JP6044521B2 - Semiconductor wafer characteristic measurement system - Google Patents

Description

  The present invention relates to an apparatus for measuring characteristics of a semiconductor wafer, and more particularly to an apparatus for measuring electrical characteristics of a semiconductor wafer using a prober using mercury as an electrode.

Conventionally, a mercury prober for measuring electrical characteristics of a semiconductor using mercury as an electrode is known. For example, Patent Document 1 discloses that the electrical characteristics of a semiconductor wafer that can measure the electrical characteristics of the semiconductor wafer without causing contamination and surface defects on the wafer by using a head contact mercury probe. A mercury prober for measuring mechanical properties is disclosed.
In the apparatus described in Patent Document 1, mercury is stored in a mercury reservoir, and the mercury is guided so as to come into contact with the semiconductor wafer from above using a conduit called a capillary tube.

Patent Document 2 discloses a mercury prober that holds a semiconductor wafer with its main surface facing downward, guides mercury used as an electrode upward from below, and contacts the semiconductor wafer.
In the apparatus described in Patent Document 2, mercury is stored in a storage container called a supply reservoir, goes up through a conduit called a siphon tube, and comes into contact with a semiconductor wafer. .

  By using the mercury prober of Patent Document 1 or Patent Document 2, electrical characteristics such as dopant concentration, resistivity, oxide film breakdown voltage, lifetime, and the like can be measured. The semiconductor to be measured is, for example, silicon, GaP, GaN, SiC, or the like.

  As a method for measuring the resistivity of a semiconductor, for example, a method for measuring CV (capacitance-voltage) characteristics is known. In order to measure the CV characteristics, a Schottky junction is formed on the surface of a semiconductor wafer as a sample, and a depletion layer is expanded inside the semiconductor wafer by applying a reverse bias voltage while changing the capacitance stepwise. To change. When the relationship between the reverse bias voltage and the capacitance is plotted on a graph, CV characteristics can be obtained.

  Furthermore, since the dopant concentration at the desired depth can be calculated from the reverse bias voltage and the capacitance, a dopant concentration profile in the depth direction in the semiconductor wafer can be obtained.

  When the measurement depth is specified in the dopant concentration profile in the depth direction in the semiconductor wafer, the dopant concentration at that depth is obtained. Further, the dopant concentration can be converted into resistivity by converting the obtained dopant concentration by a conversion formula such as ASTM STANDARDDS F723.

  Various proposals have been made to measure the resistivity of a silicon single crystal wafer using a mercury prober. For example, Non-Patent Document 1 discloses a chemical pretreatment method for p-type and n-type wafers. Furthermore, Patent Document 3 discloses that a carbon-containing compound is attached to the surface of a p-type or n-type epitaxial layer and then irradiated with ultraviolet rays in an oxygen-containing atmosphere.

  Further, in Patent Document 3, when a silicon single crystal wafer to be a sample is n-type, the surface of the wafer is oxidized in advance to form a thin silicon oxide film, and a mercury electrode is brought into contact with the silicon oxide film. It is disclosed that CV characteristics can be measured.

  As a method for forming a thin silicon oxide film on the surface of a silicon single crystal wafer, for example, in Patent Document 4, the surface of a silicon single crystal is oxidized with ozone gas by exposing the silicon single crystal wafer to ultraviolet light in an oxygen-containing atmosphere. A method is disclosed.

  Patent Document 5 proposes that a light-shielding plate that allows ozone gas to pass therethrough and shields ultraviolet rays is disposed between the ultraviolet light source and the wafer when oxidized with ozone gas.

Japanese Patent Laid-Open No. 6-140478 US Pat. No. 7,253,649 JP 2007-158314 A Special table 2002-516486 gazette JP 2013-46030 A

ASTM Standards F1392-02

  However, the reproducibility of electrical characteristics obtained using a mercury prober is poor. For example, when measuring the resistivity of a silicon single crystal wafer, the coefficient of variation CV (standard deviation σ / average value x-bar) is 0.5% or less. It is quite difficult to control. Here, CV = 0.5% is the maximum value of the coefficient of variation necessary for guaranteeing the standard center ± 15%.

