WO1993013240A1

WO1993013240A1 - System for depositing superconducting films

Info

Publication number: WO1993013240A1
Application number: PCT/US1991/009698
Authority: WO
Inventors: John H. Prince
Original assignee: Prince John H
Priority date: 1991-12-23
Filing date: 1991-12-23
Publication date: 1993-07-08
Also published as: AU1337392A

Abstract

A system for depositing a film containing as many as seven elements or more, the ratios amongst which are under precise computer cotrol. The ability to change these ratios almost instantly allows the deposition of structurally and compositionally complex crystalline or amorphous films, including superconducting films. In a preferred form the system comprises a high vacuum chamber (2) comprising a multi-pocket (6a-6g), single filament electron gun (24) and an atomic absorption rate monitoring system (20, 22). The atomic absorption system (20, 22) measures the concentrations of the depositing elements in the vapor near the film substrate (68), and frequently relays these data to the computer (14). These data are used by the computer (14) to precisely control the deposition rates of the elements being deposited. This deposition control is implemented by precisely varying the dwell times of the electron beam (1) within the pockets (6a-6g) containing the depositing elements.

Description

SYSTEM FOR DEPOSΓΠNG SUPERCONDUCTING FILMS

Background of Invention

Field of the invention:

The present invention relates to systems for forming crystalline films that act as superconductors. (As used herein, the term "film" refers to a thin film of thickness 50 A to 200,000 A or more, deposited onto thick ceramic, metallic or organic substrates.) More particularly, the present invention relates to systems that employ vapor deposition techniques for building crystalline superconducting films.

State of the art:

It is well known that high temperature superconducting materials can be formed by vapor deposition techniques. Typically, the superconducting materials are composed of hundreds - or many thousands - of discrete periods of precise atomic composition. In order to provide desired properties in the superconducting materials, A the vapor deposition rate of the various constituents, and therefore the ratio among the rates, must be precisely controlled to arrive at the right composition. Currently in this field atomic ratios can be controlled within one percent.

SUMMARY OF THE INVENTION

The present invention relates to systems for controlling vapor deposition processes for forming high-temperature superconducting thin films. More particularly, the present invention provides a system for controlling the ratio, rate and sequence with which materials are vaporized to form superconducting films.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention can be further understood with reference to the following description in conjunction with the appended drawings, wherein like elements are provided with the same reference numerals. In the drawings:

Figure 1 is a pictorial view, schematic in nature, of a system according to the present invention for producing superconducting films; 6

Figure 2 is an enlarged pictorial view of an electron- beam guidance system for use in the system of Figure 1; ^r»

Figure 3 is a plan view which schematically shows an alternative embodiment of an electron-beam guidance system for use 5 in the system of Figure 1;

Figures 4A and 4B are timing diagrams which show two sequences of clock periods;

Figure 5 shows a coordinate system for describing the 10 coordinates of the crucibles of Figure 1;

Figure 6 is a timing diagram which illustrates a sequence of dwell times versus x-y coordinates;

Figure 7 is an amplified view of Figure 6 showing the fine 15 structure of the y coordinates during dwell time n2 in pocket 2;

Figure 8 is a flow chart which illustrates the operation of the system of Figure 1; and Figure 9 is a pictorial view, schematic in nature, of internal mechanisms of the system according to the present invention for producing superconducting films.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

In the system of Figure 1, a set of seven open crucibles 6a-6g is arranged within a vacuum chamber 2. Each of the open crucibles holds a material for vaporization by electron beam 1. In the drawing, vapors rising from one of the crucibles are schematically illustrated by the upwardly flaring cone. In the operation of the system, there would be several such- vapor cones.

The system in Figure 1 further includes a microprocessor- based computer 14 that operates a guidance system for determining the landing location of the electron beam 1. The computer 14 also controls the dwell time of the electron beam at each landing location.

In practice, the dwell times can range from about five milliseconds to more than five-hundred milliseconds with a resolution of 3.2 us or better.

