KR100687796B1 - Device and method for measuring dopant profiling of semiconductors - Google Patents

Abstract

A device and a method for measuring dopant density profiling of a semiconductor are provided to measure quantitatively the dopant density profiling of a contact point with high spatial resolution. A diamond thermoelectric probe(100) includes a diamond lamination to form a thermocouple with heavily-doped silicon. An AC power source applies AC power to the diamond thermoelectric probe in order to heat a contact part of the thermocouple. A first voltage measurement unit(300) is connected with the diamond thermoelectric probe in order to measure a thermoelectric voltage by using a variation of temperature. A second voltage measurement unit(400) is connected to the diamond thermoelectric probe in order to measure a thermoelectric voltage and derive a thermoelectric constant of a surface of a semiconductor.

Description

Device and method for measuring impurity concentration in semiconductors {Device and Method for measuring dopant profiling of semiconductors}

1 is an electron micrograph of the tip of the diamond probe of SSAN dedicated SS NANOWORLD.

2 is a schematic diagram illustrating the principle of a Scanning Thermo-Electric Microscopy (SThEM) method.

3 is a perspective view showing a probe in the form of a bridge disclosed in US Pat. No. 6,518,872.

4 is a conceptual diagram schematically showing the configuration of a measuring device according to an embodiment of the present invention and a conductive diamond probe used in the measuring device.

5 is a conceptual diagram illustrating a measuring device and a measuring method according to an embodiment of the present invention.

6 is a graph showing a temperature measuring device and a frequency vs. temperature fluctuation graph previously tested by the applicant.

Fig. 7 is a graph showing signal processing and results thereof in the preceding experiment.

8 is a flowchart illustrating a method for measuring impurity concentration of a semiconductor according to an embodiment of the present invention.

 Explanation of symbols on the main parts of the drawings

10: heavy doped conical silicon 20: silicon dioxide thin film

30: diamond thin film 40: thermocouple contact portion

50: probe tip 60: contact

100: conductive diamond probe 200: AC power

300: first voltage measuring device 400: second voltage measuring device

500: semiconductor sample 501: sample surface

The present invention relates to a method for nanoscale and quantitatively measuring semiconductor density density using thermoelectric probes. More specifically, the present invention relates to three methods of heating, temperature measurement, and thermoelectric voltage measurement in a nano tip of a probe. The present invention relates to a measurement method capable of quantitatively measuring the density of impurities (dopant density profiling) at a contact point of a semiconductor surface with a spatial precision of about nanometers.

According to the International Technology Roadmap for Semiconductors (ITRS), which is often cited as a representative data for predicting the future development direction of integrated circuit manufacturing technology, the gate length of semiconductor devices is reduced to about 30 nm by 2009. Circuit fabrication technology inevitably requires structural analysis technology of semiconductor devices with nanometer scale accuracy. Integrated circuits are fabricated by doping a single crystal silicon wafer, which is basically a Group 4 element, with high spatial precision of n-type impurities, which are Group 5 elements, and p-type impurities, which are Group 3 elements. Therefore, the structural analysis technology of the semiconductor device basically means a silicon density impurity density measurement technology (dopant density profiling technology). Because of this technical importance, researches on techniques for measuring the impurity concentration distribution of semiconductors have been conducted in a wide variety of ways. Three representative methods are Scanning Capacitance Microscopy (SCM), Scanning Spreading Resistance Microscopy (SSRM), and the most recently developed Scanning Thermo-Electric Microscopy (SThEM), which will be described in detail below.

First, SCM (Scanning Capacitance Microscopy) utilizes the capacitance between the tip of the probe and the semiconductor surface as a function of the impurity concentration on the semiconductor surface. The capacitance between the probe tip and the semiconductor surface is shown in Equation 1.

Where C _e is the capacitance, Q is the charge amount, V is the potential difference, ε ₀ is the relative dielectric constant, A is the area of the probe tip, and d _c is the distance between the probe tip and the semiconductor surface. Currently the highest precision of SCM is reported to reach 10-20 nm.

However, this resolution is obtained by sacrificing the measurement sensitivity. Since the capacitance is proportional to the area of the probe tip, the sharper the tip of the probe, the higher the resolution, but the decrease in measurement sensitivity is inevitable. In addition, the SCM is very difficult to exceed the current accuracy because the probe must be kept in contact with the surface to measure the concentration of impurities.

