US3703637A

US3703637A - Electron beam inspection

Info

Publication number: US3703637A
Application number: US91654A
Authority: US
Inventors: Allan Edward Dugan
Original assignee: Western Electric Co Inc
Current assignee: AT&T Corp
Priority date: 1970-11-23
Filing date: 1970-11-23
Publication date: 1972-11-21
Anticipated expiration: 1989-11-21

Abstract

Differences between the secondary electron emission characteristics of two phases of the same material in response to bombardment by an electron beam are sensed to enable impurities, comprising one of the material''s phases, in a sample, comprising the material''s phase, to be detected. The electron beam is a low energy beam. Relative motion between the sample and the beam may be effected by sweeping the beam, by moving the sample, or by a combination of the two motions.

Description

United States Patent Dugan 1 Nov. 21, 1972 [54] ELECTRON BEAM INSPECTION 3,473,023 10/1969 Bloch ..250/49.5 PE [72] Inventor: gs z ifag yggfi m g Primary Examiner-James W. Lawrence a Assistant Examiner-C. E. Church [73] Assignee: Western Electric Company, lncor- Attorney-W. M. Kain, R. P. Miller and R. C. Winter porated, New York, NY. 22 Filed: Nov. 23, 1970 [57] ABSTRACT Differences between the secondary electron emission [211 Appl' 91654 characteristics of two phases of the same material in response to bombardment by an electron beam are [52] user. ..2s0/49.s PE,250/49.5 A sensed to enable impurities, comprising one of the [51] Int. Cl. ..G0ln 23/04 materials P a sample ""P materi- [58] Field of Search ..250/49.5 PE 49.5 A Phase be detected- The elem beam is a energy beam. Relative motion between the sample and [56] Reerences Cited the beam may be effected by sweeping the beam, by moving the sample, or by a combination of the two UNITED STATES PATENTS motions.

3,479,506 11/1969 Dorfler ..250/49.5 PE 4 Claims, 10 Drawing Figures I05 OO 144 U llk 105 J K I06 i V [lo l,

I38 I34 l30 I34 |5o Y-MAGNIFJCIATION Y-RAMP GEN. J- DUAL i AM? c J. X-MASNIFICATION T X-RAME GEN. AMP.

I36] I32 I28} I32 I4I/ \MO (202 |oz'- l.

l I 7+ no l4 4 I I 1 1 1 [em Qt \T 31 5:1127. x- MAGNIFICATION x- RAMP 'GEN. AM P 200 1 I52 I28 1 I40 I m PATENTEDunvz: I972 SHEET 2 BF 4 4205 ORI6) VACUUM lo l llllllllllllllllllll PRIMARY ELECTRON ENERGY (KeV) PRIMARY ELECTRON ENERGY (Kev) PATENTEDuuvzumz 3.703.637

7 SHEEI30F4 lo 1 l L l L; .2 .4 .5 L L 2.0 g PRIMARY ELECTRON ENERGY (Kev) EP 11:

.- z w m g 0 .1 g; 5

nr 0 u w .J o O l l l l 0 IO 7 --4O GRID BIAS (VOLTS) m m z o m 5 5 n: 0.5 7 :7 g: 7 i1 5 w Lu SECONDARY ELECTRON ENERGY(eV) SECONDARY 4* PRIMARY 1 ELECTRON BEAM msrscrron BACKGROUNDOF THE INVENTlON 1. Field of the invention The present invention relates to the detection of impurities in a sampleand, more particularly, to a method of electrically detecting, in a sample, the presence of impurities, the sample being a first phase of a material while the impurities are a second phase of the same material.

In a more specific sense, the present invention contemplates inspecting a sample of a first phase of material to detect the presence therein of a second phase of the same material by using techniques derived from scanning electron microscopy (SEM).

2. Discussion of the Prior Art The use of SEM techniques to study both surface topography and the boundary between a first material (e.g., silicon) and a second material (e.g., gold), deposited on the first, is well known. In typical SEM techniques a sharply focused beam of high energy elec' trons, usually in the range of 3 keV-40keV, is moved relatively with respect to a sample, e.g., by sweeping the beam thereover in a raster. impingement of the beam on the sample effects one or more of several types of beam-induced signals. These signals may be utilized to form an image by means of a sweep on a cathode ray tube (synchronized with the beam sweep), modulated by the beam-induced signals.

The beam-induced signals may be generated by emitted secondary electrons, emitted back-scattered electrons, emitted auger electrons, emitted photons, or may be proportional to various types of currents induced in the sample, all of these being due to impingement of the beam on the sample. It has been generally preferred to form the image on a cathode ray tube with signals generated by emitted secondary electrons, back-scattered electrons, or induced currents.

