Method for detecting particles on the surface of an object, wafer, and mask blank

Abstract

A method for detecting deposited particles (P) on a surface (11) of an object (3, 14) includes: irradiating a partial region of the surface (11) of the object (3, 14) with measurement radiation; detecting measurement radiation scattered on the irradiated partial region, and detecting particles in the partial region of the surface of the object (3, 14) based on the detected measurement radiation. In the steps of irradiating and detecting, the surface (11) of the object (3, 14) has an anti-reflective coating (13) and/or a surface structure (15) for reducing the reflectivity of the surface (11) for the measurement radiation (9), wherein the particle detection limit is lowered due to the anti-reflective coating (13) and/or the surface structure (15). Also disclosed are a wafer (3) and a mask blank for carrying out the method.

Claims

1. A method for detecting deposited particles on a surface of an object, comprising: irradiating a partial region of the surface of the object with measurement radiation; detecting scattered measurement radiation scattered at the irradiated partial region; detecting the particles on the surface of the object based on the detected scattered measurement radiation; and during said irradiating and said detecting of the scattered measurement radiation, providing the surface of the object with at least one of an anti-reflective coating and a surface structure that reduces a reflectivity of the surface for the measurement radiation, wherein the anti-reflective coating and/or the surface structure lowers a particle detection limit for the particles on the surface of the object by decreasing a full width at half maximum (FWHM) value of a haze signal for scattered light distribution due to roughness of the surface, and wherein the particle detection limit is based on an intensity threshold value that is dependent on a measure of dispersion of the haze signal.

2. The method as claimed in claim 1, wherein the object is a wafer for microlithography or a mask blank for microlithography, and wherein the particles are detected on the surface of the object.

3. The method as claimed in claim 1, wherein the measurement radiation has a predetermined measurement wavelength (λ.sub.M).

4. The method as claimed in claim 1, wherein the scattered measurement radiation is detected in a detection angle range between a first scattering angle (α.sub.1) and a second scattering angle (α.sub.2) with respect to the measurement radiation incident on the surface of the object.

5. The method as claimed in claim 4, wherein the anti-reflective coating has, in the detection angle range between the first scattering angle (α.sub.1) and the second scattering angle (α.sub.2), an angle-dependent reflectivity (R) for the measurement radiation at which a difference between a maximum value (R.sub.MAX) of the reflectivity and a minimum value (R.sub.MIN) of the reflectivity is less than 5%.

6. The method as claimed in claim 5, wherein the difference between the maximum value (R.sub.MAX) of the reflectivity and the minimum value (R.sub.MIN) of the reflectivity is less than 1%.

7. The method as claimed in claim 1, wherein the anti-reflective coating is formed as a multilayer coating.

8. The method as claimed in claim 1, wherein the reflectivity (R) of the anti-reflective coating for the measurement radiation is less than 15%.

9. The method as claimed in claim 8, wherein the reflectivity (R) of the anti-reflective coating for the measurement radiation is less than 1%.

10. The method as claimed in claim 1, wherein the surface structure is formed as a needle-type microstructure.

11. The method as claimed in claim 1, in which the object is made from silicon.

12. The method as claimed in claim 11, wherein the surface structure is formed as black silicon.

13. The method as claimed in claim 1, wherein the object is formed from an optical filter glass for filtering the measurement radiation.

14. The method as claimed in claim 1, wherein the object is made from a material that has, for the measurement radiation, an absorption coefficient of more than 1×10.sup.4 1/cm.

15. The method as claimed in claim 1, wherein the object has a thickness (d.sub.1, d.sub.2) of at least 500 μm.

16. The method as claimed in claim 1, wherein one of the particles is detected in the irradiated partial region when a scattered light intensity (I) of the scattered measurement radiation that is scattered at the partial region lies above an intensity threshold value (I.sub.S).

