The method of operational control of roughness of everglade surfaces of large dimensions by x-ray scanning and device for its implementation
(57) Abstract:The invention relates to the field of control of Everglade surfaces with manometer levels of roughness. The method of controlling the roughness includes direction on the analyzed surface collimated beam of x-rays at an angle slip = c-(c+0,4o), whereccritical angle, the scanning x-ray beam on the sample surface, the measurement of the intensity of the reflected radiation with the line of photodiodes and determination of the roughness on the measured parameters. Device for controlling roughness includes the x-ray source with a collimator, a scanning platform for placing the sample with the test surface and the photodiode array to measure the intensity reflected from the sample surface x-ray radiation. The x-ray source with the collimator is set relative to the sample surface with the provision of the slip angle collimated x-ray radiation =c-(c+0,4o), where ccritical angle. The invention allows to increase the area surveyed for one smarm. 2 S. and 11 C.p. f-crystals, 12 ill. The invention relates to the field of control of Everglade flat surfaces with nanometer level of roughness, such as, for example, a substrate of semiconductor microelectronic devices, CD-ROMs, laser, and x-ray mirrors, etc.Known methods of controlling surface roughness using atanasijevic microscopes and devices to measure the value of the surface roughness atomic force microscopes (e.g., ). However, the area covered by one atomic force measurement is extremely small (of order 3x3 µm2), which does not allow to control the whole sample in a reasonable time.There are also known methods of x-ray reflectometry, namely in the direction of the analyzed surface collimated beam of x-rays, measuring the intensity of reflected radiation by means of detectors and determining the amount of roughness on the measured parameters. It is also known a device for the implementation of this method, containing the x-ray source with collimator element relative displacement of the collimated source of radiation and the investigated surface is possible . Methods of x-ray reflectometry allow examination for one dimension relatively large area of 500 mm2but have a low spatial destructive ability.An object of the invention is to reduce the measurement time by increasing the area surveyed during one measurement up to 104mm2and the increase of spatial destructive abilities up to the value of the order of 1 mm.The problem is solved in that in the method of controlling surface roughness, including direction on the analyzed surface collimated beam of x-rays, measurement of the intensity of the reflected radiation with detectors and determining the amount of roughness on the measured parameters, a collimated beam of x-rays is directed to the surface under study at a glide angle =c-(c+0,4o), whereccritical angle, scan the x-ray beam on the sample surface in the direction perpendicular to the projection axis of the beam on the analyzed surface, and the measurement of the intensity of the reflected radiation to produce a line of detectors.For ensuring the rule, the line of detectors is a line back biased photodiodes with preamps and a modem to transfer data to the computer.To ensure accurate measurements and high spatial destroying the ability of x-rays colliery to divergence of 10-4the radian.In accordance with step diode line and the width of the sensing element of the detector cross-sectional dimension of the collimated beam of x-rays choose 1x1 mm2.When the scanning speed is chosen equal to 1 mm/tnwhere tnthe time of accumulation of the detector.The task is achieved by the fact that in the device for controlling surface roughness, including the x-ray source with collimator element relative displacement of the collimated source of radiation and the sample surface and the detector for measuring the intensity reflected from the sample surface x-ray radiation, the x-ray source with the collimator is set relative to the sample surface with the provision of the slip angle collimated x-ray radiation =c-(c radiation and a surface made in the form of a scanning platform for placing the sample with the test surface, detectors made in the form of line detectors.When this line of detectors is a line back biased photodiodes with preamps and a modem to transfer data to the computer.To reduce the readout noise of the photodiodes substrate on which are formed the sensitive elements of the photodiodes, is in thermal contact with the Peltier element.Typically, the device for controlling surface roughness further comprises a computer connected to the modem.To enable scanning of the scanning platform is equipped with a controller and stepper motor controlled by a computer.In private cases, the collimator is made slot or in the form of a parabolic reflector.In Fig. 1 shows the experimental data (crosses and dots) and theoretical curves according to formulas (1)-(3) (solid line) reflection coefficients of the two quartz samples with different roughness; CuKradiation
< / BR>In Fig. 2 shows the angular dependence of the signal-to-noise ratio for the silicon sample.< / BR>In Fig. 3A,b shows the optical diagram of the device.In Fig. 4 shows a diagram of a set is and without compensation and with compensation (b).In Fig. 7 shows the signal detector obtained by the illumination of the direct x-ray beam.In Fig. 8 illustrates the influence of the divergence of the beam.In Fig. 9a,b shows embodiments of the collimator.In Fig. 10 shows the structure of a first test object (a) and its image (b).In Fig. 11 shows the structure of the second test object (a) and its image (b).In Fig. 11 shows the structure of the second test object (a) and its image (b).In Fig. 12 shows the structure of the third test object (a) and its image (b).The way the roughness control is based on the following.According to the formula of Fresnel reflectance intensity flat, non-polarized monochromatic wave from a perfectly smooth surface is equal to :
< / BR>< / BR>< / BR>< / BR>Hererandi- the real and imaginary parts of the dielectric susceptibility, is the wavelength. Roughness exponentially reduces the reflection coefficient:
R(,) = R0()exp[-2f()], (2)
where is the root mean square height of roughness;
f () is a known function of the angle of slip .
