Download the book "Technology of production of semiconductor devices and integrated circuits" (3.82Mb). Method a
Introduction
Chapter 1. Analytical review of the literature 11
1.1. General characteristics of the ion implantation method 11
1.1.1 Theory of ranges and distribution of ions in solids 11
1.1.2. Formation of radiation defects during ion implantation 15
1.1.3. Formation of distribution profiles of radiation defects
1.2. Effect of Proton Implantation Parameters on the Microstructure, Distribution Profile, Mechanical and Electrical Properties of Silicon
1.2.1. Influence of proton energy 27
1.2.2. Proton Dose Effect 29
1.2.3. Influence of post-implantation annealing 33
1.2.4. Influence of different orientation of substrates 38
1.3. Application of proton implantation in the technology of manufacturing semiconductor devices
1.4. Conclusion on Chapter 1 45
Chapter 2. Methods for studying the structure of damaged layers 46
2.1. X-ray diffraction method 47
2.1.1. Double-crystal X-ray diffractometry method 48
2.1.2. Three-crystal X-ray diffraction method 51
2.1.2.1. Diffuse X-ray Scattering by Microdefects in Single Crystals
2.1.2.2. TRD Intensity Profiles in the Case of Single Crystals with Coulomb Type Defects
2.2. X-ray topography method 64
2.3. Transmission electron microscopy method 66
2.4. Conclusion on Chapter 2 67
Chapter 3. Objects of research and methods of experiments and measurements 68
3.1. Silicon crystals implanted with different energies and doses of hydrogen ions after implantation and annealing
3.2. Irradiation technique 69
3.3. Construction of the Distribution Profiles of Embedded Hydrogen and Radiation Defects in Silicon Using the TRIM Program
3.4. Resistivity Profile Measurement Technique 72
3.5. X-ray technique 73
3.5.1. Identification of a damaged layer using X-ray topography
3.5.2. Study of the structure of ion-implanted layers by X-ray diffractometry
3.5.3 Method for determining the integral characteristics of the damaged layer
3.5.4. Method for Obtaining Strain Profiles from Diffraction Reflection Curves
3.5.5. Determination of Microdefect Parameters from the Results of Measuring the Intensity of Diffuse X-Ray Scattering
3.6. Sample preparation technique for examination by transmission electron microscopy
3.7. Opinion on Chapter 3 89
Chapter 4 Results of a comprehensive study of ion-implanted layers and their discussion
4.1. Results of studying the effect of irradiation on the structural properties of silicon
4.1.1. Results of studying the influence of the dose and temperature of proton irradiation on the integral characteristics of the damaged layer
4.1.2. Results of studying the influence of the field of mechanical stresses on the formation of a damaged layer upon implantation of hydrogen ions into silicon
4.1.3. The results of the study of the effect of post-implantation heat treatment on the process of defect formation
4.1.4. Results of Determination of Parameters and Qualitative Changes in the Nature of Microdefects in Proton-Implanted Layers of Silicon Crystals
4.1.5. Analysis of Changes in the Characteristics of Microdefects in Silicon Crystals Irradiated with Protons during Heat Treatment
4.2. Results of a Study on the Possibility of Using Proton Implantation to Correct the Characteristics of PIP Photodiodes
4.3. Opinion on Chapter 4 158
The main results and conclusions of the dissertation 160
Literature 163
Introduction to work
In recent decades, the possibilities of traditional semiconductor metallurgy have been significantly expanded through the use of ion implantation technology, which makes it possible to introduce practically any impurities into the material with concentrations not limited by the solubility limit. At present, the main regularities of the processes occurring with this doping method have been studied quite well, its advantages and possibilities of use have been identified, and ion implantation itself has become one of the basic technological processes.
For a long time, practically the only application of ion implantation, both in planar and non-planar technologies, was the introduction of dopants into semiconductors in the production of discrete devices and integrated circuits. In recent years, the field of application of ion implantation has expanded significantly.
The relevance of the work is due to the rapidly developing areas of practical use of ion implantation in modern technology for the production of microelectronic devices. In particular, for the controlled introduction of radiation disturbances for the purpose of isolating the elements of integrated circuits, accelerating diffusion and stimulating the electrical activity of embedded impurity atoms, precision adjustment and the creation of high-resistance resistors, obtaining amorphized layers, gettering unwanted impurities, controlling the properties of a metal-semiconductor contact, etc. . But behind the desire to obtain the optimal parameters of an ion-implanted material ready for the production of a device based on it, the study of the processes occurring in a crystal during ion implantation, which is also important for microelectronic technologies, remains in the shade. With the lack of a clear understanding of the mechanism of structural changes in
Near-surface defect-saturated layers of ion-implanted structures are associated with most of the technological problems.
Studies carried out in recent years have shown the promise of irradiating silicon with light ions (hydrogen, helium) for the formation of so-called "disturbed" layers and regions. A feature of such layers is the existence in them of a large number of structural defects created by the implantation of hydrogen ions and subsequent heat treatment. Depending on the implantation modes, temperature and annealing time, it is possible to create regions saturated with various types of defects: clusters and accumulations of point defects, micropores, gas bubbles filled with hydrogen. The study of the nature and characteristics of defects arising during implantation will expand the possibilities of the ion implantation method in the field of creating new technologies and controlling the characteristics of microelectronic devices. Information about the structure of ion-implanted samples will make it possible to decide on the choice of doses and energies of ion implantation, as well as the annealing temperature in order to optimize the properties of the damaged layer for a specific practical application. All this makes this work relevant.
The purpose of this work was to identify the main regularities in the formation of the defective structure of silicon layers implanted with hydrogen ions under various conditions and its evolution during heat treatment, as well as to study the possibility of using proton implantation to modify the properties of silicon surface layers in order to improve the parameters and increase the yield of suitable electronic devices.
To achieve these goals, it was necessary to solve the following main goals :
one). Determine the dependence of the nature of defect formation on the conditions of implantation (energy and dose);
2). Find out the role of external factors (temperature, mechanical stress fields) in the formation of a defective structure during implantation;
3). Determine the structural and electrophysical characteristics of silicon layers disturbed by proton implantation and trace their changes during subsequent heat treatment;
four). Identify and determine the characteristics of microdefects that occur during the implantation of hydrogen ions and trace their evolution during heat treatment;
5). To reveal the possibilities of using the properties of damaged layers created by proton implantation to control the characteristics of silicon electronic devices.
Scientific novelty the results obtained are as follows:
New results have been obtained on changes in the structural and electrical properties of damaged layers of silicon crystals when it is irradiated with protons with energies in the range from 100 to 500 keV, doses from 10 15 to 2-10 16 cmL
For the first time, the features of the formation of damaged layers under the influence of external factors, such as temperature and mechanical stress fields, were revealed.
The change in the structural state of silicon layers implanted by protons during heat treatment in a wide temperature range (100 - 1100C) has been studied.
For the first time, the characteristics of microdefects forming the damaged layer were determined, and their changes during heat treatment were studied.