17. The method as claimed in claim 1, wherein at least said irradiating of the object with the measurement radiation and said detecting of the scattered measurement radiation are performed on a measurement apparatus for measuring at least one of microlithographic mask blanks and microlithographic wafers.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Exemplary embodiments are illustrated in the schematic drawing and will be explained in the following description. In the Figures:

(2) FIG. 1A shows a schematic illustration of a measurement apparatus for inspecting the surfaces of mask blanks or wafers for microlithography, in particular for detecting particles,

(3) FIG. 1B shows a plan view of a mask or of a wafer with a partial region that is irradiated with measurement radiation by the measurement apparatus,

(4) FIG. 2A shows a schematic illustration of measurement radiation that is used to irradiate the surface and is scattered at the surface into a detection angle range,

(5) FIG. 2B shows a schematic illustration of the surface to be inspected with a particle deposited on the surface,

(6) FIGS. 3A-3C show schematic illustrations of an object to be inspected, in which an anti-reflective coating is applied on the surface to be inspected (FIGS. 3A,3B) or in which the object has a surface structure for reducing the reflectivity of the surface to be inspected (FIG. 3C),

(7) FIG. 4 shows a schematic illustration of the frequency distributions of the scattered light intensity recorded during the inspection of a conventional surface or during the inspection of a surface as per FIGS. 3A-C, and

(8) FIG. 5 shows a schematic illustration of the reflectivities of the surface of FIG. 3A provided with two different anti-reflective coatings within a detection angle range of the measurement apparatus.

DETAILED DESCRIPTION

(9) In the following description of the drawings, identical reference signs are used for identical or functionally identical components.

(10) FIG. 1A shows a measurement apparatus 1 for examining a mask 2 or a wafer 3 using a scattered light measurement. The measurement apparatus 1 can be for example a measurement apparatus 1 offered by Rudolph, Inc. under the trade name “Reflex TT MBI” for examining masks 2 or a measurement apparatus 1 offered under the trade name “Reflex TT FSI” for examining wafers 3.

(11) The measurement apparatus 1 has a fixed machine body 4, on the upper side of which a measurement head 5 is mounted. The measurement head 5 is mounted on the machine body 4 rotatably about an axis of rotation 6 extending in the Z-direction of an XYZ coordinate system. A rotary plate 7 that is mounted rotatably about a further axis of rotation 8 extending in the Z-direction is also mounted on the machine body 4.

(12) The measurement head 5 has a measurement light source 10 in the form of a laser diode for producing measurement radiation 9 in the form of a measurement light beam at a measurement wavelength λ.sub.M of approximately 405 nm. The measurement radiation 9 emitted by the measurement light source 10 is deflected in the Z-direction, that is to say parallel to the two axes of rotation 6, 8, at a deflection mirror and is incident on the upper side of a mask 2 situated on the rotary plate 7 in a substantially round partial region T, which is shown in FIG. 1B. The partial region T can be moved over the surface 11 of the mask 2 by rotating the mask 2 about the further axis of rotation 8 and by rotating the measurement head 5 about its axis of rotation 6, with the result that the partial region T can be moved or displaced along the entire surface 11 of the mask 2. The measurement apparatus 1 or a suitably modified measurement apparatus 1 can also be used for examining a wafer 3, shown in FIG. 1B.

(13) As can be seen in FIG. 2A, the measurement radiation 9 that is used to irradiate the planar surface 11 of the mask 2 perpendicularly and thus in the direction of the surface normal is scattered back at the surface 11 in all directions of a hemisphere. However, a detector 12 mounted in the measurement head 5 detects only that measurement radiation 9 that is scattered at the substantially point-shaped partial region T in a detection angle range that lies between a first scattering angle α.sub.1 and a second scattering angle α.sub.2. In the example shown, the first scattering angle α.sub.1 is 32° and the second scattering angle α.sub.2 is 68°. In order to detect scattered measurement radiation 9 only in the detection angle range between the first and second scattering angles α.sub.1, α.sub.2 with the detector 12, an elliptical concave mirror and a pinhole mirror are integrated in the measurement head 5. Unlike in the case that is indicated in FIG. 2A, the detection angle range is rotationally symmetric with respect to the partial region T, that is to say the entire measurement radiation 9 that is scattered back at the partial region T in the detection angle range in the circumferential direction (that is to say over an angle of 360°) is captured by the detector 12.

