HIGH TEMPERATURE SUPERCONDUCTING FILMS ON ALUMINUM OXIDE SUBSTRATES

Cross Reference To Related Applications

This application is a continuation-in-part of Serial Number 494,730 filed March 16, 1990, Serial Number 495,568 filed March 16, 1990 and Serial Number 565,691 filed August 13, 1990.

Technical Field

This invention relates to high critical temperature superconducting (HTSC) films and, more particularly, this invention relates to the preparation of HTSC films on sap¬ phire (aluminum oxide) substrates by the use of stable, epitaxial buffer layers.

Background of the Invention The recent discovery of high critical temperature superconducting materials has created interest in the use of these materials in microwave devices. The HTSC metal cuprate materials cannot readily be produced in bulk with geometries and properties suitable for microwave devices. Presently, devices based on the HTSC materials are fabri¬ cated by formation of thin films on substrates by tech¬ niques similar to those used to fabricate semiconductor devices. To be useful in a microwave device the HTSC film must be grown on a substrate having low dielectric losses at high frequencies.

Silicon, a readily used substrate for semiconductor devices, cannot be used with the HTSC cuprate materials since at the temperature prevalent during deposition the cuprate reactants readily react with silicon. Sapphire (Al ₂ 0 ₃ ) would appear to be an excellent substrate since it has a very low dielectric loss and is a strong, low cost, highly crystalline material available commercially in large sizes. Sapphire would be a good substrate for bolometer

(infrared detector) applications since thin films of sap- phire have low heat capacity. Epitaxial films of a HTSC

material such as YBa ₂ Cu ₃ 0 ₇ can be grown on sapphire substrat¬ es. However, the optimum window of substrate temperatures during film deposition is relatively narrow [4], At high temperatures above about 700 K, the YBa ₂ Cu ₃ 0 ₇ thin film reacts with the sapphire, especially the Ba atoms. At low temperature, below about 650 K, it is very difficult to produce thin YBa ₂ Cu ₃ 0 ₇ films having good epitaxy.

There are other substrate materials that provide epi¬ taxial growth of thin HTSC films with high superconducting transition temperatures and low rf surface resistance, such as strontium titanate (SrTi0 ₃ ) , calcium titanate (CaTi0 ₃ ) , lanthanum alu inate (LaAlO-) , magnesium oxide (MgO) and yttria stabilized zirconia (YSZ) . However, thick substrat¬ es of some of these materials, e.g. strontium or calcium titanate, exhibit too high an rf loss or do not have high enough strength to act as a substrate for large area micro¬ wave devices. Some of these materials are only available in small sizes and/or at very high cost.

However, thin epitaxial films of these materials ex- hibit a low dielectric loss and would be useful as a buffer layer between the HTSC film and the sapphire substrate if they provided an epitaxial surface for the HTSC film and were stable and non-reactive with the HTSC film and the substrate.

List of References

1. Multilayer YBa ₂ Cu ₃ O _χ - SrTi0 ₃ - YBa ₂ Cu ₃ O _χ Films For Insulating Crossovers, Kingston et al., Applied Phys¬ ics Letters, January 8, 1990.

I . YBa ₂ Cu ₃ 0_ Films Grown on Epitaxial MgO Buffer Layers on Sapphire, Talvacchio et al., Proceedings M ² S-HTSC, Stanford, July 1989, Physica. 3. The Sputter Deposition and Characterization of Epitax¬ ial Magnesium Oxide Thin Films and Their Use as a Sap¬ phire/ YBCO Buffer Layer, Morris et al., Proc. M.A.S. Vol 169, 1990. 4. Properties of Epitaxial YBa ₂ Cu ₃ 0 ₇ Thin Films on A1 ₂ 0 ₃ {1012}, Char et al., Appl. Phys. Lett. 56(8) 19 Feb 1990, p.785-787.