  The present invention has been made in view of the above problems, and characteristics of a semiconductor wafer in which electrical characteristics with improved reproducibility can be obtained when the electrical characteristics of a semiconductor wafer are repeatedly measured using a mercury electrode. It aims at providing a measuring device.

  In order to achieve the above object, the present invention is a semiconductor wafer characteristic measuring device for measuring characteristics of a semiconductor wafer by contacting the semiconductor wafer with mercury as an electrode, the mercury reservoir containing the mercury, A device for measuring characteristics of a semiconductor wafer, comprising a conduit for guiding mercury from the mercury reservoir to a contact portion with the semiconductor wafer, wherein the mercury reservoir and at least one of the conduits are grounded. provide.

  As described above, by grounding at least one of the mercury reservoir and the conduit, static electricity generated by contacting the mercury reservoir or the inner wall surface of the conduit when the mercury moves can be eliminated via the ground wire. Therefore, electrical characteristics with improved reproducibility can be obtained.

At this time, it is preferable to ground both the mercury reservoir and the conduit.
Thus, by grounding both the mercury reservoir and the conduit, static electricity can be more efficiently removed through the ground wire.

At this time, it is preferable that the semiconductor wafer is a silicon single crystal wafer, and the characteristic of the semiconductor wafer is resistivity.
The resistivity of the silicon single crystal wafer can be suitably measured using the above measuring apparatus.

Here, in the resistivity measurement of the semiconductor wafer, the variation coefficient CV of the repeated measurement defined by the standard deviation σ / average value x-bar can be 0.5% or less.
As described above, according to the apparatus of the present invention, the coefficient of variation CV of the repeated measurement can be 0.5% or less, and the resistivity measurement of the semiconductor wafer can be performed with higher accuracy.

  As described above, according to the semiconductor wafer characteristic measuring apparatus of the present invention, static electricity generated by contact with the mercury reservoir or the inner wall surface of the conduit when mercury moves can be eliminated through the ground wire. Therefore, it becomes possible to obtain electrical characteristics with improved reproducibility.

It is the schematic which shows an example of the characteristic measuring apparatus of the semiconductor wafer of this invention. It is the schematic which shows an example of the CV characteristic measurement system for measuring a resistivity. It is a schematic explanatory drawing which shows the relationship between static electricity and a depletion layer. It is a schematic explanatory drawing which shows the influence of the electrostatic charge and discharge at the time of CV characteristic measurement. It is an equivalent circuit diagram showing the relationship between the capacitance of a depletion layer formed by static electricity and the capacitance of a depletion layer formed by applying a reverse bias voltage. It is a schematic explanatory drawing which shows the state in which the static electricity at the time of CV characteristic measurement is not charged. It is the schematic which shows another example of the characteristic measuring apparatus of the semiconductor wafer of this invention. It is a graph which shows the measurement result of an Example. It is a graph which shows the measurement result of a comparative example.

  Hereinafter, the present invention will be described in detail as an example of an embodiment with reference to the drawings, but the present invention is not limited thereto.

  FIG. 1 is a schematic view showing an example of a semiconductor wafer characteristic measuring apparatus according to the present invention. A mercury probe is used for guiding mercury used as an electrode upward from below and contacting the semiconductor wafer. , Resistivity).

  In a semiconductor wafer characteristic measuring apparatus (mercury prober) 10 shown in FIG. 1, a semiconductor wafer (for example, a silicon single crystal wafer) 1 is held on a stage 11 with a main surface to be measured facing downward, and is placed above a mercury electrode. Placed. When measuring the resistivity, mercury 31 stored in a mercury reservoir 12 called a supply container rises through the conduit 13 and contacts the main surface of the silicon single crystal wafer 1 to form a Schottky junction. .

