target 18 of 6.77 W/cm.sup.2. Plot B used a power density applied to the target 18 of 16.92 W/cm.sup.2. The deposition temperature was 200° C.

    [0064] Using a power density of 6.77 W/cm.sup.2 (FIG. 4, Plot A) provided excellent resistivities (<9 μΩ.Math.cm) at both high pressure (approx. 7.8 mTorr) and at low pressure (approx. 1.4 mTorr). However, increasing the pressure was found to further improve the resistivity (to <8.8 μΩ.Math.cm). Operating at a higher pressure also provided a more tensile film, the stress of which could be tuned using a RF bias power applied to the substrate support 12. A RF bias power applied to the substrate support 12 generally causes the average stress in the metallic layer to become more compressive (less tensile). Therefore, in order to better control the average stress of the deposited layer, it is preferable that when no bias power is applied the deposited layer is slightly tensile. That is, the average stress of a more tensile film (e.g. deposited at higher pressure) can be tuned so that the average stress is about 0 MPa. The excellent (low) resistivity values suggest that these films are dense.

    [0065] Using a power density of 16.92 W/cm.sup.2 (FIG. 4, Plot B) at a low pressure (approximately 1.4 mTorr) resulted in a metallic layer having a sub-optimal resistivity (>10 μΩ.Math.cm). However, as the pressure is increased (to approximately 7.8 mTorr), the sensitivity to the high power density operation is reduced enabling higher deposition rates while reducing the resistivity values of the deposited metallic layer (<9 μΩ.Math.cm). FIG. 5 shows how the average stress (line 50, represented by circles) and stress range (line 52, represented by triangles) in an as deposited 200 nm thick Mo metallic layer varies as a function of bias power applied to the substrate support 12. The Mo metallic layer was deposited at a temperature of 200° C., at a pressure of about 8 mTorr and with a target power density of 6.77 W/cm.sup.2. A substrate support bias power of about 240 W resulted in the Mo metallic layer having an average stress of about 0 MPa and a stress range of about 160 MPa.

    [0066] FIG. 6 shows how the average stress (line 60) in an as deposited 200 nm thick Mo metallic layer varies as a function of bias power applied to the substrate support 12. The Mo metallic layer was deposited at a temperature of 200° C., at a pressure of about 8 mTorr and with a target power density of 16.92 W/cm.sup.2. A substrate support bias power of about 95 W resulted in the Mo metallic layer having an average stress of about 0 MPa.

    [0067] Using the combination of high target power density together with high pressure (in accordance with the present invention), it was possible to deposit metallic layers that exhibited excellent stress uniformity of <120 MPa, and a thickness uniformity of <0.7 1σ % for a metallic layer having an average stress of 0 MPa. Additionally, low within wafer stress ranges of <200 MPa can be achieved at a variety of average stress while maintaining a thickness non-uniformity of <1.5% 1σ.

    [0068] Table 3 illustrates exemplary processing parameters that can be used in accordance with exemplary methods of the present invention to deposit a metallic layer exhibiting an average stress of about zero. A comparative process is also shown.

    TABLE-US-00003 TABLE 3 Comparative Process Present invention Pressure (mTorr) 2-4 6-12 Platen RF (W) 0 10-600 Cathode power density 1-4 6-20 (W/cm.sup.2) Resistivity (μΩ .Math. cm) 10-12 8-10 Stress Range (MPa) >350 <250

    [0069] Without being bound by any theory or conjecture, it is believed that operating at a high pressure and target power density (in accordance with the present invention), there is sufficient flux of energetic ions and neutral species bombarding the substrate 16 to create a dense film with minimal levels of gas incorporation. The high target (cathode) power density compensates for any additional gas scattering that might otherwise result due to the higher chamber pressure. In turn, the high target power density compensates for any attenuation in energy of the ions and neutral species in the plasma due to an increase in chamber pressure, and thus provides suitable conditions to achieve a dense film. Therefore, the combination of using a high power density applied to the target 18 with a high chamber pressure causes the species in the plasma to have an optimum energy so as to form dense metallic layers exhibiting improved resistivity, excellent thickness uniformity, and stress control with excellent stress uniformity. Such films are suitable for applications as electrode layers in resonator devices, such as BAW and SAW devices.

    [0070] Using a high chamber pressure and a high target (cathode) power density coupled with a controlled bias power applied to the substrate support 12, the present inventors have developed a method to deposit metallic layers exhibiting excellent thickness uniformity, improved resistivity (<10 μΩ.Math.cm), and reduced stress range. Whilst the examples above primarily relate to the deposition of Mo metallic layers, the same process conditions can be used to sputter deposit other metals, such as tungsten, titanium, tantalum, platinum and ruthenium. The metallic layers deposited using the present methods meet the requirements for highly uniform, low stress Mo thin films, and are suitable for use as thin electrode layers in resonator devices such as BAW and SAW devices.

    [0071] In a further exemplary method a resonator device, such as a BAW or SAW device was produced. A first metallic layer was sputtered deposited onto a substrate using the same DC magnetron sputtering methods described in relation to the first exemplary method. The substrate was a patterned silicon wafer.

    [0072] Subsequently, a piezoelectric material was deposited onto the first metallic layer. The piezoelectric material layer can be made from AlN, AlScN or other suitable piezoelectric material.

    [0073] A second metallic layer was then sputtered deposited onto the piezoelectric material layer using the same DC magnetron sputtering methods described in relation to the first exemplary method. The resulting resonator devices exhibited an improved quality factor.