Notwithstanding the rather highly developed state of the SEM art, as summarized above, there have been no known techniques for accurately and reproducibly dc tecting or observing the presence of a second phase of a material in a sample of a first phase of the same material. It is not known why such a deficiency in the SEM art exists. Nevertheless, one object of the present invention is to enable such detection to be effected.

The detection of sites of the second material phase, herein termed impurities, in the first material phase, herein termed the sample, is especially important, when the phases are different phases of tantalum. This is true when the sample is B tantalum (tetragonal) and the impurities are body centered cubic (bcc) tantalum, hereinafter called a tantalum. There is also importance in the detection of B tantalum impurities in a tantalum samples.

Today, sputtered B tantalum is considered a preferred material in the production of thin-film capacitors. Its electrical properties, especially its dielectric properties when anodized, have led to this preferred status. See US. Pat. Nos. 3,391,373; 3,382,053; and 3,275,915. Generally, such thin-film capacitors comprise a sputtered film or thin-film of B tantalum on a substrate, a portion of the film being anodized to form a B tantalum pentoxide dielectric layer. This layer is ultimately covered with a metallic film (often gold) which serves as a capacitor countere- 2 lectrode. Thus, the completed capacitor comprises, in order, a substrate, a B tantalum film, a B tantalum pentoxide layer and a counterelectrode.

Sputtered B tantalum films have been found to contain, at times, a tantalum inpurities, which impurities undesirably affect the otherwise desirable electrical characteristics of the B tantalum. Specifically, when B tantalum films containing a tantalum impurities are anodized to produce the a tantalum pentoxide capacitor dielectric layer, the anodized a tantalum sites result in a tantalum pentoxide sites which have a different (usually lower) dielectric constant than the B tantalum pentoxide layer. Further, the a tantalum pentoxide sites often (about percent of the time), for unexplained reasons, result in leaky capacitors and/or capacitors which break down at their intended voltage.

The presence of such a tantalum impurities is thought to be caused by improperly adjusted parameters of the sputtering procedures by which B tantalum films are usually produced. The exact nature of, and the interaction between, these parameters is not presently understood, however. Accordingly, at times these parameters can be adjusted to eliminate such a tantalum impurities; however, at other times such elimination proves difficult or impossible. Nevertheless, the detection of the a tantalum impurities prior to anodization of the B tantalum film at least prevents additional processing (e.g., anodization, counterelectrode deposition, etc.) of a B tantalum film which cannot be used to produce an acceptable capacitor.

Thus, the accurate, simple and expedient detection of a tantalum impurity sites in B tantalum thin films is a sought-after end, and is an object of the present invention.

B tantalumfilms for capacitors are usually produced by a so-called in-line sputtering machine in which a conveyor-like track continuously transports substrates through one or more sputtering environments and low pressure chambers which serveas air-locks. See US. Pat. No. 3,340,176 (assigned to the assignee of the present invention) for a disclosure of one type of such an in-line sputtering machine. Continuous inspection of the films as they leave the machine is a desirable end and an object of the present invention. Another object of the present invention is to utilize improved SEM techniques to realize this end.

Specifically, as noted before, SEM techniques utilize relative motion between the sample and the electron beam. Either the sample or the beam or both may be moved, although it is more typical to sweep the beam in a raster. Because the sputtered B tantalum films leave the in-line machine continuously on the track (in an arbitrarily chosen Y-direction of motion) the electron beam can be simply swept (in a X-direction) thereover. The signals produced by such scanning, with proper coordination of the scan rate and the track speed, can be used to adjust the sputtering parameters, or at least to pinpoint B tantalum films with a tantalum impurities therein. Such is also an object of the present invention.

it should be noted that considerations similar to the above, also obtain when a tantalum films are being produced. At times B tantalum impurities reside therein, and detection of such impurities is yet another object of this invention.

Moreover, the detection of impurities (i.e., one phase of a given material, not necessarily tantalum) in a sample (i.e., another phase of the same given material, again, not necessarily tantalum) is a general object of this invention.

SUMMARY OF THE INVENTION With the above and other objects in view, the present invention contemplates a new and improved method of electrically detecting impurities in a sample of a predetermined material. The impurities comprise one phase of the material and the sample comprises another phase of the same material. For example, the sample may be in the form of a sputtered thin film of 6 tantalum and the impurities may be a tantalum sites. The composition of the sample and the impurities may be reversed, (i.e., the sample may be a tantalum and the impurities may be B tantalum.

A low energy electron beam is generated, the electrons therein having energies within the range of 0.1-3.0 keV. Preferably, however, the energies of the majority of the beam electrons are at approximately 0.4 keV. The electron beam is swept across the sample, either in a raster scan or as a result of the combined motion of the beam and of the sample. Sweeping of the beam effects the emission from the sample of, inter alia, secondary electrons and back-scattered electrons. The emitted secondary electrons are captured and a current is derived proportional to the number of secondary electrons so captured.