A model is proposed for the evolution of microdefects in layers of silicon crystals damaged by proton implantation (E = 100-500 keV, D = 10 15 -2-10 16 cm*2) during heat treatment in the temperature range T = 200-1100C.
6. The effectiveness of using damaged layers of silicon crystals formed during proton irradiation and annealing to correct the characteristics of silicon RIP photodiodes is shown.
Practical significance work results:
A technique has been developed for carrying out measurements and processing experimental data to determine the parameters of damaged layers by high-resolution X-ray diffractometry.
A non-destructive method for determining the nature and characteristics of microdefects with different signs of dilation has been developed. The technique based on the analysis of asymptotic diffuse scattering makes it possible to expand the spectrum of observed microdefects.
The established role of external factors of ion implantation in the formation of a defective structure should be taken into account in the technology of creating damaged layers.
The results of determining the characteristics of microdefects in silicon crystals implanted with protons of various doses and energies and subjected to heat treatment in a wide temperature range can be used to create damaged layers with optimal properties.
A method has been developed for protecting the surface of silicon pin-photo diodes, which includes irradiation of the periphery of p-n junctions with protons and subsequent annealing. The optimal mode of irradiation and annealing for this type of devices is determined, which increases the yield.
The main provisions for defense:
Results of determining the integral characteristics of a damaged layer of silicon crystals irradiated with protons with an energy of 150 keV and doses from 2.5 x 10 15 cm" 2 to 2 x 10 16 cm" 2 in the temperature range from 50 to 610C.
Results of studying the influence of external factors of ion implantation: doses, temperatures and fields of mechanical stresses, on the process of formation of a defect structure in silicon crystals.
Results of studies of the structural and electrical characteristics of damaged silicon layers formed by irradiation with protons cE = 200, 300, 100 + 200 + 300 keV, D = 2-10 16 cm" 2 and subsequent heat treatment in the temperature range from 100 to 900C.
Experimentally established nonmonotonic dependence of the integral and electrophysical characteristics of the damaged layer on the annealing temperature.
Model of the evolution of radiation microdefects in silicon layers implanted with hydrogen ions (E = 100-500 keV, D = 10 I5 -2-I0 16 cm" 2) during heat treatment in the temperature range from 200 to 1100C.
Results of investigations of the mechanism of influence of structural and electrophysical parameters of silicon surface layers modified by proton irradiation and subsequent vacuum annealing on VAC of rip-f oto diodes; modes of optimal proton processing of diffusion pin-photo diodes with a depth of pn junctions - 3 μm.
Approbation of work
The main results presented in the dissertation were reported and discussed at the Scientific and Technical Conference of Students, Postgraduates and Young Specialists of MGIEM (TU) (Moscow, 1998), the IX International Conference "Radiation Solid State Physics" (Sevastopol, 1999), Second Russian Conference on Materials Science and Physical and Chemical Fundamentals of Technologies for Obtaining Doped Silicon Crystals "Silicon -2000" (Moscow, 2000), X International Conference "Radiation Solid State Physics" (Sevastopol, 2000), Third International Scientific and Technical conference "Electronics and Informatics - XXI century" (Zelenograd, 2000), the Third International Conference "Hydrogen Processing of Materials" (VOM-2011) (Donetsk - Mariupol, 2001), VI Interstate Seminar "Structural Foundations of Modification
materials using methods of non-traditional technologies” (MHT-VI) (Obninsk,
2001), XI International Meeting "Radiation Solid State Physics" (Sevastopol, 2001), 2nd Interuniversity Scientific School of Young Specialists "Concentrated energy flows in space technology, electronics, ecology and medicine" (Moscow, 2001) , XII International Meeting "Radiation Solid State Physics" (Sevastopol, 2002), Meeting on the growth of crystals, films and structural defects of silicon "Silicon - 2002" (Novosibirsk, 2002), IV International scientific and technical conference "Electronics and informatics - 2002 "(Zelenograd,
d.), the Third Russian Conference on Materials Science and Physical and Chemical Fundamentals of Technologies for Obtaining Doped Silicon Crystals and Device Structures Based on Them "Silicon - 2003" (Moscow, 2003).
The work was awarded the Diploma of the 1st degree, as the best scientific work presented at the scientific and technical conference - a competition of students, graduate students and young specialists in 1998 (MGIEM (TU)).
Publications
The structure and scope of the dissertation
The dissertation consists of an introduction, four chapters, main results and conclusions, and a list of references. The work is presented on 172 pages of typewritten text, contains 58 figures and 4 tables. The list of references includes 101 titles.
Theory of ranges and distribution of ions in solids
Defect formation during implantation of hydrogen ions and the effect of defects on the physicochemical parameters of silicon are a serious problem in creating crystals with desired properties. Microdefects (MD), formed as a result of coagulation of point defects and creating around themselves strong fields of elastic distortions, lead to an additional change in the properties of the crystal and their significant local inhomogeneity. Interest in the study of MDs is determined by insufficient knowledge of both the very nature of MDs, the mechanisms of their formation, and their influence on the physical properties of the crystal and, accordingly, on the main characteristics of devices based on them.
In order to study MD, as well as the possibility of using hydrogen ion implantation in the technology of manufacturing semiconductor structures, it is necessary to consider the effect of proton implantation parameters on the properties of silicon,
The method of ion implantation is universal and non-specific, it allows introducing ions of various elements into any target in strictly controlled amounts, setting the distribution of concentrations over depth by a sequence of ion doses with different energies; in many cases such distributions simply cannot be obtained by other methods.
The primary process in ion implantation is the penetration of ions into a substance and their deceleration to thermal velocities. The resulting depth distribution of interstitial atoms is called the distribution profile, which differs from the final distribution of impurity atoms, to which diffuse processes often contribute. The theory of deceleration of medium-energy ions in amorphous bodies was developed by Lindhard, Schaff, and Shiot (the LSS theory). The essence of the LSS theory is briefly reduced to the following. In the bombardment of solids by charged particles, the decisive role is played by inelastic collisions with bound electrons of the decelerating substance (electronic braking), in which the kinetic energy of the moving ion is spent on electronic transitions in atoms, as well as on the excitation of collective oscillations of electrons and elastic collisions with nuclei (nuclear braking) , in which energy is transferred to atoms as whole. Which of these effects will prevail depends on the energy and mass of the accelerated particles and the mass and serial number of the target atom.
According to the LSS theory, the distribution of ion ranges turns out to be Gaussian and is characterized by the average normal (projected) range Rp and the root mean square (standard) deviation ARP (Fig. 1.1.1). incident ion target surface Fig. 1.1.1. Schematic representation of the total run length R, normal run Rp and standard deviation ARp. The resulting ion trajectories are complex curves and have a statistical character in an amorphous substance. The total trajectory of the ion is called the free path R. If the mass of the ion N\ is much greater than the mass of the target atom M2, then the deviations are small and the ion moves almost rectilinearly. Therefore, the length of its path along the trajectory R differs slightly from Rp. If Mi M2, and the energy of the ion E is not too high, then the trajectory is tortuous and Rp is much less than R. Due to the statistical nature of the movement of ions, the values of Rp and R do not have a definite value, but fluctuate around average values.