(14) Since the measurement radiation 9 is radiated perpendicularly onto the surface 11, the measurement radiation 9 that is reflected at an ideal, entirely planar surface 11 would have to leave the surface 11 likewise in the perpendicular direction and therefore lie outside the detection angle range captured by the detector 12. In practice, every surface 11 has a roughness on a microscopic scale that depends on the quality of the polishing methods used, on the material from which the surface 11 is made, etc. FIG. 2B shows a detail of a real surface 11 having such microscopic roughness. Likewise shown is a particle P deposited on the surface 11 and a coating 13 that is applied in a section of the surface 11. The measurement radiation 9 that is scattered at the surface 11 and captured by the detector 12 has a first portion, brought about by scattered light due to the roughness of the surface 11, and a second portion that is due to defects on the surface 11, for example the particle P shown in FIG. 2B.

(15) The grating period G of a surface grating is related to the measurement wavelength λ.sub.M of the measurement radiation 9, the angle of incidence α.sub.E, of the measurement radiation 9 and the scattering angle as in accordance with the following equation (Bragg equation):
sin(α.sub.E)−sin(α.sub.S)=λ.sub.M/G.

(16) For normal incidence of the measurement radiation 9, that is to say for an angle of incidence α.sub.E of the measurement radiation of 0° and a measurement wavelength λ.sub.M of 408 nm, at a first scattering angle cu of approximately 32° and a second scattering angle α.sub.2 of approximately 68°, a maximum grating period G.sub.MAX of approximately 800 nm and a minimum grating period G.sub.MIN of 400 nm is obtained, that is to say the detection angle range specified above corresponds to a spatial wavelength range (or a range of grating periods G) between approximately 400 nm and approximately 800 nm. Said spatial wavelength range relates to the first portion of the scattered measurement radiation 9 brought about by scattered light due to the roughness of the surface 11.

(17) For scattering centers or particles P having a diameter D.sub.S of less than approximately half the measurement wavelength λ.sub.M, it is approximately true that the scattered light intensity I is proportional to the sixth power of the diameter, i.e. I˜D.sub.S.sup.6 (Rayleigh scattering). Particles scatter in an approximately isotropic fashion, meaning that the intensity of the second portion of the scattered measurement radiation 9 recorded by the detector 12 is only marginally dependent on the selected detection angle range. With a suitable selection of the detection angle range, more precisely of the first scattering angle cu and of the second scattering angle α.sub.2, which corresponds to a spatial wavelength range in which in an ideal manner the roughness of the surface 11 is minimum, the first portion of the scattered measurement radiation 9 that is captured by the detector 12 and is due to the roughness of the surface 11 can be minimized.

(18) Relatively large scattering centers, as are caused by particles P, which typically lie in a diameter range between approximately 0.25λ.sub.M and approximately 0.5λ.sub.M, result in a high scattered light intensity I in particular in objects such as conventional Si-wafers and masks that are used in the semiconductor field, while the roughness or unevennesses of the surfaces 11 result in a lower scattered light intensity I, as will be explained below with reference to FIG. 4.

(19) FIG. 4 shows what is known as the “defect count” (D.C.) of the scattered light intensity I, which has a first portion, referred to as the defect signal 17, and a second portion, referred to as the haze signal 18. The haze signal 18 represents a frequency distribution of the measurement radiation 9 that is scattered at the entire surface 11, that is to say of the entire measurement radiation 9 that is detected during the movement of the partial region T over the entire surface 11, which can be divided for example into a measurement grid during the measurement. The defect signal 17, by contrast, is the scattered light intensity I that was measured in a partial region T (for example corresponding to a grid element of the measurement grid). As can be seen in FIG. 4, the defect signal 17 has its maximum at a greater scattered light intensity I than the haze signal 18.

(20) The defect signal 17 is due to particles P and possibly to further defects on the surface 11, whereas the haze signal 18 is substantially due to the roughness of the surface 11, because greater defects such as particles P lead to a higher scattered light intensity I (see above). If a scattered light intensity I that lies above an intensity threshold value I.sub.S is measured in a partial region T of the surface 11, it is assumed that a particle P is present on the surface 11 in the partial region T, that is to say in the case of a scattered light intensity I that is measured by the detector 12 and lies above the intensity threshold value I.sub.S, then, with a specified irradiation duration, a particle P is detected.

(21) An addition or integration is performed over the scattered measurement radiation 9 recorded in the partial region T during the irradiation duration before an adjacent partial region T of the measurement grid into which the surface 11 is divided for the measurement is examined for particles P in the same way. If a scattered light intensity I that lies above the threshold value I.sub.S is measured in a plurality of adjacent partial regions T, it is possible using suitable algorithms to deduce the presence of a scratch or of a local surface defect on the surface 11. The threshold value I.sub.S of the scattered light intensity I in a partial region T of the surface 11 being exceeded is therefore a necessary, but possibly not a sufficient criterion for the presence of a particle P in the partial region T of the surface 11.