It has been found that different phases of the same material, whether tantalum or not, emit substantially different numbers of secondary electrons (but about the same number of back-scattered electrons) upon the low energy electron beam impinging thereon. Accordingly, differences in the number of secondary electrons captured yields a direct indication of the presence of the second phase impurities in the sample of the first material phase.

Accordingly, in synchronism with the relative motion between the beam and the sample, variations in the derived current are measured. Such measurement indicates changes in the number of secondary electrons captured which changes, in turn, are indicative of the number of secondary electrons emitted.

It has also been found that different phases of the same material emit, upon low energy electron bombardment, substantially the same number of back-scattered electrons. Accordingly, one of two alternative techniques may be used in carrying out the above method. First, the capturing step may be effected so that almost all of the back-scattered electrons are excluded. Such exclusion is based on the fact that the geometry of the emitted secondary and back-scattered electron configurations renders it possible to capture either one or the other or both types of emitted electrons. Second, where more convenient, the capturing step may effect the capture of both the secondary and the back-scattered electrons. In this instance, because as noted above, different phases of the same material emit substantially the same number of back-scattered electrons, the derived current may have subtracted therefrom a constant current proportional to the number of back-scattered electrons. In either case, the ultimate derived current is proportional to the number of secondary electrons and pennits the convenient detection of the presence of impurities in the sample.

BRIEF DESCRIPTION OF THE DRAWINGS Other objects and advantages of the present invention will appear upon a consideration of the following detailed description with the accompanying drawings wherein:

FIG. la is a schematic diagram of low energy scanning electron beam inspection apparatus, used in carrying out the method of the present invention, to detect impurities in samples;

FIG. lb is an alternative embodiment of FIG. la wherein the inspection apparatus thereof is used in conjunction with an in-line sputtering machine;

FIG. 2 is a schematic diagram of apparatus used to measure the secondary and back-scattered electron emission properties of a typical material comprising a sample which is to be inspected and of a typical material comprising an impurity in the sample which is to be detected by the apparatus of FIGS. la or lb;

FIG. 3 is a graph showing the back-scattered electron emission coefficient 1 versus the incident electron beam energy obtained with the apparatus of FIG. 2 for the material comprising the sample to be inspected by the apparatus of FIGS. la or lb and for the impurities in the sample;'

FIG. 4 is a graph depicting the secondary electron emission coefficient 8 versus the electron beam energy obtained with the apparatus of FIG. 2 for the material comprising the sample to be inspected by the apparatus of FIGS. la or lb and for the impurities therein;

FIG. 5 is a graph of the ration of 8 to 1 for both the material comprising the sample and the impurities therein versus the electron beam energy obtained by combining the graphs of FIGS. 3 and 4 and using average values for 8 and 1 therefrom;

FIG. 6 is a graph depicting, in the apparatus of FIG. 2, the current through an emitted electron collector as a function of the bias on a retarding grid when a 1.5 keV electron beam impinges normally on the materials comprising the sample and the impurities therein;

FIG. 7 is a graph depicting the ratio 6/eV versus the electron energy of secondary electrons emitted from both materials comprising the sample and the impurities therein upon bombardment thereof by a 1.5 keV electron beam nonnal thereto; this graph was obtained by differentiation of the graph of FIG. 6;

FIG. 8 is a graph depicting the variation of 8 for the material of the sample and of the impurities therein as a function of the angle of incidence of a 1.5 keV electron beam with respect thereto wherein 0 is normal incidence;

FIG. 9 is an enlarged view of a thin film of B tantalum having a tantalum impurities therein, these two phases of the same material being typical of the materials of the sample and impurities, respectively, to be inspected by the apparatus of FIGS. la or lb.

DETAILED DESCRIPTION Before discussing the method of the present invention and the apparatus for effecting that method, it is necessary to discuss some preliminary considerations. These preliminary considerations concern certain characteristics of a sample 14 to be inspected and of impurity sites 15 therein to be detected.

The sample 14 to be inspected by the present method may be in any form. Referring to FIG. 9, there is an exemplary sample 14 comprising a thin film 16 deposited, for example, by sputtering on a glass or other appropriate substrate 17,. The thin film 16 contains the impurity sites 15, the size of which is greatly exaggerated in FIG. 9. Of particular interest in the present invention are impurities 15 which are both a different phase of the material of the film l6 and undetectable by ordinary visual techniques, as well as impurities which are visually detectable. Exemplary materials are [3 tantalumfor the film 16 (ultimately used as a capacitor electrode) and a tantalum for the impurities 15 (which may prevent the film 16 from being properly anodized to form the capacitor). it is to be noted that different phases of the same material, whether tantalum or some other material, have approximately the same atomic number Z and that the impurities 15 of interest here are those which may be undetectable by ordinary visual means.