Next, SSRM (Scanning Spreading Resistance Microscopy) is a method of measuring contact impurity concentrations to overcome these disadvantages of SCM.The tip of the probe and the resistance of the semiconductor surface contact are the electrical resistivity and the equation of the semiconductor surface. It is a technique that takes advantage of the relationship as shown in 2.

Where R is the electrical resistance of the probe tip and the semiconductor surface contact, ρ is the electrical resistivity, and a is the radius of the contact. The principle of the method is a technique utilizing the electrical resistance is a function of the impurity concentration, as can be easily seen from Equation 2 above. The spatial precision of this technique is known to be about 25 nm, but recently a paper has been reported that a resolution of about 5 nm can be obtained. The technique requires a diamond probe to achieve stable electrical contact between the probe and silicon. Currently, NANOWORLD is commercially selling SSRM-only diamond probes as shown in FIG.

The SSRM is a fairly simple and effective method for qualitatively determining the concentration distribution of impurities. However, as can be seen from Equation 2, in order to quantitatively accurately measure the resistance ρ of the silicon contact from the measured resistance R, the radius a of the contact must be known correctly, but the tip of the nanoscale probe and the electrical contact between the silicon sample surface Knowing the size correctly is very difficult. Therefore, there are many problems in quantitatively analyzing the type and concentration of impurities.

It is well known that electrical contact occurs in the region where the tip of the probe and the silicon sample make mechanical contact with a pressure of about 10 Mpa or more. Factors affecting the size of the electrical contact are (i) the magnitude of the contact force between the diamond probe tip and the silicon sample surface, (ii) the shape of the diamond probe tip, and (iii) the native oxide on the silicon sample surface. oxide) thickness and surface state. Among these, (i) the magnitude of the contact force can be adjusted very precisely using AFM (Atomic Force Microscope). However, it is very difficult and error prone to precisely and repeatedly control (ii) the shape of the probe tip and (iii) the state of the sample surface.

For example, Figure 1 is an electron micrograph of the tip of the probe for SSRM NANOWORLD company, it can be seen that the curvature of the tip of the probe is determined by the growth pattern of the diamond grain (grain) and the thickness of the diamond thin film. The shape of the probe tip varies slightly from probe to probe because the growth pattern of the probe grain varies from probe to probe. On the other hand, the silicon exposed to air has a native oxide on the surface, which is affected by the environment and time that the silicon sample is exposed to the air. It is difficult.

Finally, Scanning Thermo-Electric Microscopy (SThEM) is the most recently developed method published in Science in 2004. The principle is described with reference to FIG. 2 as follows. First, the sample is placed in an ultrahigh-vacuum of 5 × 10 ⁻¹¹ torr, and then heated with a heat source attached to the back of the sample so that the temperature of the sample is maintained at a temperature of about 5 to 30K above ambient. It was. Subsequently, contacting the surface of the sample with a nanoscale probe tip at room temperature results in a local temperature gradient in the sample due to heat transfer to the probe tip. The local thermoelectric coefficient is measured with spatial resolution of several nanometers by measuring the thermoelectric voltage which is proportional to the local thermoelectric coefficient of the probe and sample contact generated by this temperature gradient, and the concentration distribution of semiconductor impurities is analyzed with high precision. It was.

However, this method requires an ultrahigh-vacuum state for local temperature gradients and is difficult to apply to silicon having a natural oxide film on its surface due to the use of STM. There is no In addition, it is very difficult to actually implement the method because the measurement and probe transfer do not occur simultaneously but intermittently. Therefore, the method is meaningful in that it is effective to measure the thermoelectric coefficient in measuring the impurity concentration distribution of the nanoscale, but there is a problem in that there is a big limitation in practicality.

Meanwhile, a patent related to the technical field of the present invention is US Patent No. 6,518,872 registered in 2003, which has a high resolution scanning thermal probe and method of manufacturing according to the present invention. It contains claims for the process of manufacturing probes that measure the temperature of the counterpart by making thermocouples with MEMS (Micro Electro Mechanical Systems) processing and chemical vapor deposition processes, and the bridge between the probe and the thermocouple. ), But the patent is implemented by employing a probe having a plurality of legs, which is different from a nano thermocouple implemented at the tip of the probe.

The present invention is to provide a new measurement method that is simple to use while having a higher spatial precision and quantitative precision than the existing silicon impurity concentration distribution measurement method.