As further shown in Figure 1, the computer 14 receives signals from sensor systems that monitor conditions within the vacuum chamber 2. The monitored conditions can include, for example, the evaporation rates of the heated materials, the growth rate of the vapor deposited films, and the chemical composition of the vapor deposited films. In practice, these conditions are monitored in "real time" - that is, while vapors are being deposited to form superconducting films.

The system shown in Figure 1 controls all the process variables necessary for producing films accurately. In addition to the sensor systems there are, for example, systems for holding and heating the substrates. The substrates may be crystalline ceramics, amorphous ceramics, or metals. The substrates may be mounted on a block which may be heated through the block by resistance means or heated directly by radiation from a closely mounted quartz lamp. The heater blocks may rotate, as in Figure 9 item 68, or be stationary, as in item 69. The heating system controls the ramp up of temperature, the constant temperature, and the cooldown at rates typically +- IC/min. The thickness of the film may be controlled either by the optical monitor 70 or by the integration of rate times time by the computer 14. In either case the precision will be of the order of +- 1%. Still further in the system in Figure 2, electromagnets 34, 36, and 38 generate magnetic fields that determine both the landing location and crucible-to-crucible movement of the electron beam 1. The field strengths of all three electromagnets 34, 36, and 38 are determined by the computer 14. In the particularly illustrated embodiment, the electromagnets 34 and 36 are mounted in parallel for acting on opposite sides of the electron beam 1 for controlling its lateral (y-direction) landing location, and the electromagnet 38 is mounted perpendicular to the magnets 34 and 36 for controlling the longitudinal (x-direction) landing location of the electron beam.

In practice, the coils for the electromagnets 34, 36, and 38 are high-speed, low inductance coils with ferrite cores. Also in practice, the coils for the electromagnets are immersed for cooling in a high thermal conductivity, low inductance alloy block. The block can be made, for example, of a 70:30 coppeπnickel alloy or any other mechanically durable, low vapor pressure material with high electrical resistance and good thermal conductivity.

As an alternative to the arrangement of electromagnets shown in Figure 2, Figure 3 shows two electromagnets 42 and 43 that are mounted at an oblique angle to the direction of the electron beam 1

-1. The oblique angle can range from 180 to less than 90 degrees, and the number of electromagnets can be two or four or more. In operation, the electromagnets 42 and 43 act in concert to control both the lateral and longitudinal components of the beam landing location.

In operation of the system of Figure 1, the landing location of the electron beam 1 and its dwell time at each landing location determines the composition and rate at which vapors are generated from the crucibles 6a - 6g. In practice, the computer- controlled system provides the selection and control of thousands of different evaporation combinations and sequences, so that different superconductors and combinations of different superconductors can be deposited singly and in any sequence. This sequence also can include layers other than superconductors with different required properties.

In addition, for a given dwell time, the system of Figure

1 controls the electron beam 1 either to remain at a fixed location within a given crucible or to move from location to location within the crucible. This control of the beam landing location is tantamount to enlarging the beam spot and, therefore, allows precise control over heating density within selected one of the crucibles. This feature is f

important because the elective size of the beam spot can either promote, or inhibit, dissociation of certain materials. Also, beam movement within a crucible increases the uniformity with which a crucible's contents are heated, allowing materials in the crucible to be uniformly vaporized.

The beam guidance system in Figure 1, as will be further described below, provides control during each evaporation cycle, or set of cycles, so that the landing location of the electron beam within any given crudble depends on the cyde. This feature also promotes uniform heating and evaporation of materiak from the cruάbles.

Employing the coordinate system described in Figure 5, a sequence of beam positions and associated dwell times is shown in Figure 6. In this example the electron beam dwells on crudble 6a (coordinates xl,yl) for forty clock periods; then on crudble 6b

(coordinates x2,y2) for fifty clock periods; and so forth. Finally, the electron beam dwells on crudble 6g (coordinates x2,y7) for sixty clock periods. Then the beam returns to renew the cycle at crudble 1. Figure 7 shows an example of the above-described control system being used to vary the y-direction landing location of the electron beam within a given dwell period. In this example, dwell period n2, in which the electron beam dwells on crudble 6b, is shown forillustration. This dwell period n2 is expanded in Figure 7 to show the fine detail. Both the x and y coordinates will vary but for this illustration only the y coordinate is shown. The computer 14 controls an x-y driver board which steps the coils with dwell time n2 at a rate of the order of one thousand hertz.