When an electron (hereinafter called a primary electron) of energy E, is incident upon a target, a number of different types of electrons (and photons, xrays, etc.) are emitted therefrom. This is true whether the target is the material of the film 16 of the sample 14 or is the material of the impurities 17 in the film 16. Two types of emitted electrons of interest here are back-scattered electrons and secondary electrons, for reasons to be more fully discussed below. For purposes of the present invention, if the film 16 is B tantalum and the impurities 15 are or tantalum, emitted electrons having energies greater than 40eV are considered back-scattered electrons while those electrons having energies less than 40eV are considered secondary electrons. This limit is somewhat arbitrary and will vary according to the material of the film 16 (one phase) and of the impurities 15 the other phase). The choice of this limit is discussed below with reference to H6. 6. Nonetheless, in accordance with standard definitions, secondary electrons are considered to be of lower energies than back-scattered electrons.

Back-scattered electrons originate near or at a surface of the target upon which the primary electron of energy E, is incident. Many back-scattered electrons are incident electrons which undergo nearly elastic reflection from the first few atomic layers of the target surface. It has been found that the number of backscattered electrons emitted from the surface for a given incident primary electron energy is, for the most part, some function of the atomic number Z of the material of thesurface. Moreover, different phases of the same material have substantially the same atomic number Z. Accordingly, if a surface upon which an electron is incident contains impurities of a different phase of the same material which constitutes the surface, substantially the same number of back-scattered electrons will be emitted from both the surface and the impurities therein, as shown in FIG. 3, discussed below.

Secondary electrons are emitted from a surface when an incident electron of energy E, penetrates the surface and is scattered within the material constituting the surface. Such scattering transfers energy from the incident electron to other electrons within the material. Many of these electrons to which energy is transferred migrate to the surface and are emitted from the surface if they are energetic enough to penetrate the potential barrier of the surface. It has been found that the number of secondary electrons emitted is dependent upon, among other things, parameters such as the average depth beneath the surface at which the secondary electrons are generated, the condition of the surface and the work function of the material of the surface. Assuming these factors to be generally equal, it has been found that different phases of the same material emit, upon being bombarded by low energy electron beam, substantially different numbers of secondary electrons, as more fully discussed below with reference to FIG. 4.

For purposes of this invention, a secondary electron emission coefficient 8 isdefined as the ratio of the number of secondary electrons emitted to the number of primary electrons incident upon a surface. Similarly, a back-scattered electron coefficient 'r is defined as the ratio of the number of back-scattered electrons emitted from the surface to the number of primary electrons incident thereupon. The incident electrons are referred to-as primary electrons. Referring now to FIG. 2, there is shown apparatus for performing certain preliminary steps of the method of the instant invention. These preliminary steps involve an initial characterization of the materials comprising the sample 14, specifically, the material of the film l6 and'of the impurities 15 therein which are a different phase of the same material. For convenience {3 tantalum and a tantalum are chosen for such materials, but similar characterizations may be easily effected for different phases of materials other than tantalum.

Within a chamber21l evacuable to a low pressure by standard facilities, not shown, there is contained a target holder 22, a standard electron gun 24, and an emitted electron collector system 26. The electron gun 24 is so positioned that a beam of low energy electrons 28 may be directed at the target holder 22 at varying energies and angles of incidence. The electron gun 24 is shown generally and may include standard facilities such as a cathode, an anode, electronic optics, deflection plates, etc., not shown.

The target holder 22 may comprise a conductive member 30 rotatable on an axis 32 thereof by means, not shown. The member 30 may take any convenient shape, a generally square cross-section being shown in FIG. 2. Mounted to one face A of the member 30 is a faraday cup 36 which is a very efficient electron collector, as is well known. Mounted to another face B of the member 31) is a target 41). The target 40 includes, depending upon the characterizations being effected, either a deposit 42 of the essentially pure material of the film 16 of the sample 14 to be inspected or of the material of the impurities 15 to be detected by inspection of the sample 14, the impurities 15 being a different phase of the same material as the sample film 16.

The purity of the target 4% is ascertained by standard methods.

The deposit 42 may reside on a substrate 17, which may be the same as the substrate 17 of FIG. 9. Moreover, the deposit 42 may conveniently be, but not necessarily, a thin film. Specifically, if the sample 14 includes a sputtered thin film (of B or a tantalum) so too should be the deposit 42 of the target 40; conversely, if the sample 14 includes an electrodeposited thick film, the deposit 42 of the target 40 should be such a thick film; etc. Moreover, the deposit 42 is electrically connected to the member 30 via a convenient conductor such as at 44.

Selective rotation of the member 30 about the axis 32 positions one of the two faces A or B thereof to have incident thereon the electron beam 28 from the electron gun 24. The member 30 is grounded, as shown, in series with a first electrometer 46 or other current measuring device.