(22) As can be seen in FIG. 4, the haze signal 18 has a comparatively large full width at half maximum (FWHM), such that the right edge of the substantially Gaussian haze signal 18 possibly overlaps in part with the defect signal 17. Although the right edge of the haze signal 18 makes up only a comparatively small portion of the frequency distribution, it may still be the case that the scattered light intensity I lies above the intensity threshold value I.sub.S only because of the haze signal 18, which means that a particle P is detected in the partial region although no particle P is present on the surface 11 there. The intensity threshold value I.sub.S and thus the minimally detectable particle diameter D.sub.S can thus not be made arbitrarily small to avoid errors in the detection of particles P.

(23) In order to reduce the detection limit, that is to say the minimally detectable particle diameter D.sub.S, it is necessary to separate the haze signal 18 and the defect signal 17 from one another as much as possible. This can be accomplished by reducing the width of the haze signal 18, such that the intensity threshold value I.sub.S for the particle detection limit is also reduced. Although this could also be accomplished by reducing the roughness of the surface 11, this is generally not so easily possible.

(24) For reducing the detection limit, instead of using a conventional mask blank 2, a plate-shaped object 14 is introduced into the measurement apparatus 1 of FIG. 1A, on the surface of which an anti-reflective coating 13 is applied, as is shown in FIG. 3A. As can be seen in FIG. 2B, the upper side of a coating 13 applied with a homogeneous thickness follows the rough surface 11, that is to say the applied coating 13 does not increase the roughness of the surface 11. The surface 11 that is inspected using the measurement apparatus 1 of FIG. 1A thus substantially matches the original surface 11 in terms of its roughness.

(25) On account of the anti-reflective coating 13, the reflectivity and thus the scattering effect of the surface 11 for the measurement radiation 9 is reduced, that is to say a lower first portion of the intensity I.sub.O of the intensity Ii of the measurement radiation 9 that is incident at a specific angle of incidence is reflected than in the case of a surface 11 on which no anti-reflective coating 13 is applied. On account of the reduced reflectivity, the first portion of the measurement radiation 9 that is incident on the detector 12 and is due to the surface roughness of the surface 11 is thus reduced, as a result of which the scattered light distribution measured on the surface 11 changes, as can be seen in FIG. 4:

(26) The haze scattered light distribution 18′ measured on the surface 11 that is provided with the anti-reflective coating 13 is reduced in terms of its scattered light intensity I, that is to say it is displaced to the left in FIG. 4. The defect peak or the defect signal 17′ is likewise displaced to the left, but the distance A between the maximum of the haze scattered light distribution 18′ and the maximum of the defect signal 17′ remains constant. This is based on the assumption that the entire scattered light intensity I emanating from a grid element of the surface 11 is additively composed of the scattered light portion of the surface 11, that is to say the haze signal 18 or 18′, and the defect signal 17 or 17′.

(27) In the example shown in FIG. 4, it was assumed that the reflectivity of the surface 11 is halved by the anti-reflective coating 13. Therefore, the scattered light intensity at which the haze signal 18, 18′ has the maximum reduces from 100 a.u. to 50 a.u., as is clearly evident from FIG. 4. The standard deviation or the FWHM value of the haze signal 18, 18′ also changes accordingly: In the case of the surface 11 without anti-reflective coating, the FWHM value is approximately 10 a.u. (between approximately 95 a.u. and approximately 105 a.u.). In the case of a reduction of the reflectivity of the surface 11 to a half, the FWHM value is also reduced to a half, that is to say to approximately 5 a.u. (105/2 a.u.-95/2 a.u.). The intensity threshold value I.sub.S or I.sub.S′, which is typically defined on the basis of a scattering measure of the haze signal 18, 18′ (see below), is reduced accordingly. Therefore, the intensity threshold value I.sub.S′ in the case of the haze signal 18′ that is displaced to the left can likewise be displaced to the left in a direction to the maximum of the displaced haze signal 18′, without the error rate in the detection of particles P increasing here.