  The mercury reservoir 12 is connected to the ground wire 72 through a conductive rubber sheet 71 that covers the mercury reservoir 12. The conduit 13 is connected to the ground wire 74 through a conductive rubber sheet 73 covering the conduit 13. The conductive rubber sheets 71 and 73 are mixed with a conductive material such as carbon or silver and have a low resistivity of about 0.008 Ω · cm, so that static electricity generated in the mercury reservoir 12 and the conduit 13 can be prevented. It can be easily absorbed and escaped to the ground wires 72 and 74.

  The semiconductor wafer characteristic measuring apparatus 10 of FIG. 1 can apply a voltage by a power source 14 provided between a stage 11 and a mercury electrode 31, and further, by connecting a capacitance meter 15 and a voltmeter 16. , Capacitance C and voltage V can be measured.

  The CV characteristic measurement system 100 for measuring resistivity shown in FIG. 2 includes a mercury probe 10 that contacts the silicon single crystal wafer 1 using mercury 31 as an electrode, and measurement cables 61 and 62 connected to the mercury probe 10. An LCR meter 40 that forms a depletion layer by applying a reverse bias voltage to the semiconductor wafer 1 and measures the capacitance of the depletion layer, and a GPIB (General Purpose Interface Bus) cable to the LCR meter 40 The PC 50 has a PC (Personal Computer) 50 connected thereto, and the PC 50 is installed with analysis software for calculating the resistivity from the reverse bias voltage and the capacity of the depletion layer.

  When measuring the resistivity of the silicon single crystal wafer 1, the reverse bias voltage V is stepped by using the LCR meter 40 on the mercury electrode formed with the silicon single crystal wafer 1 placed on the mercury probe 10 and forming a Schottky junction. The capacitance C is measured by the LCR meter 40 while changing the capacitance C by expanding the depletion layer inside the silicon single crystal wafer 1. Then, using the analysis software installed in the PC 50, the resistivity is calculated from the relationship between the reverse bias voltage V and the capacitance C of the depletion layer.

  When the mercury 31 used as an electrode is contaminated by particles adhering to the main surface of the silicon single crystal wafer 1 every time it comes into contact with the silicon single crystal wafer 1, the resistivity of the mercury 31 gradually increases and the electrical characteristics are measured. The value will be adversely affected. In order to prevent this, the mercury 31 in contact with the semiconductor wafer 1 is temporarily returned to the mercury reservoir 12, and the entire mercury 31 is bubbled and cleaned, and then supplied again as an electrode.

  When this bubbling is performed, the mercury 31 rubs against the inner wall of the mercury reservoir 12 and generates static electricity. When the mercury reservoir 12 is made of polycarbonate, for example, it is negatively charged. On the other hand, mercury 31 is charged to + (plus).

  Further, when the mercury 31 rises through the conduit 13, the mercury 31 rubs against the inner wall of the conduit 13 and generates static electricity. When the conduit 13 is made of, for example, polytetrafluoroethylene, it is charged to-(minus). On the other hand, mercury 31 is charged to + (plus).

  The static electricity generated in the mercury reservoir 12 or the conduit 13 is lost to the ground wires 72 and 74 via the conductive rubber sheets 71 and 73, and thus immediately disappears, and the static electricity of the mercury 31 also disappears accordingly.

  If, for example, the mercury 31 is positively charged and contacts the main surface of the p-type silicon single crystal wafer 1, positive static electricity and the p-type carrier repel each other as shown in FIG. Before starting to apply bias voltage V, depletion layer 32 is formed in the vicinity of the main surface.

  When the resistivity is measured by CV characteristic measurement in a state where the depletion layer 32 is formed, as shown in FIG. 4A, the depletion layer 33 formed by applying a reverse bias voltage and the static electricity The depletion layers 32 to be formed are connected in series. When the static electricity of the mercury 31 is discharged and decreases with the passage of time, the width of the depletion layer 32 formed by the static electricity is changed from d1 to d2, as shown in FIGS. 3B and 4B. Narrow.