The electron collector system 26 includes a hemispherical, conductive dome 48 with its open, concave side facing the target holder 22. Mounted through the dome on a radius thereof is a standard drift tube 50 through which the electron beam 28 passes in traveling to the target holder 22. The drift tube 50 is well known in the art, and merely prevents the electron beam 28 from being adversely affected by varying potentials on the various elements of the electron collector system 26.

Mounted within the concave side of the dome 48 and insulatively spaced therefrom is a suppressor grid 52.

The dome 48 is grounded in series with a first variable voltage source 54 and a second electrometer 56 or other current sensor. The suppressor grid 52 is connected to a movable arm 58 of a double pole switch 60. When the arm 58 engages a first contact 61 of the switch 60, the suppressor grid 52 is grounded in parallel with the dome 48. When the arm 58 engages a second contact 62 of the switch 60 the grid 52 is grounded in series with a second variable voltage source 64 and a third electrometer 66 or other current sensor.

The variable voltage source 54 has its positive side connected to the dome 48 and the contact 61, while the variable voltage source 64 has its negative side connected to the contact 62.

In general, if a grounded object, such as the deposit 42 of the target 40, is bombarded by an electron beam 28 of energy E, a relationship exists between the beam current i,,, the current flowing between the object and ground i the current due to back-scattered electrons i and the current due to secondary electrons i That relationship is:

It can be seen that measurement of any three of these currents permits a direct determination of n and 8. Specifically, this is true because the number of primary electrons is proportional to i,,, n is proportional to i,,, and 8 is proportional to i In operating the apparatus of FIG. 2, the member 30 of the target holder 22 is rotated on the axis 32 to first position the faraday cup 36 to receive the electron beam 28. The arm 58 is moved to engage the contact 61 and the dome 48 and the grid 52 are rendered positive (attractive of electrons), for example, by applying thereto a potential of about 120 volts from the source 54. Thus, should any secondary or back-scattered electrons be emitted by, or escape, the faraday cup 36 they are attracted to and captured by the dome and

grid

48 and 52, as will be indicated by current flow through the electrometer 56. The electron gun 24 is turned on. The current i, to ground through the electrometer 46 for the faraday cup 36 (i, i is now measured as is the current (i +1, through the electrometer 56.

It is found, as should be expected, that very few electrons are emitted by or escape the faraday cup 36. Notwithstanding the high positive potential on the dome 48 and on the grid 52, the electrometer 56 indicates substantially zero current flow to ground. Accordingly, when the faraday cup 36 is bombarded by the electron beam 28,

i.,+i,=0;thus,' (2a) This value for i, (=i is used subsequently in determining the value of 8 and 1 Specifically,

n=i.,/i,,=i,/i (2d) Next, the member 30 is rotated to position the deposit 42 of the target 40 so that the electron beam 28 impinges thereon. As noted previously, the deposit 42 may include either a substantially pure thin film of the material of the film 16 (B tantalum) or of the material of the impurities 15 (a tantalum). Bombardment of the deposit 42 which comprises the film l6 material by the electron beam 28 results in the flow of a current through the electrometer 56, the arm 58 remaining in engagement with the contact 61 of the switch 60 to apply the positive potential to both the dome 48 and the grid 52. This current, denoted i, and due to electrons captured by the positively biased dome 48 and grid 52, contains components due to both secondary and back-scattered electrons. Specifically, i, is defined as ic=ii+ii. (a) the positive bias (here, volts) being sufficiently high to effect capture of substantially all the emitted secondary and back-scattered electrons. This bias due to the source 54, may, of course, be varied depending on the material of the deposit 42, the only requirement being that substantially all of the emitted secondary and back-scattered electrons are captured.

Next, the arm 58 is moved to engage the contact 62. Initially the source 64 is adjusted to produce zero (0) volts output. Then the source 64 is adjusted to render the grid 52 increasingly more negative while the dome 48 remains biased positively. Depending upon the material of the deposit 42, the voltage source 64 is now adjusted until all of the lower energy secondary electrons are repelled from the electron collector system 26 by the negative bias on the grid 52. The higher energy back-scattered electrons get through the negatively biased grid 52 to the positively biased dome 48.

At the grid bias where all or substantially all secondary electrons are repelled, both

electrometers

56 and 66 indicate current flow denoted i, and i respectively, due to the more energetic back-scattered electrons reaching the dome 48 and the grid 52, respectively.

After the above current measurements have been made, it is a relatively simple matter to obtain the values of 6 and 1 as follows:

Because the current i is due solely to back-scattered electrons, i is defined as Thus, from equation (3),

9 i i, i and from equations (4) and (5 i, ='i, (i +i,"). 6

Dividing equations (4) and (6) by 1, (previously defined as i, i,,) in accordance with equations (2c) an (2d),

Note that it here is the actual i, measured depending on the character of the target 40, while the i, (i, i used previously is that present when the faraday cup 36 receives the beam 28 (in the latter i is used only to ob tain a measure of i,,).