In order to solve the above technical problem, an object of the present invention is to perform an impurity concentration distribution of a semiconductor surface contact by simultaneously performing local heating, temperature measurement, and thermoelectric voltage measurement on a thermocouple of a probe tip in contact with a sample surface. It is to provide a measuring method that can quantitatively measure density profiling) with spatial precision of about nanometer.

In order to achieve the above object, an impurity concentration measuring apparatus of a semiconductor of the present invention includes a diamond thermoelectric probe in which conductive diamond is laminated to the outside to form a thermocouple with heavy doped silicon;

An AC power source for locally heating the thermocouple contact unit by applying an alternating current to the thermoelectric probe;

A first voltage measuring device connected to the thermocouple probe to measure a thermocouple voltage (ΔV _tip ) by a temperature change ΔT _junc occurring at the thermocouple contact portion; And

And a second voltage measuring device connected to the diamond thin film and the semiconductor sample to measure a thermoelectric voltage (ΔV _tip-sample ) to derive a thermoelectric coefficient on the surface of the semiconductor.

The conductive diamond probe may comprise heavy doped conical silicon;

A silicon dioxide (S _i O ₂ ) thin film which is an insulator laminated on an outer surface except for an end portion of the silicon; And

A diamond thin film laminated on the outside of the silicon dioxide thin film and the silicon end and forming a tip of a probe; It is preferable that it is a probe consisting of.

In addition, in the method for measuring the impurity concentration of a semiconductor of the present invention, a thermocouple probe in which a conductive diamond thin film is laminated is contacted with a sample surface to form nano-contacts, and then an alternating current is applied to the thermocouple of the probe to localize the thermocouple contact portion. Firstly generating heat;

After measuring the thermoelectric voltage (ΔV _tip ) generated by the temperature fluctuation (ΔT _junc ) of the thermocouple contact portion by the heat generation with a first voltage measuring device, the thermoelectric voltage (ΔV _tip ) and the thermoelectric coefficient (S _si between silicon and diamond) a second step of calculating the temperature variation ΔT _tip of the probe tip using _-dia );

After measuring the thermoelectric voltage (ΔV _tip-sample ) of the probe contact formed between the tip of the conductive diamond thin film and the semiconductor sample surface with a second voltage measuring device, the thermoelectric voltage (ΔV _tip-sample ) and the temperature fluctuation (ΔT _tip) Calculating a thermoelectric coefficient (S _dia-sub ) between the diamond thin film and the sample surface using; And

After calculating the thermal coefficient of the sample surface (S _sample) by using the thermal transfer coefficient (S _dia-sub), from the thermal coefficient (S _sample) calculates the fourth step of quantitative analysis of the impurity concentration in the semiconductor; a Characterized in that made.

In the second step, the driving voltage of frequency ω is separated from the measured thermoelectric voltage to calculate the thermoelectric voltage ΔV _tip of the frequency 2ω generated by the Seebeck effect.

In the third step, the driving voltage of frequency ω is separated from the measured thermoelectric voltage to calculate a thermoelectric voltage ΔV _tip-sample of frequency 2ω, which is generated by the Seebeck effect.

The thermoelectric coefficient (S _sample ) of the sample surface in the fourth step may be obtained by excluding the thermoelectric coefficient (S _dia ) of the diamond thin film 30 from the calculated thermoelectric coefficient (S _dia-sample ), which is an impurity of a semiconductor. The concentration (n) is a function of the thermoelectric coefficient (S) (n = f (S)) and the relationship between them is plotted by experiment. Based on this, it is possible to quantitatively analyze the impurity concentration of the semiconductor according to the thermoelectric coefficient. have.

Referring to the drawings the configuration of an embodiment of the impurity concentration measuring apparatus and method of the semiconductor of the present invention as described above is as follows.

4 is a conceptual diagram schematically showing a configuration of a measuring device according to an embodiment of the present invention and a conductive diamond probe used in the measuring device, and FIG. 5 illustrates a measuring device and a measuring method according to an embodiment of the present invention. 8 is a conceptual diagram for explaining the present invention. FIG. 8 is a flowchart of a method for measuring impurity concentration of a semiconductor according to an exemplary embodiment of the present invention.

In addition, Figure 6 is a view showing a temperature measuring device and a frequency vs. temperature fluctuation graph previously experimented by the present applicant, Figure 7 is a diagram showing a graph of the signal processing and the results in the preceding experiment.