It should be noted that the contribution of nuclear deceleration dominates at low implantation energies, while that of electronic deceleration dominates at high ones. When the energy loss curves due to nuclear and electron drag are added together, the total energy loss is constant over a very wide range of incident ion energies. As a result, the total ion path R is approximately proportional to the initial energy of the incident ion.
The simplest ion distribution profile is normal, or Gaussian, for the construction of which only the first two moments are required - the projective range Rp and the standard deviation ARP . The Gaussian distribution is a satisfactory approximation to the real impurity distributions over ranges or depths in cases where these ranges are sufficiently symmetrical. However, this is not always done. Deviations from symmetry are especially noticeable in cases of bombardment of heavier targets by light ions, provided that electron losses predominate.
Various kinds of asymmetric distribution profiles can be used. The classical method for constructing the distribution of ion ranges is the Pearson IV distribution - distribution in the approximation of four parameters: Rp, ARP, distribution skewness and kurtosis p. This method is discussed in detail in. For a number of values of asymmetry in the tables of Pearson's distribution functions in dimensionless units are calculated. The tables allow one to easily construct the distribution profile of the introduced impurity for any known values of Rp, ARp, and asymmetry for a wide range of targets, ions, and their energies.
To obtain the concentration profile N(x) of an impurity, the Pearson distribution must be multiplied by the dose: N(x) = N0-f(x). (1.1.3) There is another method for obtaining the distribution of ion ranges, which is called the Monte Carlo method. The idea of the method is that some element of a solid body is modeled in a computer, the laws are set according to which the ion interacts with atoms, and then an “ion” accelerated to a certain energy is released onto such a simulated “solid body” at a random place on its surface. In other words, a computer experiment is being carried out, in which it is possible to trace the entire path of the ion, including the place where it stops. After repeating this operation many times so that the errors associated with the average statistical deviations are small, the depth distribution of ions can be plotted. The correspondence of the results of such machine experiments to the real one is determined by the correctness of setting the laws of interaction. Simultaneously with the study of ion ranges by the Monte Carlo method, it is possible to obtain such information as the depth distribution of the number of initially displaced target atoms (defect concentration). In all methods, deceleration is divided into elastic and inelastic components, without taking into account the real shell of the point structure of the atom, completely secondary processes, some other simplifying assumptions are used. As a result, the error in calculating the average values of Rp and ARP can reach 20-25%.
Thus, the distribution profile of implanted ions in single crystals depends on numerous factors: the direction of the ion beam, its divergence, the state of the surface, the perfection of the crystal structure, and the target temperature, since it affects the amplitude of thermal vibrations and the kinetics of accumulation of radiation defects.
Double-crystal X-ray diffractometry method
The method is based on registration of the angular distribution of the diffracted beam by the sample under study (that is, on the measurement of the diffraction reflection curve), with further analysis of the parameters of the resulting curve. The most informative and convenient method for studying single crystals is the method of recording rocking curves in the Bragg geometry. The presence on the crystal surface of a layer with varying degrees of damage affects the parameters of diffraction curves in different ways (percentage reflection, half-width, integral reflection coefficient, law of decay of "tails"). By the magnitude of the deviation of these parameters, when compared with theoretically calculated for an ideal sample, one can draw a conclusion about the characteristics of the damaged layer, such as the average change in the lattice parameter, the effective depth of the damaged layer, and determine the deformation profiles.
The sample under study is usually illuminated with an X-ray beam, previously monochromatized by reflection from the monochromator crystal, which remains stationary while the crystal under study rotates near the diffraction angle . The curve characterizing the dependence of the intensity of the radiation reflected by the crystal on the angle of rotation is called the rocking curve, or the diffraction reflection curve (DRC). Characteristics of the perfection of the crystal structure are the following parameters of the rocking curve: the integral reflection coefficient R, which is defined as the ratio of the total intensity reflected by the crystal under study, multiplied by the angular velocity, to the intensity reflected by the monochromator crystal; half-width of the rocking curve, i.e. full width of the curve at half height, which defines the interval of rotation angles in which the intensity decreases by half of the maximum.
Defects in crystals can affect the indicated characteristics of the rocking curves by changing the reflection curve of the crystal under study, i.e. the reflection coefficient R2 and the shape of the R(P) curve change. Comparison of the calculated and experimental rocking curves is the basis for assessing the perfection of the crystal structure.
If the crystal under study reflects according to the Bragg scheme, then in the usual case, dislocations at a density greater than 5 104 cm 2 cause the appearance of such misorientations, which can be easily seen from the broadening of the rocking curve. If the broadening is due only to misorientations, the rocking curve is the sum of individual curves shifted relative to each other by the misorientation angle, since when the crystal is rotated, different sections successively fall into the reflecting position. This broadening is independent of the Bragg angle. In this case, since the half-width of the rocking curve is usually equal to several seconds, if the monochromator and the sample are perfect crystals, then an additional broadening of one or several seconds is reliably fixed. If the rocking curve broadening is caused by the presence in the reflecting volume of sections with different values of interplanar distances dj, then it depends on the reflection angle: Db = -(L)tg9. (2.1.4) With a sufficiently developed substructure, when dislocations are grouped into flat networks, the rocking curves from individual subgrains can be separated, and the total rocking curve will have several maxima. The distance between them is equal to the misorientation around an axis parallel to the sample rotation axis.
If the size of the subgrains is greater than the thickness of the semi-absorption layer, then each subgrain reflects independently of the neighboring ones, and the total area of the rocking curve, which consists of several maxima, will be the same as for a perfect crystal. If their size is less than the thickness of the semi-absorption layer, then the subgrains that are not screened by the subgrains lying above them, which have already completely or partially left the reflecting position, can also make a significant contribution to the total reflected intensity. As a result, the total scattering volume and the angular interval of reflection increase significantly, which leads to a strong increase in the integral reflection coefficient, which in the limit tends to the integral reflection coefficient corresponding to the kinematic theory.
However, the method of recording diffraction reflection curves in a two-crystal scheme has a significant drawback. This method is integral, since the recorded intensity is collected from a wide region of the reciprocal space along the section of the Ewald sphere. In this case, it is impossible to distinguish between the contribution of the diffraction (coherent) and diffuse (incoherent) scattering components to the intensity of the rocking curve. When studying thin layers, the contribution of diffuse scattering from structural imperfections of the damaged layer (clusters of radiation point defects, partially amorphized zones, etc.) to the resulting intensity is large. This makes it difficult to unambiguously interpret the results obtained. The separation of these effects requires a detailed analysis of the intensity distribution in the vicinity of the reciprocal lattice site, which can be implemented on a three-crystal X-ray diffractometer. 2.1.2. Method of three-crystal X-ray diffractometry The possibilities of another X-ray diffraction method in studying the structure of thin damaged layers can be greatly expanded if a third analyzer crystal is introduced into the diffraction scheme, as shown in Figure 2.1.1.