Statement of the Prior Art

Kingston et al. [1] disclose the use of SrTi0 ₃ as an insulating layer between two YB ₂ Cu ₃ O _χ films, one of which was deposited on an MgO substrate. MgO is not the best sub¬ strate for microwave applications due to its high dielec¬ tric loss. MgO does not provide a fully compatible lattice match with HTSC films such as YBa ₂ Cu ₃ 0 ₇ and polished MgO surfaces degrade in air with humidity. Talvacchio et al. [2] grew YBa ₂ Cu ₃ 0 ₇ films on epitaxial MgO buffer layers on sapphire. This layered structure failed as a microwave device since the MgO was excessively moisture sensitive. The HTSC film did not have sufficient orientation and there were too many random grains. Morris et al. [3] formed an MgO buffer layer by sputter deposition at 5.5 to 7 Pa (about 40-50 m Torr) a fairly low pressure. The resulting YBCO film grown on the epitaxial MgO has high normal state resistance and a broad superconducting transition.

Statement of the Invention

The invention provides a HTSC buffer layered structure on sapphire that is stable and suitable for microwave en¬ vironments or for use in the fabrication of bolometers. The layered structure with a thin SrTi0 ₃ or CaTi0 ₃ buffer layer has a sharp transition and good thermal response, making it suitable for bolometer applications. The HTSC film can be grown over a wide temperature range without the HTSC reacting with the sapphire substrate or the buffer reacting with the substrate or the HTSC film. The layered structures of the invention provides the highest microwave performance of any sapphire supported HTSC film reported to date. Others have reported surface resistance measuring on a par with copper. The layered sapphire supported HTSC structure of the invention exhibits rf surface resistance 2 to 3 times better than copper at 10 GHz and 77 K and much better than copper at lower temperatures. The YBCO films grown on the buffer layer exhibit low normal state resist-

ance and a narrow superconducting transition.

The buffer layer in the HTSC structure of the inven¬ tion is grown in a high pressure process such as laser ab¬ lation, sputtering or metallo-organic compound vapor depo- sition (MOCVD) . Talvacchio et al. [2] used electron beam deposition of the MgO film. Electron beam epitaxial depo¬ sition is conducted at relatively high vacuum, about 10 ^"5 Torr. The low partial pressure of oxygen is believed to result in an oxygen deficient buffer layer film. These films are chemically active and react with water which de¬ grades the quality of the overlying HTSC film. The buffer layer in the process of the invention is deposited at high¬ er pressure which results in inert buffer layers with good epitaxial qualities. The SrTi0 ₃ and CaTi0 ₃ buffer layers are excellent intermediate substrates for forming high per¬ formance, stable HTSC films.

These and many other objects and attendant advantages of the invention will become apparent as the invention be¬ comes better understood by reference to the following and detailed descriptions when considered in conjunction with the accompanying drawings.

Brief Description of the Drawings

Figure 1 is a schematic drawing of a layered HTSC-Buf- fer-Sapphire structure produced according to the process of this invention;

Figure 2 is a series of curves showing Surface Resis¬ tance, R _s of epitaxial YBa ₂ Cu ₃ 0 ₇ films as a function of tem¬ perature at 10 GHz; Figure 3 is a series of plots of Surface Resistance, R _s vs. frequency with Cu data at 77 K and Nb data at 7.7 K;

Figure 4(a) is a c-axis scan of the YBa ₂ Cu ₃ 0 ₇ film;

Figure 4(b) is a ø-axis scan of the [103] peak of the YBa ₂ Cu ₃ 0 ₇ film; Figure 5(a) is the mutual inductance response of the YBa ₂ Cu ₃ 0 ₇ film at 1 A;

Figure 5(b) is the mutual inductance response of the

YBa ₂ Cu ₃ 0 ₇ film at 100 mA; and

Figure 6 is a graph of surface resistivity, R _s versus temperature.

Detailed Description of the Invention

Referring now to Figure 1, the layered structure 10 is composed of a sapphire base or substrate 12, a thin buffer layer 14 and a HTSC layer 16. The buffer layer 14 and the HTSC layer 16 may be laid down in a pattern to form a de- vice by use of a mask formed by conventional lithographic techniques or by use of shutters or shields.

Sapphire substrates are commercially available in high purity and in a variety of thicknesses and shapes. The support need only have sufficient thickness such as 0.1 millimeters to provide a mechanically strong substrate. Substrates having thicknesses above 1.0 millimeters would usually not be utilized since they add cost and weight to the device without providing any other benefit.