Accordingly, from equation (ll) measurement of i, (by a meter, a CRT, etc.) gives an indication of i,, albeit a mirror measurement.

Referring now to FIG. 3, obtained by the abovedescribed use of the apparatus of FIG. 2, there is shown a graph representing the varying back-scattered electron emission coefficient 1 versus the electron energy li of the electron beam 28 for both the material of the film 16 and the material of the impurities 15. This graph contains information relating to B tantalum, which is an example of the material of the film 16, and a tantalum which is an example of the material of impurities 15. Because the impurities are merely a different phase of the material of the sample 15, as may be seen from FIG. 3, the value of 'n (as well as i, i, i,") for varying primary electron energies E, from about 0.2-2.0 keV for the two is substantially the same. Accordingly, as previously discussed, the use of backscattered electrons to detect the presence of the a tantalum impurities 17 in the B tantalum sample 15 is exceedingly difficult. Again, this difficulty is due to the fact that the atomic numbers Z of both phases of tantalum are substantially the same. For materials other than tantalum, this similarity of 'n is experienced over an E, range ofO. l-3.0 keV.

Referring now to FIG. 4, also obtained via the abovedescribed use of the apparatus of FIG. 2, there is shown a graph similar toFIG. 3 except that the value of 1; for both tantalum phases is plotted against the energy E, of the electrons in the beam 28. As can be seen from the figure, for electron energies between 0.2-2.0 keV, the value of 8 for a and B tantalum is significantly different. The difference in the respective values of 6 is a maximum at incident electron energies at about 0.4keV and is a minimum at energies greater than 2.0keV. For non-tantalum materials the E, range of 0.1-3.0 permits easy difi'erentiation of 8 for the two phases. The maxima and minima 8 differences are at a different E,,, which can be easily calculated as discussed below.

In fact, referring to FIG. 5 (a combination of FIGS. 3 and 4) which is a plot of the ratio of 8 to 1 (both averaged for both materials) versus the energy of the electrons in the beam 28, it may be seen that, at about 0.4keV, 8 is nine times greater than n. Thus, it may be said that with the same amount of noise present in measuring either 8 or n, the signal-to-noise ratio of the 8 measurement is'nine times greater than the signal-tonoise ratio of the '1 measurement, at beam electron energies of 0.4keV. .At other beam electron energies, the signal-to-noise ratio, while less than 9, is about 3 or greater.

Referring now to FIG. 6, there is shown a graph of (a) the current (i, i through the

electrometers

56 and 66 when the arm 53 engages the contact 62 versus (b) the bias on the grid 52 due to adjustment of the variable voltage source 6 3 for both a and B tantalum. There is, as can be seen, a measurable difference between the values of the collector current curves for the two materials up to about 40 volts bias on the grid 52.. The assumption is made here that this difference is due to secondary electron emission differences between the two materials, their back-scattered electron differences, as shown by FIG. 3, being negligible and both varying substantially linearly. The fact that at a negative 40 volt bias the curves merge leads to the conclusion that, at least for purposes of a and h tantalum, secondary electrons are those electrons having energies less than about 40eV. Such a definition of secondary electrons is accordingly used herein with reference to a and I8 tantalum. Specifically:

Secondary electrons energies less than 40eV Back-scattered electrons energies greater than In fact, referring to FIG. 7, tests show that the energy distribution of the secondary electrons emitted from a and 3 tantalum is one wherein most of these electrons have energies within the approximate range O-ISeV, with the majority of these being at about 4eV.

From materials other than a and B tantalum, variations in the graphs of FIGS. 3-7 will exist but the result is the same. Where the materials are different phases of the same material, a graph similar to MG. 3, using the apparatus of FIG. 2, shows that 11 for the two materials is about the sameand varies linearly for'low, varying E,s. Moreover, 8 for the same 18, will be measurably different for the two phases, similar to FIG. 4. Also, a graph similar to FIG. 5 will show significant signal-tonoise ratio improvements of 8 over 1 Moreover, use of the apparatus of FIG. 2 produces graphs similar to FIGS. 6 and 7, indicating that at low primary electron energies E,, an easy differentiation between secondary and backscattered electrons can be made.

It should be noted that FIGS. 3-7 result from the bombardment of a and )3 tantalum deposits 42 in the target 40 with an electron beam 28 of 1.5keV. It has been found that substantially the same curves, with typical shifts of only fl percent are generated when the electron beam 28 energy E, is within the preferred range of 0.2-2.0 keV. Similar minimal shifts are observed for materials other than tantalum.