The purpose of this crystal is to analyze the angular distribution of X-rays reflected by the crystal under study. On perfect analyzer crystals, such an analysis of the angular distribution can be carried out with an accuracy of fractions of a second. The resulting three-crystal rocking curves reflect the nature of the structural changes that have taken place in the near-surface layers of the crystal, since have a high sensitivity to the type and characteristics of defects in single crystals. Thus, it is possible to judge the type of defects already on the basis of the very type of intensity profiles measured by the TRD method. Moreover, the high resolution of the TRD method makes it possible to extract very accurate quantitative information about the characteristics of defects.
The difference between the TRD method and conventional three-crystal schemes, in which the first two perfect crystals serve to collimate and monochromatize the radiation incident on the third sample crystal, is that the sample under study acts as the second crystal, and the third (perfect) analyzer crystal performs sweep of the angular distribution of radiation diffracted by the second crystal (Fig. 2.1.1). The sample crystal is deviated from the exact Bragg condition by an angle a, and the analyzer crystal is rotated in a certain angular range near the exact Bragg angle. The X-ray intensity recorded by the detector during the rotation of the third crystal is the TRD spectrum. With this recording scheme, the spectrum usually consists of three peaks, which, according to the established terminology, are called the main, pseudo and diffuse peaks. The angular positions of the peaks are determined by the laws of crystal rotation and diffraction geometry.
Construction of the Distribution Profiles of Embedded Hydrogen and Radiation Defects in Silicon Using the TRIM Program
The processes of defect formation during ion implantation depend on many factors: the target temperature, the dose and energy of implanted ions, their chemical activity, the mass ratio of the ion and target atoms, and the orientation of the substrate. It is not always possible to take into account the influence of all these factors. The TRIM (Transport of Ions in Matter) program makes it possible to make approximate estimates of the primary processes of ion implantation and makes it possible to visualize how the ion will penetrate into the target and what the consequences will be.
Calculations of the distribution profiles of hydrogen ions and radiation defects in depth, made using the TRIM program, are based on the Monte Carlo method. The essence and accuracy of this method are described in Chap. 1, paragraphs 1.1.1, 1.1.3. The TRIM program takes into account only the effect of energy on the impurity ion distribution profile, regardless of the number of introduced ions. Therefore, to collect the necessary statistics, when calculating the distribution profile, an arbitrary number of introduced ions is selected. In this work, to ensure satisfactory calculation accuracy, the number of ions was taken equal to 10000. The spread of the mean range values due to statistical fluctuations inherent in the Monte Carlo method when calculating for 10000 ions using the TRIM program is 1 nm. This number of ions was equal to the implantation dose, which is set as an input parameter of the program. The average defect formation threshold Ej for silicon is 20 eV. The thickness of the target layer, in which the distribution profile is calculated, was assumed to be from 2 to 7 μm, depending on the energy of the introduced ions. Every 2000 particles, the number of impurity ions entering the layer is recalculated into the concentration of ions in this layer (cm 3). Next, the fractions of silicon and impurity atoms are calculated with respect to all particles in a given layer. During the next cycle, collisions are simulated taking into account the probability of interaction between impurity and matrix atoms.
After reading the input data and calculating the necessary parameters, the program proceeds to the cycle of the incident particle, during which collisions are considered and new directions of motion are determined: energy losses during collisions are calculated, then the possibility of the formation of primary knocked-on atoms (PKA) is considered. The change in the ion trajectory due to the elastic interaction with the atom and the loss of energy by the ion due to the inelastic interaction with the electrons of the target atom are taken into account. The process is repeated until the ion energy is more than 0.001 of the initial one. If a PVA is formed, then its data are written to list 1. If the motion of the incident particle stops, then the program passes from the cycle of the incident particle to the cascade cycle. The structure of the cascade cycle is similar to the structure of the cycle of an incident particle. Information on PVA is transferred to list 2, and information on newly knocked-on atoms is written to list 1. After the program has finished working with atoms from list 2, list 1 is merged with the reduced list 2. This procedure is repeated until list 2 will not be exhausted. Then, depending on the number of introduced ions, the program proceeds either to the cycle of the incident particle or displays the results of the calculation.
As a result of calculations performed using the TRIM program, the dependences of the concentration of hydrogen ions and radiation defects on the depth of the implanted layer were obtained at various implantation energies in the range of 100–500 keV and combined irradiation. The program assumes that the same number of vacancies and interstitial atoms are formed during implantation (see Chap.], Section 1.1.3), so the resulting profiles are output relative to one of the types of point defects. 3.4. Resistivity Profile Measurement Technique
The cut is made at a certain angle by grinding with diamond paste with a grain size of not more than 1 micron. The cut sample is mounted in the carriage of an automatic single-probe unit, which provides intermittent sample supply with a step corresponding to a depth displacement of 1 µm. A direct current is passed through the sample having low-resistance non-rectifying current contacts. A probe is placed on the surface of the sample with an oblique cut, which was a tungsten needle with a tip having a rounded radius of about 1 μm. During measurements, a positive potential is applied to the resistor included in the probe circuit. The measured value is the potential of the probe, which varies depending on the location of the touch point of the probe relative to the edge of the oblique cut. To measure the probe potential, we used a direct current electrometric amplifier with an input resistance of 10 Ω.
Results of studying the influence of the dose and temperature of proton irradiation on the integral characteristics of the damaged layer
To reveal the effect of the dose and temperature of proton irradiation on the characteristics of the damaged layer, we studied silicon crystals 0.4 mm thick, with (100) surface orientation, implanted with hydrogen ions with an energy of 150 keV and doses of 2.5-1015 cm"2, 5 -1015 cm 2, 1-1016 cm "2, 2 10 cm. The temperatures of the samples during irradiation were 50C, 140C, 230C, 320C, 430C, 550C, 610C. The studies were carried out using a two-crystal X-ray diffractometer in a dispersionless scheme (Fig. 3.5.2). As a result of the X-ray diffraction experiment, the diffraction reflection curves (DRCs) presented in Figs. 4.1.1 - 4.1.3. According to the experimental DRCs, using the technique described in Section 3.5.3, the quantitative parameters of the ion-implanted layers were obtained: the average effective thickness and the average relative strain.
For all doses of hydrogen implantation into silicon, the nature of the recorded diffraction reflection curves changed with respect to the ideal curve (Fig. 4.1.1 - 4.1.3). As can be seen, the main difference between these curves and the curve corresponding to reflection from a non-irradiated crystal (Fig. 4.1. 1) consists in the appearance (in addition to the main maximum) of an additional intensity oscillation characterizing the formation of a disturbed layer (Fig. 4.1.2, 4.1.3). In all cases, the curves are asymmetric, and the intensity is greater on the side of angles smaller than the Bragg angle than on the opposite side. For all indicated doses at temperatures from 50 to 550C, coherent oscillations are clearly visible from the side of small angles, characterizing the deformation of a positive sign, and the peak from the damaged layer is clearly pronounced (Fig. 4.1.2 a, b, Fig. 4.1.3, curve b) . It can also be seen that the additional intensity increases with increasing dose from 2.5 1015 to 2 1016 cm2.