4 and 5, in the semiconductor impurity concentration measuring apparatus according to the embodiment of the present invention, the conductive diamond thin film 30 to form a thermocouple with the heavy doped silicon (10) to the outside The stacked diamond thermoelectric probe 100, an alternating current power supply 200 for locally heating the thermocouple contact portion 40 by applying an alternating current to the thermoelectric probe 100, and a temperature variation of the thermocouple contact portion 40 ( A thermoelectric voltage (ΔV _tip ) generated by ΔT _junc ) and a first voltage measuring device 300 connected to the thermoelectric probe 100, the diamond thin film 30 and the semiconductor sample 500 are connected to thermoelectric. A second voltage measuring device 400 for measuring a voltage ΔV _tip -sample to derive a thermoelectric coefficient of a semiconductor surface is included.

The conductive diamond probe 100 includes a heavy doped conical silicon 10 and a silicon dioxide (S _i O ₂ ) thin film 20 which is a non-conductor laminated on an outer surface except for an end of the silicon 10. And a diamond thin film 30 stacked on the outside of the silicon dioxide thin film 20 and the end of the silicon 10 and forming the tip 50 of the probe 100, so that a current flows through the conical silicon 10. As the conductive diamond thin film 30 flows, the thermocouple contact portion 40 is locally heated by the Joule effect.

In addition, as shown in Figure 8 the impurity concentration measurement method of a semiconductor according to an embodiment of the present invention,

The thermocouple probe 100 having the conductive diamond thin film 30 laminated thereon is brought into contact with the sample surface 501 to form a nano-contact point 60, and then an alternating current is applied to the thermocouple of the probe 100 to provide a thermocouple contact point. A first step (S1) of locally generating the unit 40,

After the thermal voltage (ΔV _tip) caused by temperature variation (ΔT _junc) of the thermocouple contact point (40) by the heating was measured by first voltage measuring device 300, the thermal voltage (ΔV _tip) and silicon ( A second step S2 of calculating a temperature variation ΔT _tip of the probe tip 50 by using the thermoelectric coefficient S _si-dia between 10) and the diamond 30;

After measuring the thermoelectric voltage (ΔV _tip-sample ) of the probe contact point 60 between the conductive diamond thin film 30 and the semiconductor sample surface 501 by the second voltage measuring device 400, the thermoelectric voltage (ΔV _tip) and a third step of deriving a _{sample) (S3), - -sample} ) and temperature variation (ΔT by using the _tip) thermal coefficient (S _dia between the diamond film 30 and the sample surface 501

A fourth step after calculating the thermal coefficient (S _sample) of a sample surface 501 using the thermal transfer coefficient (S _dia-sample), quantitative analysis of the impurity concentration in the semiconductor from the calculated thermal coefficient (S _sample) It consists of (S4).

In the second step S2, the driving voltage of frequency ω is separated from the measured thermoelectric voltage to calculate a thermoelectric voltage ΔV _tip of frequency 2ω generated by the Seebeck effect.

In the second step S2, the thermoelectric coefficient S _si-dia between the silicon 10 and the diamond thin film 30 may be conveniently obtained by experiment, but may be theoretically obtained by Equation 3 below.

S _{si - dia} = S _si -S _dia

Where S _si Is the thermoelectric coefficient of silicon 10, S _dia Is the thermoelectric coefficient of the diamond thin film 30.

On the other hand, the thermoelectric coefficient (S _si ) of the silicon 10 and the thermoelectric coefficient (S _dia ) of the diamond thin film 30 can be obtained according to Equation 4 below according to the concentration and type of impurities, respectively.

Where n and p are electron and hole concentrations, μ _n and μ _p are electron and hole mobility, S _n and S _p is the Seebeck coefficient of electrons and majors.

In addition, the S _n and S _p can be obtained by the following equation.

,

,

Where k _B is the Boltzmann constant, e is the charge, r _e and r _h are the scattering coefficients for electrons and holes, and N _c and N _v are the conduction and balance coefficients for electrons and holes.

The temperature variation ΔT _tip of the probe tip 50 in the second step S2 may be obtained by Equation 6 below.

Here, ΔV _tip is the thermoelectric voltage at the thermocouple contact portion, S _si-dia is the thermoelectric coefficient between the silicon 10 and the diamond thin film 30.

In addition, in the same manner as in the second step S2, the thermoelectric voltage ΔV of the frequency 2ω generated by the Seebeck effect by separating the driving voltage of the frequencyω from the measured thermoelectric voltage in the third step S3. _tip-sample ).

In the third step S3, the thermoelectric coefficient S _{dia -sample} between the diamond thin film 30 and the sample surface 501 may be obtained by Equation 7 below.