FIG. 8 shows the variation in 6 for a and B tantalum with a 1.5keV electron beam 28 as the angle of incidence of the beam 28 with respect to the target 40 is varied. Note that represents a beam 28 perpendicular to the target 40. The curves for the two materials indicate that the angular dependence of 8 for both tantalum phases is similar. Both 8s increase about 40 percent as the incident angle increases from 0-40. Tests indicate that the shape of these curves are substantially independent of the energy of the beam 28 over the range 0.22.0 keV. Similar results are obtained with non-tantalum materials.

Thus, it is worthy of note that:

' a. at low (0.22.0 keV) E,, 17 for two different phases of the same material is practically indistinguishable (FIG. 3); the same is substantially true at E, greater than 2.0keV;

b. at low (0.2-2.0keV) E,, 8 for the two phases is different and the difierence is easily detectable (FIG. 4); at E, greater than 2.0, 8 differences are practically indistinguishable;

c. at low (0.2-20 keV) E,, 8/1; is 3 to 9, indicating 8 is a better indicator of detection with a given amount of noise present (FIG. 5);

d. a sharp line of practical demarcation exists at low (02-20 keV) and E, may be easily defined between secondary and back-scattered electrons (FIGS. 6 and 7); and

e. statements (a) (d), above, are independent of the angle of incidence of an electron beam of low (0.22.0 keV) E,.

Referring now to FIG. 1, there is shown one embodiment 100 of the low energy scanning electron beam inspection system of the present invention. This system 100 resembles a scanning electron microscope in that it uses a scanning electron beam as a probe.

An electron beam generator 102 comprises the usual electron source 103, grid 104, anode 105 and first and second pairs of

deflection plates

106 and 107, respectively. The deflection plates 106 control the sweep of the beam 110 in X direction; the plates 107 control the Y sweep. An electron beam 110 generated by this arrangement 102 is directed at and impinges on the sample 14.

Knowing that the secondary electron emission coefficient 8 of the material of the film 16 and of the impurities therein (a different phase of the same material) are quite different at low primary electron energies E, permits detection of these impurities by a detector subsystem.

The detector sub-system includes a secondary electron collector 122, which comprises a conventional scintillator 123, a light pipe 124 and photomultiplier 125 which produces, at an output 126 thereof, a current proportional to the number of secondary electrons emitted by the bombarded sample 15. The scintillator 123 may be positioned so that it receives only secondary electrons and very few back-scattered electrons, as is known.

Another arrangement, not shown, would be to position the scintillator 123 to receive both secondary and back-scattered electrons and to make an adjustment in the current of the output 126 to eliminate the component (i.e., i," i,") of that current due solely to the back-scattered electrons. As noted previously, use of the apparatus of FIG. 2 permits calculation of i," i," for any E,. Accordingly, the constant i, i," i," may be continuously subtracted from the current output 126 of the photomultiplier. Because i, is the same for both phases of the material in the sample 14 at low E,, only i, differences are impressed on the dual amplifier 150.

X-ramp and Y-

ramp generators

128 and 130, respectively, have their

outputs

132 and 134, respectively, connected to an X-magnification control 136 and a Y- magnification control 138. The X-magnification control 136 is connected to the deflection plates 106 to control the X motion component of the raster scan of the electron beam on the sample 14. The Y-magnification control 138 is connected to the deflection plates 107 to permit control of the Y motion component of the raster scan of the electron beam 110.

Outputs 132' and 134', respectively, of the X- and Y-

ramp generators

128 and 130, respectively, are connected to an amplifier section 140. The amplifier section 140 includes an amplifier 141 coupled to the output of the X-rarnp generator 128. The output 142 of the amplifier 141 is connected to deflection plates 144 in a standard cathode ray tube 146 for controlling the X scan thereof. Thus, the signal applied to the plates 144 is the same as that applied to the plates 106 (divided, of course, by any magnification factor supplied by the control 136).

The Y scan of the cathode ray tube 146 is controlled by deflection plates 148. The deflection plates 148 are controlled by the output 149 of a dual amplifier 1.50 in the amplifier section 140.

The dual amplifier 150 receives the output (on 134) of the Y-ramp generator 1.30 and the output 126 of the photo-multiplier 125. The time-varying output of the Y-ramp generator 130 is amplitude-modulated by the output of the photo-multiplier 126 in the dual amplifier 150. Accordingly, the signal applied to the deflection plates 148 is an amplitude-modulated signal which at"- fects the Y component of the raster scan of the cathode ray tube 146 proportionally to the number of secondary electrons emitted by the sample M. Of course, the Y-ramp component applied to both sets of plates 107, M8 is the same, except for, in addition to modulation, any magnification supplied by the control 130.