The buffer layer is an epitaxial layer having suffi- cient thickness such that the growth of HTSC film is influ¬ enced solely by the buffer layer and not by the sapphire substrate. The buffer layer can have a minimum thickness of a monolayer or less. However, it is preferred that the buffer layer cover the sapphire surface and any anomalies on the surface and suitably has a thickness of at least 50 Angstroms. Thicknesses above about 3000 Angstroms are un¬ necessary and may interfere with the microwave device. The buffer can be formed from crystalline materials having a close epitaxial match to the HTSC material such as stron- tiu . titanate, calcium titanate, lanthanum alu inate, mag¬ nesium oxide or yttrium stabilized zirconia (YSZ) . Stron¬ tium titanate and calcium titanate are preferred since it appears that they form the best HTSC films.

In the process of the invention, the thin layer of buffer is formed by deposition in a chamber having a high percentage, usually from 10 to 100% of oxygen. The pres¬ sure in the chamber is higher than practiced in the elec-

tron beam deposition process and in other deposition pro¬ cesses. The pressure in the chamber is at least 40 m Torr and can be as high as 10 Torr, usually about 100 to 500 m Torr. The buffer layer may be formed by a variety of vapor deposition techniques such as on- or off-axis sputtering, metallo-organic vapor deposition (MOCVD) or laser ablation. Laser ablation is preferred since it appears to pro¬ vide a buffer layer that provides the highest quality HTSC films. This may be due to several factors. Laser ablation is conducted at a fairly high pressure. The deposition chamber contains from 20 to 100% oxygen at a pressure typi¬ cally from 100 Torr up to several Torr. The higher oxy¬ gen pressure could provide a more stable, more perfect crystalline epitaxial layer. Laser ablation generates its own plasma. The charged ionic species may assemble into a more ordered crystal form. The laser is pulsed during las¬ er ablation. During the non-pulsed period the growing crystal can relax to allow the metal and oxygen atoms to assume their positions in the crystal lattice of strontium titanate or calcium titanate.

The HTSC film is preferably grown by off-axis sputter¬ ing, laser ablation or MOCVD. Any of these procedures can be used to form the buffer layer. A common chamber can be used to form the buffer layer and HTSC film by two consecu- tive laser-ablation depositions or two consecutive off-axis sputtering depositions.

In off-axis sputtering the sapphire substrate or the buffer layer coated sapphire substrate is placed on a heat¬ ed substrate holder in a sputtering chamber at an angle of at least 40°, usually 90°, from the sputtering source. The substrate is heated to a temperature of from about 650 K to 800 K. The chamber contains from 10-50% of an oxidizing gas such as oxygen or nitrous oxide. The vapor source is a composite ceramic in the correct stoichiometric ratio for the film such as a Y_,Ba ₂ Cu ₃ alloy or a SrTi0 ₃ crystal or pressed powder source. Deposition is usually conducted over several hours at a high pressure between 100 and 500

Torr. A post deposition treatment in oxygen at a tempera¬ ture from 400 K to 600 K can be conducted in case of the HTSC film.

The films of high critical temperature (T _c ) supercon- ducting materials (HTSC) prepared in the present invention are mixed metal cuprates or mixed metal bismuthates having a T _c above about 30 K, usually above 70 K which belong to four families: the rare earth cuprates, the thallium-based cuprates, the bismuth-based cuprates and the alkaline earth bismuthates. The HTSC materials have an ordered lattice and are usually crystalline ceramics of the general formula

M"x ¹ My ² CuzOn where M ¹ is a Group IIIA metal, Group IIIB metal, Group VB or a rare earth metal, M ² is a Group IIA metal and x, y, z and n are integers. Usually the ratio of y:x is about 2:1 and the ratio of z:x is at least 3 usually from 3:1 to 6:1. The oxygen is present in an amount to satisfy valency of the metals and n is usually no more than 20, typically about 5-15. M ₁ can be a Group IIIA metal such as Yttrium (Y) or lanthanum (La) , a Group IIIB metal such as Thallium (Tl) , or Group VB metal such as Bismuth (Bi) or a rare earth met¬ al such as Erbium (Er) , Cerium (Ce) , Praseodymium (Pr) , Samarium (Sm) , Europium (Eu) , Gadolinium (Gd) , Terbium (Tb) , Dysprosium (Dy) , Holmium (Ho) , Ytterbium(Yb) , Lute- tiu (Lu) or combinations of these metals. M ² is a Group IIA metal such as strontium, barium, calcium or mixtures thereof.