According to the method described in Section 3.5.3, a program was compiled for calculating the integral characteristics of the damaged layer Leff and Da/a directly from the experimental DRCs for the MATLAB software package. The results of calculating the integral characteristics for all samples are shown on the dependencies Leff(T), Da/a(T), Leff(B), Aa/a(D) (Fig. 4.1.4, 4.1.5).
Analyzing the temperature dependence of Leff and Da/a (Fig. 4.1.4 a, b) it can be seen that the effective thickness and relative deformation of the damaged layer increase, reaching a maximum value at a temperature of 430C. Moreover, at an irradiation dose of 2 10 cm ", the value of Leff increases 2.7 times with an increase in the irradiation temperature, while at lower doses it increases almost 4 times. The relative deformation increases on average 1.3 times with an increase in the temperature of proton irradiation from 50 C to 430 C. With a further increase in temperature, the values of Iff and Da/a sharply decrease.
It is obvious that the formation of a damaged layer in a crystal is caused by two competing processes of evolution of primary radiation defects. After silicon atoms are knocked out of the equilibrium position and interstitial atoms and vacancies are formed, their recombination can take place, in which case the defects disappear. In another case, due to diffusion processes, primary interstitial atoms and vacancies can move away from each other and form stable radiation defects in the form of pairs, clusters, etc.
An analysis of the dose dependence of Leff and Da/a shows an increase in the values of these quantities with dose, and the largest changes in the effective thickness and relative deformation of the damaged layer with dose occur at temperatures up to 140C (the steepest slope of the curve, Fig. 4.1.5 a, b), c 1.8 and 1.3 times, respectively.
In silicon irradiated with protons in the temperature range of 300 - 450C, shallow hydrogen-containing donors are formed, according to . During such high-temperature implantation in silicon, the supersaturated solution of implanted hydrogen decomposes and interacts with radiation defects and impurity atoms. This interaction leads to the formation of electrically active defects that exhibit the properties of shallow donor centers. The structure and parameters of these centers depend on the hydrogen concentration.
The control of diffusion layers is carried out mainly according to such parameters as the depth of occurrence of the formed p–n–transition, conductivity of the surface layer, and surface concentration of impurity atoms.
The most common depth control method p–n–transition is a method of staining a thin section . To measure the depth of an impurity of the order of a few micrometers, it is less convenient to use a spherical section.
Rice. 9.3. Scheme of the process of manufacturing a spherical section on a plate with a diffusion layer:
1 - semiconductor plate; 2- р–n- transition;
3 – steel ball; 4 – painted R– section area
It is made by rotating a steel ball with a diameter of 35 - 100 mm, pressed against the surface of the plate. The formation of a sphere occurs due to the fact that an abrasive suspension is fed into the place of contact between a rotating ball and a crystalline plate or diamond powder is applied to the surface of the ball in the form of an emulsion. For greater measurement accuracy, the grain diameter of the abrasive material should not exceed 1 µm. To reveal the boundaries р–n-transition, the depth of the spherical hole must be greater than the depth of occurrence р–n– transition. The border is revealed by staining (darkening) R– areas due to oxidation in an etchant consisting of 48% hydrofluoric acid with a small addition (up to 0.05–0.1%) of 70% nitric acid.
The depth of the diffusion р–n–transition:
x j = l 2 /(4D) (9.4)
where l- the length of the chord of the contour of a spherical section (Fig. 9.3), measured with a microscope; D- ball diameter.
To improve the accuracy of measurements, several sections are made (up to 5), and the results obtained are averaged.
The most common method for measuring surface resistance is the four-probe method. . The measurement error of surface resistance usually does not exceed 5 - 10%. To determine the surface concentration of the dopant, it is necessary to know the nature of the distribution of impurities in the diffusion region, which depends on the conditions of the process.
Ion implantation
Ion implantation (ion doping) is the process of introducing ionized atoms into the target with energy sufficient to penetrate into its near-surface regions. The successful application of ion implantation is determined mainly by the possibility of predicting and controlling the electrical and mechanical properties of the formed elements under given implantation conditions.
Purpose and application of ion implantation
The most common application of ion implantation in the technology of IC formation is the process of ion doping of silicon. It is often necessary to implant atoms into a substrate, which is covered with one or several layers of various materials. They can be both thin layers of heavy metals (for example, Ta or tantalum silicide TaSi 2) and dielectrics. The existence of a multilayer structure can cause sharp drops in the doping profile at the interface between individual layers. Due to the collision of ions with atoms of near-surface layers, the latter can be knocked out into deeper regions of the alloyed material. Such "fragmentation effects" can cause a deterioration in the electrical characteristics of finished devices.
In many cases, to obtain the required distribution profile of the dopant in the substrate, a method is used based on the preliminary driving of ions with their subsequent thermal dispersal in the target. In this case, implantation is carried out with low ion energy.
The total trajectory of the ion is called the path length R, and the distance traveled by the implanted ion before stopping in the direction perpendicular to the target surface, projected by the path length R p .
9.5.2. Application of ion implantation in VLSI technology. Creating shallow transitions
Formation requirement n+ Shallow layers for VLSI can be easily satisfied using the As ion implantation process. Arsenic has a very short projected path length (30 nm) during conventional implantation with ion energies of 50 keV.
One of the progressive trends in the development of VLSI is the creation of CMOS transistors. In this regard, it is of great importance to obtain small p+ - layers. It is very difficult to form such layers by implantation of B+ ions.
The solution of the problem associated with the implantation of boron at a shallow depth is facilitated in practice by using BF 2 particles as implantable particles. The dissociation of the ВF 2+ molecule during the first atomic collision leads to the formation of low-energy boron atoms. In addition, the use of the BF 2 molecule has an advantage in the process of annealing the structures.
The value of the depth and cultivation of the arable layer of soil for plants.
The power of the arable layer of soil is one of the indicators of fertility and its cultivation. The larger it is, the higher its fertility and crop yields.
Obtaining high and stable yields of agricultural crops is possible only if the needs of plants in water and food are uninterruptedly and completely met. All food (except carbon dioxide in the air) and water enter the plant through the roots from the soil. It is therefore understandable that the exceptional influence that is given in agriculture to the creation of the most favorable soil conditions for the growth and development of agricultural plants. All agrotechnical methods that make up the systems of soil cultivation and the use of fertilizers in crop rotation are ultimately aimed at this. Under the influence of agrotechnical measures carried out during the agricultural use of the soil, its properties change significantly. The direct impact of cultivation methods and the use of fertilizers on the condition and properties of the soil is limited to its upper layer of a certain thickness. It is constantly exposed to tillage implements. Loosening and wrapping this layer with tillage tools provides a stronger effect on its properties. Organic and mineral fertilizers introduced into the soil are distributed, in this soil layer there is an intensive activity of soil microorganisms, which play a leading role in the life of the soil, creating the conditions for its fertility.
On old-arable soddy-podzolic soils, it is especially clearly seen how sharply the upper (arable) layer differs from the underlying soil layers both in appearance and in properties. It is characterized by a looser texture, an increased content of humus and nutrients available to plants, low acidity, and high biological activity.