(28) The intensity threshold value I.sub.s or I.sub.s′ can be fixed for example in dependence on the variance, the FWHM value, the standard deviation or another measure of dispersion of the haze signal 18, 18′. For example, the intensity threshold value I.sub.s or I.sub.s′ can be defined as what is known as a 3 σ value, that is to say a scattered light intensity I.sub.s or I.sub.s′ that is more than three times the standard deviation σ of the haze signal 18, 18′—measured starting from the maximum of the haze signal 18, 18′—is considered to be associated with the defect signal 17 and is evaluated to mean the presence of a particle P. The measures of dispersion of the haze signal 18, 18′ can also be converted into one another, for example for the standard deviation σ, from which the intensity threshold value I.sub.s or I.sub.s′ is determined, and the FWHM value illustrated in FIG. 4, the following relationship applies: FWHM value≈2.3548 σ.

(29) The anti-reflective coating 13 applied on the plate-type object 14 is in the example shown in FIG. 3A a multilayer coating having a plurality of individual layers 13A, 13B with a high and low refractive index in alternation for the measurement wavelength λ.sub.M, the layer thicknesses of which are selected such that destructive interference occurs for measurement radiation 9 at the measurement wavelength λ.sub.M of 405 nm. The reflectivity R of the object 14 in the detection angle range between the first scattering angle cu and the second scattering angle α.sub.2 can here be lowered to less than 5.0%, as is shown in FIG. 5 for two anti-reflective coatings. The reflectivities R illustrated by a dash-dotted line or a dashed line correspond to two differently constructed anti-reflective coatings 13, which are applied in each case onto an object 14 made of silicon in the form of a silicon wafer. The exact layer construction of the anti-reflective coatings 13 will not be discussed in more detail here. One of the two anti-reflective coatings 13 is optimized for this detection angle range, with the result that the difference between a maximum value R.sub.MAX of the reflectivity R and a minimum value R.sub.MIN of the reflectivity R in the detection angle range lies at less than approximately 5%, that is to say for the reflectivity R in the detection angle range: R.sub.MAX-R.sub.MIN<5.0%.

(30) Unlike in the case that is illustrated in FIG. 5 by the dashed-dotted line for the reflectivity R of the anti-reflective coating 13, the minimum reflectivity R.sub.MIN is not necessarily achieved at the first scattering angle cu and the maximum reflectivity R.sub.MAX is not necessarily achieved at the second scattering angle α.sub.2. Ideally, the reflectivity R of the anti-reflective coating 13 is (approximately) constant over the entire detection angle range, as is the case for the reflectivity R of the anti-reflective coating 13 that is illustrated in FIG. 5 by way of the dashed line. It is to be understood that a corresponding anti-reflective coating 13 can also be optimized for measurement radiation 12 in the UV wavelength range, for example at a measurement wavelength λ.sub.M of for example approximately 248 nm. For the optimization of the anti-reflective coating 13, it is necessary to know the refractive index of the object 14.

(31) As an alternative to the above-described example, in which the plate-type object 14 is a silicon wafer, the plate-type object 14 can for example also be an optical filter glass, in particular a long-pass filter glass that is sold commercially by Schott under the trade name RG1000 and has a refractive index of approximately 1.54 for measurement radiation 9 at the measurement wavelength λ.sub.M of 405 nm. The object 14 in the form of the long-pass filter glass has, at a thickness d.sub.1 of 3 mm, a residual transmittance of less than 10.sup.−5, that is to say practically no measurement radiation 9 is reflected from the rear side of the object 14. On account of the anti-reflective coating 13 at the front side of the object 14, the reflectivity R of the surface 11 can be reduced by more than a factor 18 as compared to an uncoated surface 11.

(32) The thickness d.sub.1 of the object 14 and the dimensions thereof are selected such that it fits in the measurement apparatus 1 of FIG. 1A such that the object 14 can be measured therein. Instead of a wafer or an optical filter glass, the object 14 used can also be different materials, for example as are customary for mask blanks 2, for example quartz glass, titanium-doped quartz glass or a glass ceramic (in the case of mask blanks for EUV lithography). In particular, a conventional mask blank 2 can be prepared for the measurement by way of an anti-reflective coating 13 being applied thereon or by way of it possibly being microstructured, as will be described in more detail below.