These fast changes are possible because the steps are very small. Thus during the dwell period n2 the y-direction position of the beam can be changed a number of times , thereby allowing the beam size to be programmed for each material to be vaporized. In the example a fairly smooth function of these changes is shown but the changes can assume any form within the limits of the x-y coordinates of the crudble.

Figure 7 also shows examples of the y-direction landing location of the beam being varied from the first cycle n2 to the second cyde n2 and so forth. In this example, the same function is repeated but its location is slightly changed. As noted above this feature promotes unifrom heating and evaporation of certain l^ό

-materials over a number of cycles.

Figures 4A and 4B show examples of two sequences of dod periods. In Figure 4A, the dwell times nl and n7 are each fifty periods long and, within each dwell time, the beam coordinates are fixed. These dwell times can be adjusted by adding or subtracting periods based on information from the sensors that monitor process conditions. So, for instance, Figure 4B shows the dwell time nl shortened by ten dodc periods and dwell time n2 lengthened by ten dodc periods. Thus, that drawing illustrates the general case of varying dwell times in successive landing locations.

To extend the preceding example, when the beam energy is fixed, the energy deposited in each crudble is directly related to the proportion of time that the beam spends in each crudble. So, when ten dock periods are subtracted from dwell period nl and ten dock periods are added to dwell period n2 as in

Figures 4A and 4B, the energy ratio will go from 50:50 to 40:60.

As the following illustration will show, as a practical

matter the energy can be adjusted in exceedingly fine amounts. For example, with a dodc period of 3.2 us and a 32 ms dwell period

(which is typical) the number of clock periods will be 10,000. Thus, l l

subtraction of ten clock periods from nl and addition of those clock periods to n2 would change the ratio from 1:1 to 9,990:10,010. Because computers can just as easily subtract (and add) one clock period the ratio could as easily be 9,999:10,001. This is by no means the practical limit to resolution, because a clock period can be subtracted (or added) every tenth cyde, giving a ratio of 9,999.9:10,000.1.

In addition, by operating the internal clock generator of the computer to run at twice as fast, for example, the clock periods will be half as long. As a further example, with the internal clock generator running at ten times the speed, the clock periods will be one tenth as long. These measures increase the resolution of the dwell times by a factor of twenty over that described above.

The very large resolving abilities inherent in the above- described techniques imply that with a fast feedback loop (such as obtains with the atomic absorption method coupled with the computer) the accuracy of the film composition, on an instantaneous basis, will be within one part in ten thousand. On a continuing basis, because of fast updates, the accuracy will be of the same order.

In order to secure this level of accuracy two further components need to work together with a high degree of perfection. The first is the sensor system, which indudes the atomic absorbtion method; and the second is the evaporation control algorithm within the computer, which are both described below.

Evaporation control methods for the above-described system will now be described. In this particular embodiment the following algorithm and its derivatives are used for controlling the ratios between the film materials. As described in more detail later, atomic absorption for each material in the film is used to individually control the evaporation rate of that material.

In practice, each atomic absorption channel uses a hollow cathode Lamp with an emission line for that particular material. This emission line is split into a main beam, which passes right through the material vapor, and a reference beam, which is split off at the hollow cathode lamp, for use for reference purposes only.

In the illustrated embodiment, the main beams 20, 22 are directed through a central point just under the substrate. This provides substantially optimum control of the rate for all materials because the rate at the substrate will be most nearly equal to the rate 12,

just below it. The direction of the beams through a singular central point is an advantage because all path lengths through the vapors are equal and, therefore , substantially equal amounts of the different component materials are simultaneously monitored. In addition, with path lengths equal, beams can be interchanged without the re- calibration needed when the main beams go through different paths. This is an advantage when changing from experiment to another.