Due to the differences in the secondary electron emission characteristics of B and a tantalum, as previously described, as the electron beam 110 scans across the film 16 and the impurities 15 in that film 16 areas of a and B tantalum produce visibly different outputs on the cathode ray tube 146.

Suitable results are also obtained if an alternative embodiment 200 is used in conjunction with an inline sputtering machine 201. Such use is most convenient, because the output end 202 of the in-line machine 201 is maintained in an evacuated condition. Accordingly, the electron beam generator 102 may be placed physically in one of the evacuated chambers 203 near the output end 202 of the in-line machine 201.

Because the samples 14 move continuously through the in-line machine, a single set of deflection plates, for example, the X plates 106, may be used. Specifically, the X plates may be used simply to scan the beam 110 back and forth (with blanking, if desired) across the samples 14 as they move, due to the movement of a conveyor-like track 204'through the chamber 203, Y scanning being provided by the motion of the samples 14. An appropriate sensor 204 may be used to sense the speed of the track 204 (and of the samples 14) to provide a Y signal to the dual amplifier 150. This Y signal is modulated by the output 126 of the photomultiplier 125 to effect deflection of the beam in the CRT 146 via deflection plates 148. Such sensor 204, accordingly, may replace the

elements

107, 138, 13 0, and 130 and 134' in FIG. 1. Of course, Y scanning may be provided as in FIG. 1.

It should be obvious to one skilled in the art that the CRT 146 is not an essential element of the present invention. Having a signal available at the output 126 of the photomultiplier 125 proportional to 8 permits the use of other expedients such as meters, etc., to detect the impurities 15. In fact, the output 126 may be conveniently connected to a computer facility, programmed to make judgements concerning the acceptability of the samples 14 based on the number of impurity sites l-present in the film 16.

In any event, the output 126 may be used either to:

a. adjust the sputtering parameters of the inline machine 201 to eliminate the unwanted impurities 15, or, where such elimination is impossible because of unknown factors,

b. detect and prevent the further processing of films 16 having such impurities.

It is to be understood that the above-described embodiments are simply illustrative of the principles of the invention. Various other modifications and changes may be devised by those skilled in the art which will embody the principles of the invention and fall within the spirit and scope thereof.

Whatis claimed is:

1. A method of detecting, in a sample of a first phase of a material, the presence of a second, different phase of the material, said first and second phases having the same atomic number Z, which comprises the steps of:

impinging said sample with a low energy electron beam, the electrons in said beam having energies falling within the approximate range 0.2-2.0 keV for effecting the emission from said sample of secondary electrons, substantially all of which have energies within the range 0-15 eV, and backscattered electrons in approximately equal numbers from both phases of said material;

capturing said secondary electrons to the substantial exclusion of said back-scattered electrons to thereby derive a current proportional to the number of said emitted secondary electrons; and relatively moving said beam and the sample and simultaneously detecting, in synchronization with said relative motion, variations in said derived current f0 indicating changes in the number of said emitte secondary electrons to detect the presence of the second material phase in the sample.

2. The method of claim 1 wherein said material is tantalum and said first and second phases are respectively the a and )3 phases thereof.

3. The method of claim 1 wherein said material is tantalum and said first and second phases are respectively the B and 0: phases thereof.

4. The method of claim 1 wherein the energy of the majority of the electrons in said beam is approximately 0.4 keV.

L-566-PT Patent No. Dated November 21,

Inventor(s) Allan Edward Dugan It is Certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

r In the specification, Column 3, line 18, (i.e. should 1 read --i.e. Column 9, line 1, the equation should read i i i -5 line 29, "i is a constant" should read "1, is a constant--; line 36, i i K K K should '-j t K3|'"''o Signed and sealed this 3rd day of April 1973.

(SEAL) Attest:

EDWARD M.FLETCHER,JR. ROBERT GOTTSCHALK Attesting Officer Commissioner of Patents

Claims

1. A method of detecting, in a sample of a first phase of a material, the presence of a second, different phase of the material, said first and second phases having the same atomic number Z, which comprises the steps of: impinging said sample with a low energy electron beam, the electrons in said beam having energies falling within the approximate range 0.2-2.0 keV for effecting the emission from said sample of secondary electrons, substantially all of which have energies within the range 0-15 eV, and back-scattered electrons in approximately equal numbers from both phases of said material; capturing said secondary electrons to the substantial exclusion of said back-scattered electrons to thereby derive a current proportional to the number of said emitted secondary electrons; and relatively moving said beam and the sample and simultaneously detecting, in synchronization with said relative motion, variations in said derived current for indicating changes in the number of said emitted secondary electrons to detect the presence of the second material phase in the sample.

2. The method of claim 1 wherein said material is tantalum and said first and second phases are respectively the Alpha and Beta phases thereof.

3. The method of claim 1 wherein said material is tantalum and said first and second phases are respectively the Beta and Alpha phases thereof.