The examples of practice of the invention will be di- rected to the YBaCuO materials of the general formula Y.,Ba ₂

Cu ₃ 0 ₇ , but the invention is equally applicable to other HTSC materials of the cuprate family or bismuthate family such as the bismuth cuprate of the general formula BiSrCaCuO or the thallium cuprate of the general formula TlBaCaCuO. The invention will now be illustrated by specific ex¬ amples. 500 Angstrom thick films of YBa ₂ Cu ₃ 0 ₇ (123) were grown on 500 Angstrom thick buffer layers of SrTiO- or

CaTi0 ₃ . The structural and electrical properties are char¬ acterized by x-ray diffraction data, SEM images and AC mu¬ tual inductance responses.

Both laser ablation and off-axis sputtering techniques were utilized in growing low surface resistance "123" films on A1 ₂ 0 ₃ with SrTi0 ₃ buffer layers. In the case of laser ablation about 1.8 Joule/cm ² energy density of 248 nm wave¬ length pulsed excimer laser beam was focused on a SrTi0 ₃ or an "123" pellet. Other deposition parameters were 200 m Torr of oxygen pressure and 750 C substrate temperature. In the case of off-axis sputtering two sputtering guns were mounted face to face and the substrates were glued on a heater that faces perpendicular to the both SrTi0 ₃ and "123" targets. Oxygen pressure of 40 m Torr and Argon pressure of 160 m Torr were used at the substrate temperature of 740 C.

In order to measure surface resistance at microwave frequency a parallel plate resonator was formed by sand¬ wiching two 1 cm by 1 cm "123" thin films face to face with a 12.5 thick Teflon dielectric in between. This resona¬ tor generates a series of transverse electromagnetic modes. The advantage of this method is that the current and field distribution can be calculated and the relation between the measured Q factor and the surface resistance R _s can be de- duced in a straightforward fashion. This method has been successfully used to measure 20μΩ for Nb films at 4.2 K at 10 GHz. The resolution of this method is about 5 μΩat 10 GHz.

Figure 2 shows the temperature dependence of surface resistance R _s of the "123" films grown on SrTi0 ₃ buffer lay¬ ers on Al ₂ 0 ₃ (1102) substrates. The circles denote the data on a pair of films made by laser ablation and the crosses represent the data on films grown by the off-axis sputter¬ ing technique. They have a residual resistance of about 65 μΩ at low temperature and about 800 μΩ at 77 K at 10 GHz. The actual resonance frequency (w) was about 11 GHz and the usual R„(w)∞ w ² relation was used to scale back to 10 GHz.

These surface resistance values are compared to those of Cu at 77 K and Nb at 7.7 K in Figure 3. The resistance range 65 to 200 μΩ between 10 K and 50 K is lower than the

Nb 7.7 K value at 10 GHz. The 800 μΩ at 77 K is also at least a factor 10 better than that of Cu at 77 K.

For structural information x-ray diffraction data of a c-axis scan and a ø scan of the (103) peak are shown in Figures. 4a and 4b. The c-axis scan shows that the samples are well aligned in the c-axis direction. The existence of (200) peaks means that there are some a-axis oriented grains in the sample. A SEM (Scanning Electron Microscope) image reveals that these a-axis oriented grains reside on the surface mostly as isolated grains. The rocking curve of the (005) peak is about 2.4 degree wide, which is not much different from the epitaxial "123" films on bare sap¬ phire.

The big diff rence in microstructure between epitaxial "123" thin films with SrTi0 ₃ buffer layer and films without SrTi0 ₃ buffer layer can be found in the ø-scan of (103) peak in Figure 4b. The sharp peaks every 90 degree means that the a-axis and b-axis are also aligned in the sapphire plane (1102). The full width at half maximum, φ, of the peaks in the ø-scan is about 3.8 degree. In comparison, ø of good epitaxial films on MgO, SrTi0 ₃ is about 1.5 de- gree. However, unlike the data of the "123" thin films on bare sapphire [4] , these peaks do not have the shoulders. In other words, the "123" films with a SrTi0 ₃ buffer layer have better in-plane epitaxy than those without a SrTi0 ₃ buffer layer. It is known that the I-V characteristics of the grain boundary of two misaligned grains have Josephson junction behavior. It was further found by a weakly coupled grain model that these grain boundaries lead to higher surface resistance as well as longer penetration depth. Improved in-plane epitaxy can be interpreted as reduction of grain boundaries, which results in lower surface resistance.