Here, ΔV _tip-sample is the thermoelectric voltage at the probe contact point 50 which is the contact point between the conductive diamond thin film 30 and the semiconductor sample surface 501, and ΔT _tip is the temperature variation value of the probe tip 50 obtained in the third step. to be.

In the fourth step S4, the thermoelectric coefficient S _sample of the sample surface 501 is obtained by separating the thermoelectric coefficient S _dia of the diamond thin film 30 from the calculated thermoelectric coefficient S _dia-sample . The impurity concentration (n) of the semiconductor is a function of the thermoelectric coefficient (S) (n = f (S)) and the relationship between the two is plotted by experiment, based on which the impurity of the semiconductor according to the thermoelectric coefficient Concentration can be analyzed quantitatively.

Hereinafter, the operation of the impurity concentration measuring apparatus and method of the semiconductor according to an embodiment of the present invention as described above will be described in detail.

In order to apply the method for measuring the semiconductor impurity concentration proposed by the present invention to the semiconductor sample 500, the conductive diamond probe 100 is required due to the physical properties of silicon. In order to make stable electrical contact by contacting the tip 50 of the probe 100 to the semiconductor silicon surface 501, the local pressure of the contact point should be 10 Gpa or more, which is a stress beyond the yield strength of tungsten or steel. to be. Therefore, in order to achieve stable electrical contact through mechanical contact with the silicon surface 501, a material having a high electrical conductivity while having a yield strength far exceeding 10 Gpa is required. Ideal as such a material is conductive diamond. Since diamond is the hardest of all natural materials and more chemically stable than titanium or platinum group metals, diamond can maximize the material's strengths in nanotechnology, where maintaining the size and shape of the nanometer scale is key. Material.

Therefore, first, silicon dioxide (S _i O), which is a non-conductor in heavy doped conical silicon using micro electro mechanical systems (MEMS) and nano-fabrication processes, on the tip 50 of the probe 100, is used. ₂ ) is laminated and a conductive diamond probe laminated with a conductive diamond thin film to be bonded to the end of the silicon on the silicon dioxide thin film. The tip of the probe should be formed at about 100 nm, which is nanoscale.

The impurity concentration measurement of a semiconductor according to an embodiment of the present invention utilizing the conductive diamond probe 100 may be performed by energizing a thermocouple formed by the silicon 10 of the probe and the conductive diamond thin film 30 with an alternating current. At the same time, the probe 50 is generated at the contact point 60 of the probe 100 and the semiconductor sample 500 while the tip 50 of the probe 100 is brought into electrical contact with the semiconductor sample 500. It is used to measure thermoelectric voltage, which is to perform three functions of (a) heating of probe tip, (b) temperature measurement, and (c) thermoelectric voltage measurement at the nanoscale probe tip 50 simultaneously. This is achieved by quantitatively measuring the thermoelectric coefficient of the surface contact 60.

That is, the present invention relates to a new measurement method capable of simultaneously performing three functions of heating, temperature measurement, and thermoelectric voltage measurement of nano thermocouples located at the probe tip 50.

The first requirement of this measuring method is to measure the thermoelectric voltage induced by the temperature change of the contact while heating the thermocouple electrically. Until now, thermocouples have been used only as passive sensors that measure the temperature of two different metal or semiconductor contacts. This is because, when the thermocouple is electrically heated, the thermoelectric voltage generated in the thermocouple is much weaker than the voltage required for the electric heating.

However, the applicant has theoretically induced and experimentally verified that it is possible to measure the thermoelectric voltage generated at the contact when the contact of the thermocouple is heated with an alternating current as described in the prior patent application 2005-0116008. Based on this. This is briefly described as follows.

The basic principle is to periodically heat the junction of the thermocouple probe using Joule heat generated by AC current, and measure the Seebeck voltage generated at the junction of the thermocouple according to the temperature change of the junction.

That is, when the contact of the thermocouple probe is energized with an alternating current having a frequency of ω, Joule heat is generated at the frequency of 2 ω, and the temperature at the contact of the thermocouple probe also vibrates at a frequency of 2 ω. Accordingly, a thermoelectric voltage having a frequency of 2ω is generated at the junction of the thermocouple probe. Since the frequency of the thermoelectric voltage is 2ω and the frequency of the driving voltage required to supply the heating current is ω, the voltage having a frequency of 2ω is separated. By measuring, the thermocouples can be electrically heated while simultaneously tracking the temperature of the thermocouples.