The analytical form of the function f() has investigated many autorino to take [4-9]:
< / BR>Characteristic appearance of the angular dependence of the reflection coefficient shown in Fig. 1 for two values . The glide angle at which the reflection coefficient falls sharply, equal is called the critical angle. It is clear that the roughness of the sample at different points of its surface can be estimated by measuring the reflection coefficient when the slip angles, large critical. Suppose you want to distinguish between two sections of the surface roughness1and2. The greater the difference R(,1)-R(,2), the better the conditions for distinguishing between these areas. From this point of view, as shown in Fig. 1, it is necessary to choose the largest possible angle of slip. However, on the other hand, the larger the slip angle, the less the intensity of the reflected wave, and the greater the influence of the noise fluctuations of the output signal of the detector. Thus, there must exist an optimal value of the slip angle . Find this value.Let the number of photons in the incident start is equal to N. Then the difference between the amplitudes of the signals reflected from the surfaces with roughness1and2equal
< / BR>The amplitude of the random fluctuations of the output signal of the detector in a first approximation, proportional, and the average amplitude noise detector :
< / BR>You can enter the parameter q, which has the meaning of the signal-to-noise:
< / BR>The angular dependence of q() is shown in Fig. 2, which implies that the optimum glide angle is equal to approximately 0.3ofor silicon. Almost the same optimal value for quartz and aluminum. For samples with levels of roughness over the optimum value of slip angle should be maintained quite accurately, in order to obtain a sufficiently high signal-to-noise. For samples with roughness of the order of and less of this condition is not very restrictive since the large slip angles optimal, say 0,5oand-0.6ogive almost the same signal-to-noise.The principle of the method and device for control of the surface roughness is clear from Fig. 3. Parallel x-ray beam 1 of square cross-section 1x1 mm2falls on the sample 2 with the angle of slip . The size of the sample D, the glide angle and the width of the beam in the vertical plane d are related in the obvious relation d = Dsin. If the sample surface is flat, the reflected beam is parallel, so that information about each esti. Thus, if you want to have a spatial resolution in the plane of the sample, is equal to the spatial resolution of the detector must be equal = sin. For many practical applications it is enough to have a spatial resolution = 1 mm Spatial resolution of the detector is limited to step diode line 5, which in our case is equal to 12 μm, so that a value of = 1 mm corresponds to the glide angle = arcsin(/) 0,5-0,6o. This value slightly exceeds the optimum value of slip angle found above. However, for Everglade surfaces with roughness of the order of this difference can be considered negligible.The device allows you to explore the 6 samples with a maximum size of 100x100 mm2. The maximum number of resolution elements in the direction retinoscope beam is 100. To have the same number of resolution elements along the orthogonal axis, we have limited the width of the beam in the sample plane the size of 1 mm, which also corresponds to the width of one sensor element of the detector in this direction. Thus, the total number of resolution elements in the image of the object, 100x100 mm2, reanimowa platform 7 (scanner) to the controller 8 stepper motors 9-12, the detector 13, the collimator 14 with the x-ray source 15 and the computer 16 with the software. The scanner 7 is designed to align the sample 6 with respect to the incident x-ray beam and for one-dimensional scanning with respect to stationary detector 13 and the x-ray beam. Alignment is the angular orientation of the sample relative to two mutually perpendicular axes in the plane of its surface and linear movement in the vertical direction. Task alignment refers to the alignment of the sample surface under the required glide angle of 0.5oand-0.6owith respect to the incident x-ray beam and the positioning in the vertical direction of complete overlap of the beam. A necessary condition for the alignment is proper orientation of the sample around an axis parallel to the direction of scanning, as this determines the appearance of the shear distortion in the image. These three operations are performed stepper motors 14 and 16 with gearboxes. Angular step for both rotational motions is equal to 6 coal.S., and moving in the vertical plane is performed with a step of 2 μm. Scanning is also stepper motor with linear increments of 5 microns. All Chetyrkin, and full time obtain a single image, can be changed to match the accumulation time of the detector. For example, if the collection time of the detector is set equal to 1c, the linear scanning speed must be equal to 1 mm/s, which corresponds to the rotation speed of the stepper motor, 200 steps/sec. In turn, the dwell time must be chosen based on the need to provide acceptable signal-to-noise 2-10, which may vary from sample to sample depending on its roughness and material. For example, in our first experiments, the