The light intensity from both the main and reference beams is measured with photomultiplier tubes that are equipped with optical narrow band-pass filters where transmitted bands are of the same wavelength as the emission lines. In practice, the signals are filtered with digital low-pass filters that are programmable from the computer 14. The filtered signals are calibrated and used to obtain a ratio of the intensity of the attenuated main beam to the intensity of the main beam before evaporation. This ratio, which is always less than one, measures the absorption of the spectral line by the evaporated material in the path of the main beam.

This above-described ratio is used by the computer 14 for calculating the actual atomic flux rate which, for present purposes, is defined as the rate of deposition of a given material (in

Angstroms per second) onto a fixed substrate. In practice, the atomic . 4

flux rate can be related to the above-discussed absorption ratio in a non-linear manner by an experimentally derived working curve. This working curve is found by performing an experiment comparing the readings on a quartz crystal microbalance with the atomic absorption readings while the rate of evaporation of the particular material varies. The whole operation is performed and the data is recorded by the computer so that the working curve can be calculated and stored in retrievable mathematical form for use while making depositions.

The above-discussed actual atomic flux rate is subtracted from the desired atomic flux rate for each material, resulting in an error term. The error term comprises the argument for subsequent calculations by the computer 14, with the result being output signals that minimize the magnitude of the error term. These calculations can be made as follows:

For each crudble (1, . n, . ,7):

lastError(n) = Error(n) (Save the previous Error for later)

Error(n) = Desired Atomic Flux Rate(n) - Actual Atomic Flux Rate(n) This error value is provided to a control loop which calculates the electron beam power which is needed for each particular crudble according to the formulas below:

prop(part) = propFartor(n) * Error(n)

integPart(n) = integPart(n) + integFactor(n) * Error(n)

diffPart(n) = diffFactor(n) * ( Error(n) - lastError(n) )

beamPower(n) = gainFartor(n) * ( propPart(n) + integPart(n) + diffPart(n) )

In practice, the computer 14 measures and calculates all channels at once. When all of the powers of all of the beams are added to one another, the sum equals the total beam power:

totalBeamPower = BeamPower(l) + beamPower(2) + ...

The dwell time for each crudble is determined by taking the proportion of time that crudble uses out of the Total Beam Power, multiplying by 100 and converting to an integer. This /&

number, called the dwell count, is in units of 3.2 us.

dwellCount(n) = beamPower(n) * 100

After the adjustments are made, the process is repeated, within a period of one tenth of a second to one second.

Several modifications can be made to the above- described process for increasing its reliability. For instance, the above-desάrbed algorithm can be simplified so that the dwell time assodated with the material most out of balance is adjusted in the first cyde and then, in sequential cydes, the dwell times for the material next most out of balance is adjusted, and so on, with the objective of bringing the ratios into balance. Because the balance will never be exact, this is a continuing, iterative process for updating the dwell times for each material.

Second, because the process is non-linear, and because the working curve is experimentally known, the control algorithm is made to work on the curve, or in fact on piece-wise linear approximations of it. Also the rate of power change, or slew rate, is adjusted in a non-linear manner to provide damping. Together with high and low limits to the power a high degree of control is provided for the evaporation of each material.

In addition to these modifications for in- process control, an important feature has been added for control after the evaporation is completed, for use in future experiments. This feature is data logging, in which all process variables are recorded in a form in which they can be later retrieved for review and editing. It has been found that because of the sensitivity of certain process variables small variations can make a significant difference to the results, and the data log provides a means to allow close scrutiny of the variations. After review, the variables can be modified to change process parameters so that a subsequent run can be made to conform better to existing or new requirements.

The rate of evaporation depends uniquely for each material on the amount of energy deposited on it. As described above this energy can be changed with a high resolution within each cycle by changing the number of clock periods ( Dwell Count) the electron beam is made to dwell on that material. However, because the sensing method (described later) can determine predsely the rate of evaporation for each material, and the total evaporation is the sum, this sum can be fed back on an instant basis to control the total It

-energy deposited.