The mutual inductance response of a film as a function

of temperature is shown in Figures 5a and 5b. It was mea¬ sured at 1 kHz with a maximum magnetic field- of 0.6 Gauss on the film surface. The 86.5 K, where the mutual induc¬ tance transition starts, is in good agreement with the tem- perature where the dc resistivity goes to zero.

Low surface resistance 65 μΩ at 4.2 K and 800 μΩ at 77 K at 10 GHz were exhibited by the epitaxial YBa ₂ Cu ₃ 0 ₇ films on 500 A thick buffer layers of SrTi0 ₃ on A1 ₂ 0 ₃ {1102} sub¬ strates. Improved microwave surface resistance data are believed to be the results of better in-plane epitaxy due to the presence of the SrTi0 ₃ buffer layer.

Strontium titanate has some reactivity with the compo¬ nents of the growing HTSC film limiting the temperature that can be used during growth of the HTSC film to about 760°C. High quality HTSC film can be grown at higher tem¬ perature.

Though strontium titanate has a lattice spacing of about 3.92 Angstroms which is close to the approximate 3.9 Angstroms lattice spacing of YBa ₂ Cu ₃ 0 ₇ . _x , it is far from the lattice spacing of sapphire (about 3.5 Angstroms). This lattice mismatch can cause some grains to grow in undesired directions.

Strontium titanate buffers are usually deposited in a thickness of at least 100 Angstroms which increases micro- wave losses and is not conducive to the formation of the smoothest HTSC films.

Higher quality HTSC thin films are produced in accor¬ dance with the invention on calcium titanate substrates. The calcium titanate can be a bulk substrate or a thin buf- fer film or a substrate of different composition such as silicon or sapphire.

Calcium titanate has a lattice spacing of about 3.82 Angstroms which is intermediate to the lattice spacing of YBa ₂ Cu ₃ 0 ₇ . _χ (~ 3.9 Angstroms) and sapphire (~ 3.5 Angstroms). Calcium titanate films can be grown with less strain. Thinner films have less dislocations and crystal defects. Epitaxial films of calcium titanate which are 100 Angstroms

in thickness or less provide excellent substrates for grow¬ ing thin HTSC films. Thinner buffer films provide lower microwave losses.

Furthermore, since calcium titanate is less reactive with the ions forming the HTSC film, the deposition temper¬ ature can be substantially higher, e.g. about 800°C which contributes to the quality and performance of the HTSC film. Preliminary measurements indicate that the critical current density of the HTSC film are about 2 times better than comparable films grown on strontium titanate buffer films on sapphire.

Figure 6 illustrates the data from 3 different meas¬ urements of the surface resistance of 500 Angstrom thick 123 YBCO films grown on 100 Angstrom thick CaTiθ ₃ buffer layers. The surface resistivity was scaled to 10 GHz.

In a measurement of x-ray scattering, it was found that the full width at half maximum, ø, of the peaks in the ø-scan is about 3.8 degree. In comparison, ø of good epi¬ taxial films on MgO or SrTi0 ₃ is about 1.5 degree. However, unlike the data of the "123" thin films on bar sapphire [4], these peaks do not have the shoulders. In other words, the "123" films with a CaTi0 ₃ or SrTi0 ₃ buffer layer have better in-plane epitaxy than those without a CaTi0 ₃ or SrTi0 ₃ buffer layer. It is known that the I-V characteristics of the grain boundary of two misaligned grains have Josephson junction behavior. It was further found by a weakly coupled grain model that these grain boundaries lead to higher surface resistance as well as longer penetration depth. Improved in-plane epitaxy can be interpreted as reduction of grain boundaries, which results in lower surface resistance.

The mutual inductance response of a film as a function of temperature was measured at 1 kHz with a maximum magnet¬ ic field of 0.6 Gauss on the film surface. At 86.5 K, the temperature where the mutual inductance transition starts, is in good agreement with the temperature where the dc re¬ sistivity goes to zero.

Improved microwave surface resistance data and criti¬ cal current density are believed to be the results of bet¬ ter in-plane epitaxy due to the use of the CaTi0 ₃ buffer layer. It is to be realized that only preferred embodiments of the invention have been described and that numerous sub¬ stitutions, modifications and alterations are permissible without departing from the spirit and scope of the inven¬ tion as defined in the following claims.