First, the local heating of the probe tip is described as follows. As shown in FIGS. 4 and 5, when an alternating current having a frequency of ω is passed from the AC power source 200 to the conductive diamond thin film 30 from the heavy doped silicon 10 of the probe 100. Due to the Joule effect, a heat generation phenomenon with a frequency of 2ω occurs in all parts through which the current passes, and thus a temperature variation with a frequency of 2ω occurs. It is more noteworthy here that the contact portion 40 of the thermocouple where the heavy doped silicon 10 and the diamond thin film 30 meet rapidly generates an intensive heat generation phenomenon as the area through which the current passes rapidly decreases. . Therefore, the temperature fluctuation mainly occurs in the thermocouple contact portion 40 located at the tip 50 of the probe 100.

Next, the temperature measuring method of the probe tip 50 will be described. The temperature fluctuation ΔT _junc concentrated on the tip 50 of the probe 100 is the contact portion 40 of the heavy-doped silicon 10 and the conductive diamond thin film 30 by the Seebeck effect. Generates a thermoelectric voltage (ΔV _tip ) having a frequency of 2ω. At this time, since the frequency of the driving voltage applied by the AC power source 200 is ω and the frequency of the temperature fluctuation is 2ω, if the frequency ω caused by the driving voltage is removed from the generated voltage, the silicon 10 of the probe and the diamond thin film ( The thermoelectric voltage ΔV _tip at the contact portion 40 of 30 may be separated and measured. At this time, using the doped silicon (10) and the thermoelectric coefficient (S _si-dia ) of the diamond film 30, calculated by the equation (5), the temperature variation of the probe tip from the measured 2ω voltage (ΔT) _tip ) can be measured.

Finally, a method of measuring the thermoelectric voltage on the silicon sample surface 501 by measuring the thermoelectric voltage will be described below. Since the size at the tip 50 of the diamond thin film 10 is very small, ˜10 ⁻³ μm ³ , the temperature variation at the contact 60 of the silicon sample surface 501 to measure the thermoelectric coefficient is heavy doped. (heavy doping) has a value very close to the temperature fluctuation in the contact portion 40 of the silicon 10 and the diamond thin film 30, and thus the contact between the diamond thin film 30 and the silicon sample surface 501 ( 60), a thermoelectric voltage (ΔV _tip ) having a frequency of 2ω is generated by the Seebeck effect. The thermoelectric voltage also has a frequency different from ω, which is the frequency of the driving voltage, with a frequency of 2ω, so that separate measurement is possible as described above.

The thermoelectric voltage (ΔV _tip ) between the conductive diamond thin film 30 and the silicon sample surface 501 measured as described above and the temperature value (ΔT _tip ) of the probe tip 50 measured above are measured. It is possible to quantitatively analyze impurity concentrations and types.

The same signal detection method as this measurement technique has already been proved through experiments using gold (Au) and nickel (Ni). 6 is a result of experimentally verifying this using gold (Au) and nickel (Ni). In the above experiment, the temperature measured by the 3ω method, which is well known as a reference method for measuring the temperature of the conductor, was used as a reference for comparing the measurement results. As a result, it can be seen that the temperature amplitude measured at the thermocouple contact point is slightly smaller than the temperature amplitude measured by the 3ω method. The reason for this is because the cross-sectional area of the conducting wire through which the electric current passes is widened in the portion for measuring the thermoelectric voltage, and the temperature of the thermocouple contact point is actually slightly lower than the temperature of the conducting wire because of the amount of heat transferred to the terminal required for the measurement. The above experiment is a relatively simple experiment, but experimentally proved that the thermocouple can be used as an active sensor rather than a passive sensor.

On the other hand, since the measuring method of the present invention must perform three functions of heating, temperature measurement, and thermoelectric voltage measurement of the thermocouple, the thermoelectric voltage at the thermocouple contact can be detected at two or more pairs of terminals. Should be Figure 7 shows the method and results of the signal detection experiments associated with this. An alternating current was applied to terminals 1-2, while thermoelectric signals were detected at all terminals at terminals 1-2, 2-3, 3-4, and 4-1. The measurement results are as shown in the graph below in FIG. Note that the amplitude of the signal detected at terminal 1-2 is the largest, the signal magnitudes of terminals 2-3 and 4-1 are almost the same, and the value of the signal measured at terminal 3-4 is the smallest. The reason for this is as follows.