The increase in the thickness of the arable layer has a positive effect on the water regime of the soil. With its increase, the soil can more fully use the precipitation. On the soil with a deep, highly cultivated arable layer, even with heavy rains, most of the precipitation, as a rule, manages to penetrate into the thickness of this layer and lingers in it, then the excess moisture in excess of the field moisture capacity gradually goes into the underlying layers. On the other hand, on soil with a shallow topsoil, under the same relief conditions, with the same surface condition and the same agricultural use of the soil, showers of rain are usually of little use, since most of the precipitation falls on the soil surface. With an increased amount of precipitation, the soil with a shallow arable layer quickly becomes waterlogged, the plants on it suffer from an excess of moisture and a lack of oxygen in the soil. At the same time, on the nearby soil with a deep arable layer, although this soil contains more moisture than the first, the plants develop normally, no signs of suffering from excess moisture are found. On such soil, cultivated plants resist drought better and suffer less from excessive rainfall.
With an increase in the thickness of the arable layer, the nutritional conditions of cultivated plants improve. Even in very poor soil, the nutrient content is usually hundreds of times higher than what is used by agricultural plants annually at the highest yields. Despite such large reserves of nutrients in the soil, plants do not always have the opportunity to meet their food needs in a timely and complete manner. The predominant part of the nutrients necessary for plants is found in the soil in inaccessible forms - in organic residues, in humus, in the composition of soil microorganisms, and also in sparingly soluble mineral compounds. Only as a result of the processing of these constituent parts of the soil by microorganisms, as well as the decay of the bodies of dead microorganisms, nutrients are obtained in the form of readily soluble compounds available to plants. This useful activity of soil microorganisms can proceed normally only under favorable soil conditions for them - if the soil contains the food they need, heat, moisture, air (oxygen), and in the absence of increased soil acidity. In highly compacted or waterlogged soil, due to a lack of oxygen, the vital activity of microorganisms beneficial to plants is suppressed. Under such conditions, another group of microorganisms develops in the soil, the products of their vital activity not only are not used by agricultural plants for nutrition, but can even adversely affect growth and development.
The number of microorganisms in the soil is extremely high. But in such enormous quantities, soil microorganisms develop under favorable conditions of temperature and humidity only in the arable layer. In the underlying soil layers, the activity of microorganisms is sharply weakened. The predominant part of soil microorganisms needs organic matter as a source of energy necessary for their vital activity and as the main source of substances they need to build a body.
The subsurface layer of soddy-podzolic soils, represented in most cases by a podzolic horizon, contains very little organic matter and microorganisms cannot develop intensively in it, primarily due to a lack of food. Another reason for the strongly suppressed activity of microorganisms in the subsurface layer should be considered a lack of oxygen. Finally, the activity of microorganisms in the subsurface layer is often inhibited due to the increased acidity of the soil in this layer. For these reasons, the activity of microorganisms in soddy-podzolic soils is most pronounced only within the arable layer.
Consequently, the greater the thickness of the arable layer, the greater the biologically active layer, in which, thanks to the vital activity of beneficial soil microorganisms, the food necessary for cultivated plants is uninterruptedly prepared from spring to autumn.
Increasing the thickness of the arable soil layer means increasing the biologically active layer and creating greater opportunities for providing agricultural plants with nutrients. However, it would be a gross mistake on this basis to oppose an increase in the thickness of the arable layer to the use of fertilizers. In early spring, at low temperatures, microorganisms do not work. Industry comes to the aid of agriculture. It provides agriculture with mineral fertilizers that contain nutrients for plants in forms available to them. On cultivated soils with a deep arable layer, the positive effect of fertilizers on the yield is enhanced.
For normal soil nutrition of agricultural plants, the power of development of their root systems and the distribution of roots in the soil in depth are of great importance. The power of development of root systems depends on the level of soil fertility, on the degree of its cultivation. On soddy-podzolic soils, in all agricultural plants, the bulk of the roots (up to 80-90% of their total mass) is located within the arable layer. In the same layer during the entire life of plants there is a predominant part of thin roots covered with root hairs, i.e., the active, absorbing part of the root systems, through which food from the soil enters the plant. This is explained by the fact that nutrients in forms available to plants are contained mainly in the arable layer. The greater the thickness of the arable layer, the greater the volume of cultivated soil is covered by a dense network of roots and the more fully the soil nutrition of plants is provided. On soils with a shallow arable layer, plants have to cover their soil nutrition requirements mainly at the expense of a very limited, obviously insufficient layer.
On cultivated soils with favorable physical and agrochemical properties of subsurface layers, grain crops can consume more than 50% of moisture, 20-40% of nutrients from subsurface horizons.
In the presence of a deep arable layer, cases of death of winter crops under unfavorable wintering conditions are an exception. On such soils, winter crops, as a rule, safely endure even the most difficult wintering conditions. This is explained by the best physical properties of the soil with a deep arable layer, the absence of prolonged autumn waterlogging on them, and the good development of winter crops in the autumn period.
On soils with a deep arable layer, such a phenomenon as the loss of clovers is much less often observed under unfavorable wintering conditions.
With an increase in the thickness of the arable layer, the efficiency of other agrotechnical methods of cultivating crops increases. Consequently, it can be concluded that only in the presence of a deep arable layer and a highly cultivated soil can quite favorable conditions be provided for the growth and development of agricultural plants. They react differently to the thickness of the arable layer and the depth of processing. The first group of crops that bury responding to deep tillage include: beets, corn, potatoes, alfalfa, clover, vetch, broad beans, sunflower, vegetables. The second group of crops that respond moderately to deep tillage include: winter rye, winter wheat, peas, barley, oats, awnless rump. The third group of crops that respond little or not at all to deep tillage includes flax and spring wheat. On soils with a thick arable layer, crop yields are higher.
Techniques for increasing the power of the arable layer. At the beginning of the last century, on the predominant part of arable lands of soddy-podzolic soils, the depth of the arable layer did not exceed 14–15 cm, and on a large area it was no more than 12 cm. brought to 20-22 cm. It is considered economically beneficial to have a arable layer thickness of 30-35 cm. However, it should be borne in mind that an increase in the thickness of the arable layer is not limited to an increase in the depth of processing, it is mandatory to apply organic, mineral and lime fertilizers, sow green manure cultures.
The technology for creating and cultivating a deep arable layer of soddy-podzolic soils provides for leaving the arable layer in its original place, loosening and cultivating the underlying layers. This is especially important to observe with a shallow arable layer.
Currently, there are several ways to deepen the arable layer of the soil.
- Plowing the underlying soil layer with its removal to the surface.
- Complete wrapping of the arable layer with simultaneous loosening of a part of the subsurface layer.
- Loosening to a set depth without wrapping with a plow without skimmers and without mouldboards or chisel plows.
- Deepening by simultaneous plowing of a part of the subsurface layer to the arable layer and the use of loosening of the subarable layer.
- Soil cultivation by longline plows with mutual displacement of horizons.