Here, as shown in FIG. 5A (corresponding to FIG. 4A) and FIG. 5B (corresponding to FIG. 4B), the capacitance of the depletion layer 32 formed by static electricity is C32, If the capacitance of the depletion layer 33 formed by applying a reverse bias voltage is C33, these are connected in series, so the capacitance C of the entire depletion layer is
1 / C = 1 / C32 + 1 / C33 (1)
It is expressed.

At this time, when the static electricity of the mercury 31 is discharged over time and the width of the depletion layer 32 is reduced from d1 (FIG. 4A) to d2 (FIG. 4B),
C32 = Aε ₀ ε _Si / d (2)
Therefore, the capacity C32 increases. As the capacitance C32 increases, the capacitance C of the entire depletion layer increases from the equation (1). Here, A is the electrode area, ε ₀ is the vacuum dielectric constant, and ε _Si is the relative dielectric constant of Si.

Since the capacitance C of the entire depletion layer and the impurity concentration N (W) at the depth W have a proportional relationship of the equation (3),
N (W) = C ³ / (qε ₀ ε _Si A ² ) / (dC / dV) (3)
As the capacitance C of the entire depletion layer increases, the impurity concentration N (W) also increases. Then, the resistivity that is inversely proportional to the impurity concentration N (W) becomes low. Here, q is the charge amount of electrons.

    That is, if the resistivity of the p-type silicon single crystal wafer 1 is measured while the mercury 31 is charged positively, the measured resistivity is high during the initial period when a large amount of static electricity remains, and the time The measured resistivity decreases after the static electricity is discharged after elapse of time. For this reason, for example, if the time from when the mercury 31 contacts the silicon single crystal wafer 1 to when the measurement is started is not kept constant, the variation in the resistivity measurement result increases.

  On the other hand, in the mercury prober 10 of the present invention, static electricity generated in the mercury reservoir 12 or the conduit 13 is immediately discharged to the ground wires 72 and 74 through the conductive rubber sheets 71 and 73, and accordingly disappears. The static electricity of mercury 31 also disappears. As a result, as shown in FIG. 6, even when mercury 31 is brought into contact with the p-type silicon single crystal wafer 1, the depletion layer 32 caused by static electricity is not formed, but only the depletion layer 33 by the application of the reverse bias voltage V is formed. Therefore, the original capacity C can be measured.

In FIG. 1, both the mercury reservoir 12 and the conduit 13 are grounded by covering both the mercury reservoir 12 and the conduit 13 with a conductive rubber sheet, but either the mercury reservoir 12 or the conduit 13 is grounded. By covering with a conductive rubber sheet, if either one of the mercury reservoir 12 or the conduit 13 is grounded, the mercury 31 can be prevented from being charged.
If both the mercury reservoir 12 and the conduit 13 are grounded, charging of the mercury 31 can be prevented more efficiently.

When the relationship between the measured reverse bias voltage V and the capacitance C is plotted on a graph, a CV characteristic is obtained. Further, by substituting the reverse bias voltage and the capacity into the following formulas (4) and (5), for example, the depth W in the silicon single crystal wafer 1 and the dopant concentration N (W) at the depth W can be calculated. Therefore, a dopant concentration profile in the depth direction can be obtained.
W = Aε ₀ ε _Si / C (4)
N (W) = 2 / (qε ₀ ε _Si A ² ) * {d (C ⁻² ) / dV} ⁻¹ (5)

  When the measurement depth is specified in the dopant concentration profile in the depth direction, the dopant concentration at that depth is obtained. Moreover, the dopant concentration can be converted into the resistivity by converting the obtained dopant concentration by a conversion formula shown in ASTM STANDARDDS F723 or the like.