First, the current tends to flow through the path of least resistance. That is, when current passes between terminals 1-2, as soon as possible (as shown in FIG. 7) inward (in the upper quadrant of the upper figure), the shorter path passes. Therefore, it is assumed that the current density is higher toward the inside as shown in the figure. Since the local heat generation is proportional to the square of the current density according to Joule's law, the temperature amplitude of the contact point of gold and nickel, which is the second upper limit in the quadrangle of the upper drawing, is the largest, and toward the outer side (four upper directions). The temperature amplitude becomes small.

Secondly, the temperature measured in a thermocouple measures the temperature of a contact on a path with the smallest electrical resistance, as in the case where a current passes through two different metals or semiconductors. Thus, the signal measured at terminal 1-2 is the temperature amplitude of the two quadrant contacts, the terminals 2-3 and terminal 4-1 are the temperature fluctuations of the one quadrant and three quadrant contacts, respectively, and are measured at terminals 3-4. The signal is the temperature amplitude of the quadrant.

The above experiments infer that the three functions of thermocouple heating, temperature measurement and thermoelectric voltage measurement can be performed simultaneously. On the other hand, it is noteworthy that a significant temperature difference was measured even in a square thermocouple contact having a side length of 2 m. This shows that it is possible to assume that the temperature of the thermocouple contacts is constant only if the probe tip thermocouple contacts are very small in size.

6 and 7 and FIG. 4 are as follows. Applying an alternating current to terminals 1-2 corresponds to applying an alternating current to heavy doped silicon 10 and conductive diamond thin film 30. Detecting a signal at terminals 1-2 corresponds to detecting a thermoelectric voltage at the thermocouple contacts 40 of the heavily doped silicon 10 and the conductive diamond thin film 30. Detecting a signal at terminal 2-3 corresponds to detecting a thermoelectric voltage ΔV _tip at the contact 60 of the conductive diamond thin film 30 and the silicon sample surface 501. An important point here is that the difference between the signal detected at terminal 1-2 and the signal magnitude detected at terminal 2-3 will become smaller as the size of contact 60 becomes smaller. Therefore, the smaller the contact point 50 of the diamond probe 100 used in the present invention will be able to improve the quantitative measurement accuracy. Although the temperature distribution may be inconsistent even when the size of the thermocouple junction is expected to be small, about 100 nm, it is easy to measure the temperature distribution inside the thermocouple junction using a sample with a well-known thermoelectric coefficient such as platinum (Pt). As it can, it is easy to correct it.

Therefore, the impurity concentration of the silicon sample surface 501 and the temperature between the diamond thermoelectric probe 100 and the silicon sample surface 501 as well as from the thermoelectric voltage (ΔV _tip ) and the temperature value (ΔT _tip ) of the probe _tip 50 are determined. It is possible to quantitatively analyze types.

As described above, the current impurity concentration measurement method utilizing the scanning probe microscope (SPM) qualitatively measures the spatial distribution of impurity concentrations, but the semiconductor impurity concentration measurement method of the present invention has a higher spatial resolution. An important advantage is that the impurity concentration at the junction can be measured quantitatively.

In the method for measuring the concentration of semiconductor impurity according to the present invention, first, since the thermoelectric coefficient of the contact surface between the diamond probe and the silicon sample is measured, the measured signal has a constant value irrespective of the contact surface size, so that the signal is measured as the capacitance. Scanning Spreading Resistance MicroScopy (SSRM) has a high spatial resolution compared to Scanning Capacitance MicroScopy (SCM), where the size of P is proportional to the area of the contact surface. Second, the measured signal size is affected by the size of the probe and silicon sample contact surface. It is superior in quantitative measurement, and finally, spatial resolution is expected to be close to SThEM (Scanning Thermo-Electric MicroScopy), which is much better in terms of practicality, simplicity and quantitative measurement.