When choosing a method for deepening and cultivating the arable layer of soddy-podzolic soils, the following indicators must be taken into account: 1) characteristics of the arable layer (thickness, fertility, granulometric composition); 2) characteristics of the subsurface layers: composition (podzolic, illuvial, parent rock), depth, granulometric composition, agrophysical and agrochemical properties (humus content, nutrients, environmental reaction, content of mobile aluminum and ferrous iron).
The most accessible way to increase the thickness of the arable layer is to plow the underlying soil layer and bring it to the surface. It is carried out with conventional plows. At one time, no more than 2-3 cm of the podzolic layer should be plowed. On soils with an arable layer of more than 20 cm, it is deepened by 1/5 of its thickness. In order to prevent a decrease in crop yields from plowing the podzolic horizon to the arable one, it is necessary to apply 80-100 t/ha of organic fertilizers, lime fertilizers to neutralize excess acidity and mineral fertilizers in accordance with the planned yield. Such an application will improve the physical properties and biological activity of the soil and neutralize acidity. The best place to deepen the arable layer by plowing podzolic is a fallow field intended for sowing winter rye and a field for planting potatoes. It is impossible to deepen the arable layer with the involvement of the podzolic horizon under crops such as sugar beet, corn, wheat and flax, even with the application of fertilizers, as this leads to a decrease in their yield.
On soils with a shallow podzolic horizon, when deepening the arable layer, some care must be taken, given that the podzolic layer has unfavorable physical and biological properties, contains almost no plant nutrients in an assimilable form, and has an increased acidity. In this case, the podzolic horizon is not turned out and mixed with the arable horizon, but only loosened. With such a deepening, the layer turns to the depth of the humus layer, and the horizon below it is loosened by soil subsoilers by about 10-15 cm. Later, as the podzolic horizon is cultivated, it can be partially plowed to the arable horizon with an ordinary plow. The gley horizon should not be plowed into the humus horizon, as it contains acidic salts that are harmful to agricultural plants. On such soils, good results are obtained from deepening the arable layer with plows with subsoilers, plows without mouldboards, plows with cut mouldboards and chisel plows. Deepening by loosening the lower layer in place (without eversion) significantly increases aeration, enhances the vital activity of microorganisms, and accumulates in the soil digestible food products for plants both due to the decomposition of organic substances and due to the oxidation of mineral compounds. One of the effective ways to gradually increase the thickness of the arable layer is to deepen it by simultaneously plowing a part of the arable layer to the arable layer and loosening the subarable layer.
It is possible to radically change the arable layer when plowing with longline plows with mutual displacement of soil horizons. This method can be effective if there is a sufficient amount of organic, mineral and lime fertilizers on the farm, otherwise there may be a significant decrease in crop yields. Increasing the thickness of the arable layer requires large material and monetary costs, which is not always within the power of farms.
The results of long-term stationary and short-term field experiments indicate that there are not enough good reasons to recommend gradually deepening the arable layer to 25-30 cm or more. Deepening is advisable only on well-cultivated arable lands under conditions of intensive use of fertilizers, periodic liming and cultivation of crops that respond well to deep cultivation.
On average, for the rotation of a seven-field crop rotation without deepening, 59.1 centners per hectare were obtained, for a deepening of 5 cm - 59.8 centners per hectare, i.e., productivity is almost the same. However, the deepening of the arable layer due to the plowing of podzolic soil leads to high costs of fuel and lubricants for its implementation, and on soils littered with stones, to breakage of plows.
In most farms of the republic, the humus layer of arable soils is 20 cm or more, it is inefficient to deepen it by adding podzolic soil, but it should be cultivated and only in overcompacted areas should the subarable layers be decompacted with non-moldboard implements, preferably with inclined racks. On soddy-podzolic light loamy soils with a humus layer thickness of 20-22 cm, it is possible to obtain cereals 4.5-6.0 t/ha, potatoes - 35-40, root crops - 60-80, hay of perennial grasses - 10-12 t/ha.
Polishing
To improve the quality of surface treatment of semiconductor wafers and reduce the depth of a mechanically disturbed layer, a polishing process is carried out. The polishing process differs from the grinding process in the technological mode, grain size and type of abrasive, as well as the material of the polishing pad. Processing takes place using a free abrasive. The polishing process is carried out on soft polishers, which are hard disks covered with soft material. Micropowders of synthetic diamond, aluminum oxide, chromium oxide, silicon dioxide are used as an abrasive. The polishing material must retain abrasive particles during the processing of the plates. The process of polishing plates can occur in several stages. To begin with, micropowders with a coarser grain size are used. At subsequent stages, after the operation of cleaning from the traces of the previous processing, the material of the polishing pad is changed and finer micropowders are used. The load on the semiconductor wafers is slightly increased. The aqueous suspension is thoroughly mixed during the entire polishing process. The last stage of polishing is of great importance. It makes it possible to remove the background of particles from the surface of the plates, which occurs at the first stages of polishing, and significantly reduce the depth of the mechanically disturbed layer. Chemical-mechanical polishing methods can also be used, which are distinguished by high chemical activity in relation to the processed semiconductor material, .
We polish the plate in several stages, from the working side:
· Preliminary polishing with ASM-3 diamond paste on soft tissue to the depth of the damaged layer of 6-9 microns.
· Re-polishing with ASM-1 diamond paste on soft tissue to the depth of the damaged layer of 4-6 microns.
· Final polishing with ACM-0.5 diamond paste on a soft cloth to the depth of the damaged layer of 3-1 microns., .
Chemical-mechanical polishing
The removal of the residual mechanically disturbed layer from the surface of the substrate is necessary to obtain an atomically perfect structure of the surface layer; therefore, the next technological process is the chemical treatment of the wafers. All types of pollution can be classified according to two criteria: their physical and chemical properties (organic, inorganic, saline, ionic, mechanical, etc.) and the nature of their interaction (physically and chemically adsorbed) with the semiconductor materials on which they are located.
Physically adsorbed contaminants include all types of mechanical particles (dust, fibers, abrasive, metal inclusions), as well as all types of organic materials associated with the substrate surface by physical adsorption forces. The removal of organic contaminants requires a more complex cleaning process, since when heated, they decompose and release gaseous substances that worsen subsequent technological processes.
Chemically adsorbed contaminants include various types of oxide and sulfide films on the surface of the plates, cations and anions of chemicals. Thus, for complete cleaning of the substrate from contaminants, a number of successive operations are used, each of which removes several types of contaminants. Etching is a mandatory technological operation.
When silicon is etched, nitric acid plays the role of an oxidizing agent.
Hydrofluoric (hydrofluoric) acid, which is part of the etchant, converts silicon oxide to silicon tetrafluoride. For etching, which gives a mirror surface of the plates, a mixture of these acids is used in a ratio of 3: 1, the etching temperature is 30 ... 40 ° C, the etching time is about 15 s.
Chemical-mechanical polishing is carried out in two stages:
· Primary polishing with a suspension of Aerosil, SiO 2 (grain size 0.04 - 0.3 µm), to a depth of the damaged layer of 2-1 µm.