Hence not only can the ratios be determined and controlled instantly but by the same mechanism so can the total rate, and hence the results are precisely deposited films.

The prinάples, preferred embodiments, and modes of operation of the present invention have been described; however, the invention is not limited to the particular embodiments discussed, which are illustrative rather than restrictive. For example, the above- described system can be used to make amorphous films as well as crystalline films; to make optical films; to make magnetic alloy films; and so on. Also a system can be devised in which there can be than one electron-beam gun, each having one or more pockets, which are simultaneously controlled by the above-described methods. Again, a system can be devised in which the method of vaporizing the materials may (for example) be a laser instead on an electron beam, and in which mirrors instead of magnets control the landing locations; control can again obtain through the methods described above. Variations may be made in the above-described embodiments without departing from the scope of the present invention as defined in the following daims.

Claims

1 -1WHAT IS CLAIMED IS:

1. A system for controlling the vaporization of materials that can be vapor deposited for forming super-conducting films, comprising: vacuum chamber means in which superconducting films can be grown by vapor deposition techniques; open crudble means mounted in the vacuum chamber means for holding materials to be vaporized; electron beam means for generating an electron beam for vaporizing the materials held in the open crudble means; and beam guidance means for controlling a) the locations at which the electron beam lands on materials in the crudbles, b) the dwell time of the electron beam at each landing location, and c) the sequence of landing locations.

2. A system according to Claim 1 further including sensor means for measuring the rate at which each of the materials held in the crudbles vaporize.

3. A system according to Claim 1 further including computing means for calculating the time period that the electron beam should dwell on each crudble for each cycle and each sequence of the materials to be vaporized.

4. A system according to Claim 1 wherein the beam guidance means selects the dwell time at each landing location within the range of about five (5) milliseconds to about five-hundred (500) milliseconds, with a resolution greater than about 3 us.

5. A system according to Claim 3 wherein the computing means provides updating information for the beam guidance means each cyde.

6. A system according to Claim 1 wherein the beam guidance means indudes a first pair of electromagnets that are mounted in parallel for acting on opposite sides of the electron beam for controlling its lateral landing location.

7. A system according to Claim 6 wherein the beam guidance means further indudes at least one electromagnet which is mounted perpendicular to the first pair of electromagnets for controlling the longitudinal landing location of the electron beam. Ά\

8. A system according to Claim 6 wherein the coils for the electromagnets are high-speed, low indurtance coils and wherein the system indudes cooling means for immersing the coils in liquid for cooling.

9. A system according to Claim 7 wherein the cooling means further includes a high thermal conductivity, low inductance alloy block

10. A system according to Claim 3 wherein the computing means controls the field strengths and activation periods of the electromagnets.

11. A system according to Claim 10 wherein the computing means controls the electromagnets so that the electron beam sweeps within each crudble.

12. A system according to Claim 10 wherein the computing means controls the electromagnets so that the electron beam walks in discrete steps within each crudble.

13. A system according to claim 1 wherein the beam guidance means indudes two or more electromagnets that are mounted at an oblique angle to the direction of the electron beam to act in concert for controlling both the lateral and longitudinal components of the landing location of the electron beam.

14. A system according to claim 1 wherein the beam guidance means includes a computing means for controlling the electron beam to move firom location to location within a crudble for any given dwell period.

15. A system according to Claim 1 wherein the beam guidance means includes means for providing control during each cycle so that the landing location of the electron beam within any give crudble varies depending upon the cycle.

16. A system according to Claim 1 wherein the beam guidance means includes means for providing control of the x- and y direction of the electron beam to sweep within a crudble within a give dwell period.

17. A system according to Claim 2 wherein the sensor means is atomic absorption device and is combined in a multiplidty AS

of channels such that the beams are of equal length and intersect at a common point.

18. A system as in Claim 3 wherein the computing means includes means for data logging of past runs and for reediting

for future runs.