dwell time was chosen to 1c.The detector 13 is a line from 2580 back biased photodiodes with preamps and a modem to transfer data to the computer. The geometry of the sensing elements is shown schematically in Fig. 5. A silicon substrate on which are formed the sensitive elements, cooled and stabilized to a temperature of approximately 10oC using microglobuline Peltier. Cooling significantly reduces dark noise and allows you to record individual x-ray quanta . To protect sensitive components from exposure to visible light input window of the detector closed serwotka by subtracting the reference values, measured and recorded in the computer memory before starting measurements. The result of such compensation is shown in Fig. 6. Full length of the sensitive area of the detector is about 30 mm, so that only a small part of it is used for measurements. As can be seen from Fig. 7, only 150 separate delicate items from 2580 contribute to the image.There are several reasons why the incident x-ray radiation must be well collimated in the vertical plane. First of all, the slip angles at each point of the sample surface must be the same to have the same signal from different points with the same roughness. Secondly, a parallel beam does not introduce distortion in the image. In addition, the divergence of the incident beam has a negative impact on the spatial destructive ability. The scheme shown in Fig. 8 explains the situation. For example, suppose you want the resolution of the roughness of the order is not adversely affected by the divergence of the incident beam. Then from Fig. 1 it follows that for slip angles on the order of 0.5oand-0.6othere must be a limit < 0,02o-0,3o. On the other hand, that the divergence does not affect prostranstvennogo element detector. In our case, D 100 mm, and a = 0.01 mm, so that < 10-4radian 0.006o. Therefore, the spatial resolution is much more limiting factor than the resolution of the roughness. Therefore, a collimator positioned between the x-ray tube and the sample (Fig. 4), designed to reduce the divergence of up to 10-4the radian. In the first stage, we used the simplest slit collimator (Fig. 9a). Thus the divergence of the beam of rays coming within one point of the surface of the sample, equal to 0.1 mm/1000 mm=10-4radian, that is equal to the amount necessary to ensure the spatial resolution. On the other hand, the divergence of the emergent cone of rays is equal to 1 mm/1000 mm=10-3radian 0,06othat is approximately twice the value obtained above for the resolution of roughness. Therefore, when using slotted collimator is a bit distorted on the brightness of the image, and it is brighter at the detector and, consequently, the darker side of the collimator. Best image quality is obtained using the parabolic collimator shown in Fig. 9D.Below Prieur 1. The test object was prepared on a silicon substrate 100 mm by depositing a tungsten layer with a thickness in the form of parallel strips with a width of 10-15 mm (Fig. 10a). The roughness of the substrate, measured by fitting theoretical dependence of the reflection coefficient (1)-(3) to the experimental data, was equal to the Difference of the thicknesses of layers of tungsten in the adjacent bands is manifested in the difference of values of slip angles, the corresponding sharp decline in the curve of reflection: the thicker the layer, the further down the scale of corners begins a sharp decline reflection. Corners sharp drop of the reflection coefficients is bounded below critical angle of silicon 0.22oand above the critical angle of tungsten 0.56o. This led to a significant difference between the reflection coefficients of the neighboring bands, but not as big as follows from the Fresnel equations (1), since the roughness of the tungsten layer increases with its thickness.The intensity incident on the sample is collimated start was about 2105imp/s, so that the count rate of the reflected beam does not exceed 10 counts/s with an element permissions on the sample surface. The intensity noise of the detector was measured value of 0.2-0.3 IPM/s with one senses the project for not less than 1c, that provided the signal-to-noise in the image element of order 10. In order to scan the entire surface of the sample, the length of the scan was chosen equal to 110 mm, which when selected integration time, respectively, the scan time 110c.The first obtained image is shown in Fig. 10B. The scanning direction coincides with the vertical margins of the picture. It is easy to see a sample of round shape with parallel stripes on it. Dark segment in the upper right corner corresponds to the uncovered part of the plate, the reflection from which the oil compared to the coated parts. Right dark, straight, slightly oblique relative to the right margin boundary corresponds to the edge of the protective gap detector, preventing the defeat of the sensitive elements of the direct x-ray beam. The slope of this boundary is caused by a small misalignment between the scanning direction and a line slit.If the sample surface was perfectly flat, and the x-ray beam had a divergence in the vertical plane, the dark and light areas of the image belongs to the regions with large and small roughness. In General, when interpreting nabludaetsya reflection and the divergence of