· Final polishing with zeolite suspension, to the depth of the disturbed layer of 1-0.5 microns., .
O P: I;. C "À.", 3 and E images
Union of Soviet
Sotsmalmstmmeskmh
2 (5l) M. Cl.
State Committee
Council of the USSR Ministry of Culture for the Affairs of Kzoretenky and Postcards (43) Published on 10/25/78.
Zh. A. Verevkina, V. S. Kuleshov, I. S. Surovtsev, and V. F. Synorov (72) Lenin Komsomol (54) METHOD FOR DETERMINING THE DEPTH OF A DISTURBED LAYER
SEMICONDUCTOR PLATE
The invention relates to the production of semiconductor devices.
Known methods for determining the depth of a damaged layer are based on a change in the physical or electrophysical parameters of a semiconductor material during sequential mechanical or chemical removal of the damaged layer.
So, the method of plane-parallel (oblique) sections with etching consists in the successive removal of parts of the damaged layer, chemical etching of the remaining material and visual control of crack traces. fifteen
The method of cyclic etching is based on the difference in the etching rates of the damaged surface layer and the volume of the semiconductor material and consists in accurately determining the volume of etched material over a certain period of time.
The microhardness method is based on the difference between the microhardness of the damaged layer and the volume of the semiconductor material and consists in layer-by-layer chemical etching of near-surface layers of the material and measurement of the microhardness of the remaining part of the semiconductor wafer.
The method of infrared microscopy is based on different absorption of radiation
IR range by semiconductor wafers with different depths of the damaged layer and consists in measuring the integral transmission of IR radiation by a semiconductor wafer after each chemical removal of a layer of material.
The electron diffraction method for determining the depth of a damaged layer is based on preparing an oblique section from a semiconductor wafer and scanning an electron Fo beam II on the section from the surface of a single crystal to the point from which the diffraction pattern does not change, followed by measuring the distance traveled.
However, in the known methods of control, one should note either the presence of expensive and bulky equipment, or
599662 the use of aggressive and toxic reagents, as well as the duration of the result.
There is a known method for determining the depth of the damaged layer in a semiconductor S ynastin by heating the semiconductor, Qrm it consists in that the conductor plate with the damaged layer is placed in a vacuum chamber in front of the input window of the exoelectron receiver, with which the exoelectron emission from the semiconductor surface is measured.
To create an electric field that pulls the ecoelectrons, a grid is placed over the surface of the semiconductor, onto which a negative voltage is applied. Further, when the semi-conductor is heated from its surface, an eco-electron emission occurs, which is measurable with the help of a receiver1 and additional equipment (shi (eco-cavity amplifier and pulse counter), while the temperature position and intensity of the emission peaks are determined by the depth of the damaged layer. 25
This method requires the presence of vacuum Equipment, and in order to obtain emission spectra, it is necessary to create a discharge in the chamber no worse than 10 Torr. The creation of such SR conditions before the actual process of determining the density of the damaged layer leads to the formation of the final result only through
40-60 miE „In addition, according to this method, it is impossible to simultaneously determine the 35 crystallographic orientation of the semiconductor wafer.
The purpose of the present invention is to simplify the process of determining the depth of the damaged layer, while simultaneously determining the crystallographic orientation of the semiconductor plate.
This is achieved by the fact that the plate is heated from B high-frequency lope until the appearance of the skeen effect and held for 2-5 s, after which the depth of the damaged layer and the orientation of the single-crystal plate are determined from the average maximum length of the traces of oriented propagation channels and their shape.
The drawing shows the dependence of the average maximum area of traces of orientated penetration channels on the silicon surface of orientation (100) on the depth of the damaged layer „
During induction heating of a semiconductor plate (with simultaneous initiation of intrinsic conduction in the semiconductor), a skin effect appears on the periphery of the latter, which is detected by the appearance of a brightly glowing rim on the plate. When holding the wafer in the specified conditions for 2-5 s, it was found that on both sides of the periphery of the semiconductor wafer, figures are formed in the form of triangles for the semiconductors oriented in the plane, and rectangles for the (100) orientation.
These figures are traces of oriented proppant channels.
The formation of channels is apparently due to the interaction of ponderomotive sip electric fields with cracks and other defects in the near-surface layer of the semiconductor, leading to the breaking of interatomic bonds in the defect zone. X-spectrons are further accelerated in a strong electric field, ionize atoms along the way, causing an avalanche, and , thus propagating my crystal along the defect.
It has been found experimentally that the maximum length (n area) of surface traces of oriented penetration channels depends on the size (extension) of the defect itself in the structure of the semiconductor. Moreover, this dependence is linear, i.e., the larger the size of the defect, for example, the length of the cracks, the greater the area of the trace of the oriented propagation channel that has arisen on this defect.
Example When polishing silicon wafers with diamond pastes with successively decreasing grain diameter, a calibration curve is preliminarily constructed. The values of the depth of the damaged layer in silicon, determined by any of the known ones, fall along the y-axis. ny methods, for example, cyclic etching. On the abscissa axis, the average maximum length (area) of the traces of penetration, corresponding to a certain depth of the disturbed layer. For this purpose, plates with a diameter of 40 mm, taken from various stages of polishing, were used. placed on a graphite substrate in a cypindrical RF inductor with a diameter of 50 mm of the installation with a ZIVT power and an operating frequency of 13.56 MHz. The plate is kept in the IC field for 3 s, after which the average maximum length (area) of the trace of the melt channel is determined on a microscope of the MII-4 type using 10 fields of view
Compiled by N. Khlebnikov
Editor T. Kolodtseva TehredA. AlatyrevProofreader S. Patrusheva
Order 6127/52 Circulation 918 Subscription
UHHHfIH State Committee of the Council of Ministers of the USSR for inventions and discoveries
113035, Moscow, Zh-35, Raushskaya emb., d, 4/5
Branch PPP Patent, Uzhhorod, st. Design, 4 songs. In the future, with a partial change in technology, i.e., for example, when changing the type of machine, polishing material
> grit of diamond paste, etc., one of the plates is removed from a certain stage of the technical process and subjected to HF processing, as described above. Further, using the calibration curve, determine the depth of the damaged layer and make adjustments to the s technology. Orientation is also controlled visually after RF processing.
The timing of the process of determining the depth of the damaged layer and the orientation of the semiconductor, according to the proposed technical solution, shows that the entire process from its beginning (placing the plate into the RF inductor) and until the final result is obtained takes
The implementation of the described method in semiconductor production will make it possible to carry out express control of my
29 bins of the damaged layer on both surfaces of the semiconductor wafer with one-time determination of its crystallographic orientation, to reduce the use of aggressive and toxic reagents and, thereby, improve safety and working conditions.
Claim
A method for determining the depth of a damaged layer of a semiconductor wafer by heating a semiconductor, characterized in that, in order to simplify the process and simultaneously determine the crystallographic orientation, the wafer is heated in a high-frequency field until the appearance of the skin effect and kept in this way for
2-5 s, after which it is oriented along the average maximum length of the traces. the depth of the damaged layer and the orientation of the single-crystal layer BbK