(33) As was described further above, it is also possible to prepare a wafer 3 for the measurement in the measurement apparatus 1 of FIG. 1A to lower the signal-to-noise ratio or to lower the particle detection limit. FIG. 3B shows such a silicon wafer 3 on which an anti-reflective coating 13 is applied that can be constructed for example as in the thesis “Entspiegelung von Silizium-Photodioden nach dem Vorbild der Nanooptik von Mottenaugen”, D. Gabler, T U Ilmenau, 2005, cited in the introductory part, and that can consist for example of silicon nitride Si.sub.xN.sub.Y.

(34) On account of such an anti-reflective coating 13, the reflectivity R for the measurement radiation 9 at the measurement wavelength λ.sub.M of 405 nm can be lowered approximately by a factor 4 with respect to a surface 11 made from uncoated silicon. The reflectivity R of uncoated silicon in the visible wavelength range is typically more than approximately 30%, at a measurement wavelength λ.sub.M of 405 nm is approximately 50%, and the reflectivity R of the surface 11 that is provided with the anti-reflective coating 13 of FIG. 3B is less than approximately 5% (cf. FIG. 5). The absorption coefficient of crystalline silicon at the measurement wavelength λ.sub.M is more than 10.sup.5 1/cm. A commercially available wafer 3 has a thickness d.sub.2 of approximately 650 μm, such that practically no measurement radiation 9 arrives at the rear side of the wafer 3 anymore and is reflected thereby.

(35) It is to be understood that, on account of an optimization of the anti-reflective coating 13 for example by surface structuring or by the selection of a different measurement wavelength λ.sub.M, the reflectivity R of the wafer 3 or of the surface 11 for the measurement radiation 9 is further reduced and thus the detection limit for particles P in the scattered light measurement can be lowered further.

(36) FIG. 3C shows a wafer 3, in the case of which a surface structure 15 is applied for reducing the reflectivity R of the surface 11, said surface structure 15 in the example illustrated being what is known as black silicon, that is to say a needle-type microstructure that was produced by high-energy bombardment with ions or with ultrashort laser pulses. The surface structure 15 in the form of the black silicon increases the absorption of the measurement radiation 9 at the surface 11 and thus significantly reduces the reflectivity R thereof for measurement radiation 9 at the measurement wavelength λ.sub.M of 405 nm, specifically to less than approximately 2%, that is to say by a factor of approximately 20. In the case of the wafer 3 shown in FIG. 3C, the particle detection limit can therefore be significantly lowered as compared to a conventional, non-surface-structured wafer 3.

(37) Instead of a conventional wafer 3, the anti-reflective coating 13 or the surface structure 15 may also be provided on a wafer 3 that differs from a conventional wafer in that an epitaxially vapor-deposited silicon layer was applied onto the surface 11 to reduce the number of defects. Roughness measurements have shown that in the case of a wafer 3 having such an epitaxial silicon layer, the surface roughness is increased with respect to a conventional silicon wafer, as a result of which the particle detection limit increases (see above). By reducing the reflection of the surface 11 of such a modified silicon wafer 3, the particle detection limit can be lowered to an acceptable value. Even in other types of layers that are applied on a wafer 3 serving as a substrate or on a mask blank 2, the particle detection limit can be lowered in the above-described manner.

(38) It is to be understood that a combination of a surface structure 15 and an anti-reflective coating 13 that is applied onto the structured surface 11 can also be used to reduce the reflectivity R of the surface 11. To reduce the reflectivity R of the surface 11, it is also possible to use for example a grating structure or a moth-eye structure instead of a needle-type microstructure.

(39) In principle, the reflectivity R of the surface 11 can also be reduced by a gradual or possibly continuous transition of the refractive index taking place from the environment of the object 2, 3, 14, typically air, to the material of the object 2, 3, 14, that is to say a modification of the surface 11 that produces (as an effective medium) such a refractive index profile or such a transition generally leads to a reduction in the reflectivity of the surface 11. Corresponding effective media can therefore be used to reduce the reflectivity R of the surface 11 for the detection of particles P that is described further above.

(40) In summary, it is possible in the manner described further above, that is to say by using a test object 2, 3, 14 having a suitably modified surface 11 that is reduced in terms of its reflectivity R, to lower the particle detection limit such that particles P having a smaller particle diameter D.sub.S can be detected without hereby increasing the error rate.