Claims (12)

A diamond thermoelectric probe in which conductive diamond is externally laminated to form a thermocouple with heavy doped silicon; An AC power source for locally heating the thermocouple contact unit by applying an alternating current to the thermoelectric probe; A first voltage measuring device connected to the thermocouple probe to measure a thermocouple voltage (ΔV _tip ) by a temperature change ΔT _junc occurring at the thermocouple contact portion; And And a second voltage measuring device connected to the diamond thin film and the semiconductor sample to measure a thermoelectric voltage (ΔV _tip-sample ) to derive a thermoelectric coefficient on the surface of the semiconductor. 2. The method of claim 1, wherein the conductive diamond probe comprises: heavy doped conical silicon; A silicon dioxide (S _i O ₂ ) thin film which is an insulator laminated on an outer surface except for an end portion of the silicon; And A diamond thin film laminated on the outside of the silicon dioxide thin film and the silicon end and forming a tip of a probe; Impurity concentration measuring apparatus of a semiconductor, characterized in that consisting of. A first step of forming a nano-contact by contacting a thermocouple probe having a conductive diamond thin film deposited on a sample surface, and then locally generating a thermocouple contact by applying an alternating current to the thermocouple of the probe; After measuring the thermoelectric voltage (ΔV _tip ) generated by the temperature fluctuation (ΔT _junc ) of the thermocouple contact portion by the heat generation with a first voltage measuring device, the thermoelectric voltage (ΔV _tip ) and the thermoelectric coefficient (S _si between silicon and diamond) a second step of calculating the temperature variation ΔT _tip of the probe tip using _-dia ); After measuring the thermoelectric voltage (ΔV _tip-sample ) of the probe contact formed between the tip of the conductive diamond thin film and the semiconductor sample surface with a second voltage measuring device, the thermoelectric voltage (ΔV _tip-sample ) and the temperature fluctuation (ΔT _tip) A third step of calculating a thermoelectric difference (S _dia-sub ) between the diamond thin film and the sample surface using; And A fourth step of calculating a sample after thermal coefficient _(sample S) of the surface, quantitative analysis of the impurity concentration in the semiconductor from the calculated thermal coefficient _(sample S) by using the thermal transfer coefficient (S _dia-sub); Impurity concentration measurement method of a semiconductor, characterized in that consisting of. The thermoelectric voltage (ΔV _tip ) in the second step is a thermoelectric voltage of 2ω frequency generated by the Seebeck effect calculated by separating the driving voltage of frequencyω from the measured thermoelectric voltage. A method for measuring the impurity concentration of a semiconductor. 4. The method of claim 3, wherein the thermoelectric coefficient (S _si-dia ) between the silicon and diamond thin film in the second step is determined by experiment. 4. The method of claim 3, wherein the thermoelectric coefficient (S _si-dia ) between the silicon and diamond thin film in the second step is obtained by the following equation. [Equation] S _{si - dia} = S _si -S _dia Where S _si is the thermoelectric coefficient of silicon and S _dia is the thermoelectric coefficient of the diamond film. The impurity concentration measurement of the semiconductor according to claim 6, wherein the thermoelectric coefficient S _si of the silicon and the thermoelectric coefficient S _dia of the diamond thin film are calculated by the following equations according to the concentration and type of the impurity, respectively. Way. [Equation]

(Wherein n and p are the electron and hole concentrations, and μ _n μ _p is the movement of electrons and holes even, S _n and S _p is Seebeck coefficient of electron and major The method according to claim 7, wherein S _n and S _p is the impurity concentration measuring method of a semiconductor, characterized by the following formula. [Equation]

,

Where k _B is the Boltzmann constant, e is the charge, r _e and r _h are the scattering coefficients for electrons and holes, and N _c and N _v are the conduction and balance coefficients for electrons and holes being) According to claim 3, wherein the temperature variation in the probe advanced from step 2 (ΔT _tip) is measured how the impurity concentration of the semiconductor, characterized in that to obtain, by the equation below on. [Equation]

Where ΔV _tip is the thermoelectric voltage at the thermocouple contact and S _si-dia is the thermoelectric coefficient between the silicon and diamond films. The thermoelectric voltage (ΔV _{tip - sample} ) in the third step is a thermoelectric voltage having a frequency of 2ω caused by a Seebeck effect calculated by separating the driving voltage of frequencyω from the measured thermoelectric voltage. Method for measuring the impurity concentration of the semiconductor, characterized in that. The method of claim 3, wherein the thermoelectric coefficient (S _dia-sample ) between the diamond thin film and the sample surface in the third step is obtained by the following equation. [Equation]

Where ΔV _tip-sample is the thermoelectric voltage at the probe junction, the contact point between the conductive diamond film and the semiconductor sample surface, and ΔT _tip is the temperature variation of the probe _tip . The semiconductor according to claim 3, wherein the thermoelectric coefficient (S _sample ) of the sample surface in the fourth step is obtained by separating the thermoelectric coefficient (S _dia ) of the diamond film from the calculated thermoelectric coefficient (S _dia-sample ). Method for measuring impurity concentration of

Images