the incident beam. The divergence of the incident of start-up manifested in a smooth regular increase in the brightness of the picture from left to right, the more, the smaller the slip angle of rays incident on the plate surface. Variation of flatness should appear smooth, but a chaotic change in brightness due to smooth local change of the slip angle of the beam.An important feature of the presented image is a high contrast boundaries between adjacent bands. This phenomenon cannot be explained simply by changing the angle of reflection of the rays on the transition step. Indeed, the elevation at these boundaries is very small, of the order assuming that the length of the steps of the order of the spatial resolution of the device, i.e. of the order of 1 mm, the angle of reflection is just radian. Such angular deviations of the rays lie outside the scope of the sensitivity of the method. If we assume that the change of the reflection angle on the step comparable to the threshold for this method the value specified in the previous section, i.e. of the order of 10-4radians, the length of the step is equal to just more than an order of magnitude smaller than the spatial resolution of the instrument. From the above-mentioned is actionsa value. In fact, a homogeneous structure with a height of the order of a few nanometers can not be detected by conventional optical or scanning electron microscopy. Therefore, the developed instrument can be useful not only to visualize the spatial distribution of roughness, but also for visualization of regular structures with nanometer-level heights.Example 2. The test object was cooked on a flat quartz substrate 70 mm with an average roughness and the maximum variation of flatness less than 410-5radian (8 coal.C.). Six parallel to each other tungsten bands of width 6, 4, 3, 2, 1.5 and 1 mm thick were deposited on the surface of the substrate, as shown in Fig. 11a. Due to islet random nature of the deposition, the surface roughness of the strips, as expected, was significantly greater than the roughness of the substrate: about Over the created spatial structure was applied a thin layer of tungsten with a thickness of about it was Expected that the roughness of this layer will be substantially less than the roughness strips. X-ray measurements yielded a value that is less than the roughness of the substrate. As a result, the substrate was formed a regular space the second test object shown in Fig. 11b. You can see all six lanes, but the contrast of the last three most narrow of them is substantially less than the first three wide. Stripes in the picture looks dark, that is, the reflection coefficient of these sections is smaller than the rest of the sample. This proves that the nature of the observed contrast in the greater roughness of the bands, and not in their greater thickness. In fact, assume that the roughness of all sections of the same sample. Then for a fixed angle of glide reflection from a thick layer of tungsten is more than fine. Therefore, the lines would appear in the image light, i.e., the back of the observed picture.Example 3. The third test object was formed by selectively etching the quartz substrate, similar to the one that was used for the first test object. As a result, on the substrate were created in areas with a high roughness of the order shown in Fig. 12A. The image of this test object shown in Fig. 12B differs from the previous lower signal-to-noise and lower spatial resolution, since the critical angle for quartz in two times less than for tungsten.Sources of information
1. European patent N 0794406, CL G 01 and CH. 1, M: Nauka, 1972.4. M. A. Isakovich. Scattering of waves by a statistically rough surface. JETP 23, N 3(9), S. 305-314, 1952.5. L. Nevot, P. Croce. Caracterisation des surfaces par reflexion rasante de work by rayons X. Application a l etude du polissage de quelques verres silicates. Revue de Physique Appliquee 15, No.3, pp. 761-779, 1980.6. B. Vidal, P. Vincent. Metallic Board for x-rays using classical thin-film theory. Applied Optics 23, No.11, pp. 1794-1801, 1984.7. B. Pardo, L. Nevot, J-M. Andre. Matrical formalism for interfacial roughness analysis of LSMs. Proc.SPIE 984, pp. 166-172, 1988.8. F. Stanglmeyer, B. Lengeler, W. Weber, H. Gobel and M. Schuster. Determination of the dispersive correction f'(E) to the atomic form factor from x-ray reflection. Acta Cryst. A 48. pp. 626-639, 1992.9. D. K. G. de Boer. Influence of the roughness profile on the specular reflectivity of x-rays and neutrons. Physical Review B 49, No.9, pp. 5817 - 5820, 1994.10. I. P. Dolbnya, S., Kurylo. The absolute spectral sensitivity of the photodiode lines with the x-ray range 7-12 Kev. Preprint BINP SB RAS, Novosibirsk, 1991. 1. The method of controlling surface roughness, including direction on the analyzed surface collimated beam of x-rays, measurement of the intensity of the reflected radiation with detectors and determining the amount of roughness on the measured parameters, wherein the collimated beam of x-ray radiation direction is scanned by the x-ray beam on the sample surface in the direction perpendicular to the projection axis of the beam on the analyzed surface, and the measurement of the intensity of the reflected radiation to produce a line of detectors,
2. The method according to p. 1, characterized in that = 0,5-0,6o.
3. The method according to p. 1 or 2, characterized in that the line of detectors is a line back biased photodiodes with preamps and a modem to transfer data to the computer.4. The method according to any of paragraphs.1 to 3, characterized in that the x-ray radiation colliery to divergence of 10-4the radian.5. The method according to any of paragraphs.1 to 4, characterized in that the cross-sectional dimension of the beam of collimated x-ray radiation 1 x 1 mm2.6. The method according to p. 5, characterized in that the scanning speed is chosen equal to 1 mm/tnwhere tnthe time of accumulation of the detector.7. Device for controlling surface roughness, including the x-ray source with collimator element relative displacement of the collimated source of radiation and the sample surface and the detector for measuring the intensity reflected from the sample surface x-ray radiation, wherein the source of inthenews the Oia collimated x-ray radiation =c(c+0,4o), whereccritical angle, the element of the relative movement of the source of collimated radiation and the sample surface is made in the form of a scanning platform for placing the sample with the test surface, and the detectors are made in the form of a line of detectors.8. The device according to p. 7, characterized in that the line of detectors is a line back biased photodiodes with preamps and a modem to transfer data to the computer.9. The device under item 8, characterized in that the substrate on which are formed the sensitive elements of the photodiodes, is in thermal contact with the Peltier element.10. The device under item 8, characterized in that it further comprises a computer connected to the modem.11. Device according to any one of paragraphs.7 to 10, characterized in that the scanning platform is equipped with a controller and stepper motor controlled by a computer.12. Device according to any one of paragraphs.7 to 11, characterized in that the collimator is made slot.13. Device according to any one of paragraphs.7 to 11, characterized in that the collimator is made in the form of a parabolic reflector.
SUBSTANCE: method includes recording number of particles emitted by radioactive layer on basis of number of voltage or current pulses recorded by counting device, then to measuring detector a flow of ionizing radiation is directed from calibrating standard electrode and also registered is number of particles, position of covered electrode is change no less than two times, by turning it in horizontal plane around its axis for arbitrary angle, while repeating measurement of pulses number, while measurement time is selected to be such that number of recorded pulses was no less than 3600 pulses for each measurement position, and then selection of necessary number of electrodes is calculated for forming electrode system in chamber.
EFFECT: higher precision, higher safety.
FIELD: MEASUREMENT TECHNOLOGY.
SUBSTANCE: electromagnet wave is induced by means of directed aerial. The wave is incident to dielectric plate. Brusterain angle of incident wave is defined from minimum value of reflected wave and value of dielectric permeability is calculated. Power of incident and reflected waves are measured and the value of reflectivity and specific conductivity are calculated as well as value of dielectric loss of dielectric plate. Then incident angle of electromagnet wave is increased till achieving value providing total internal reflection of electromagnet wave and attenuation of intensity is measured at normal plane relatively direction of wave propagation. Factors of normal attenuation and thickness of dielectric plate are calculated. Method allows to find complex dielectric permeability and thickness of dielectric plates free of dielectric substrates.
EFFECT: improved reliability.
FIELD: electrical measurements.
SUBSTANCE: device is proposed for measurement of dielectric and magnetic permeability as well as thickness of spin coatings on surface of metal and can be used in chemical industry for inspecting composition and properties of liquid and solid media. Electro-magnetic field is induced in body of dielectric material to be inspected which material is applied onto dielectric substrate, by means of sequent excitation of slow surface waves: two E-waves are excited at different, but having almost the same value, wavelengths λr1 and λr2 and one H-wave having wavelength of λr3. Attenuation of field intensity is measured t normal plane in relation to direction of wave propagation by means of receiving vibrators system for different values of base d between them. Normal attenuation factors αE1,αE2 and αH are found from ratio of E(y)= E0 exp[-α(y) y]. Magnetic and dielectric permeability and thickness of magneto-dielectric coating are found from relations of and where has to be phase coefficient of H-wave.
EFFECT: improved precision of measurement.
FIELD: measuring engineering.
SUBSTANCE: meter determines dielectric permittivity and thickness of the oil layer by measuring at two angles unequal to the Brewster angle.
EFFECT: simplified design and expanded functional capabilities.
FIELD: radiometric testing.
SUBSTANCE: counting of electric pulses of all the detectors stops simultaneously as soon as any detector registers no less than specified number of electric pulses caused by ionizing radiation.
EFFECT: improved reliability.
2 cl, 1 dwg