By measuring the dopant concentration or resistivity obtained in this way repeatedly at the same location of the silicon single crystal wafer 1 a plurality of times, for example, 10 times continuously, the reproducibility of the resistivity measurement can be changed by the coefficient of variation CV (σ / x -bar). The coefficient of variation CV is calculated using, for example, the following equation (6) by obtaining an average value x-bar and a standard deviation σ from 10 measured values. In the case of repeated measurement, the mercury electrode is once separated from the silicon single crystal wafer 1 for each measurement, and is joined again.
Coefficient of variation CV = (σ / x-bar) × 100 (%) (6)

  By using the mercury probe 10 of the present invention, the variation coefficient CV can be reduced to 0.5% or less.

  So far, the embodiment of the present invention has been described by exemplifying a mercury prober that guides mercury used as an electrode from below to contact the semiconductor wafer. However, the present invention is not limited to this, and the upper side of the semiconductor wafer is described above. The same can be applied to a mercury prober that guides mercury so that it comes into contact with the water.

FIG. 7 shows a mercury prober 20 that stores mercury 31 in a mercury reservoir 22 and guides the mercury 31 so as to come into contact with the semiconductor wafer 1 placed on the stage 21 from above using a conduit 23.
In FIG. 7, the mercury reservoir 22 is connected to the ground wire 82 via a conductive rubber sheet 81 covering the mercury reservoir 22, and the conduit 23 is connected to the ground wire via a conductive rubber sheet 83 covering the conduit 23. 84.
The semiconductor wafer characteristic measuring apparatus 20 of FIG. 7 can apply a voltage by a power source 24 provided between the stage 21 and the mercury electrode 31, and further, by connecting a capacitance meter 25 and a voltmeter 26. , Capacitance C and voltage V can be measured.

  When measuring the resistivity of the semiconductor wafer 1 using the mercury probe 20, the reverse bias voltage V is continuously changed on the mercury electrode formed with the Schottky junction with the semiconductor wafer (for example, silicon single crystal wafer) 1. The capacitance C is changed by expanding the depletion layer inside the semiconductor wafer 1 and measuring the capacitance C with an LCR meter, and the resistivity is calculated from the relationship between the reverse bias voltage V and the capacitance C of the depletion layer. .

  EXAMPLES Hereinafter, although an Example and a comparative example are shown and this invention is demonstrated more concretely, this invention is not limited to these.

(Example)
A p-type silicon epitaxial wafer having a diameter of 200 mm and a resistivity of 12 Ωcm is prepared, the mercury reservoir 12 is connected to the ground wire 72 via the conductive rubber sheet 71, and the conduit 13 is connected to the ground wire 74 via the conductive rubber sheet 73. Using the connected mercury probe 10 and the CV characteristic measurement system 100, the central portion of the p-type silicon epitaxial wafer was repeatedly measured 10 times. The results are shown in FIG.

The average value of the measurement results shown in FIG. 8 was 12.03 Ω · cm, and the standard deviation was 0.0209 Ω · cm.
The coefficient of variation CV (σ / x-bar) was 0.17%, which sufficiently satisfied the target of 0.5%.

(Comparative example)
The center of a p-type silicon epitaxial wafer having a diameter of 200 mm and a resistivity of 12 Ωcm, using the same mercury prober 10 and CV characteristic measurement system 100 as in the embodiment except that the earth line is not connected to the mercury reservoir 12 and the conduit 13. The part was repeatedly measured 10 times. The results are shown in FIG.

The average value of the measurement results shown in FIG. 9 was 12.31 Ω · cm, and the standard deviation was 0.1235 Ω · cm.
The coefficient of variation CV was 1.00%, which was higher than the target of 0.5%.

  The present invention is not limited to the above embodiment. The above-described embodiment is an exemplification, and the present invention has substantially the same configuration as the technical idea described in the claims of the present invention, and any device that exhibits the same function and effect is the present invention. It is included in the technical scope of the invention.

  According to the semiconductor characteristic measuring apparatus of the present invention, when the semiconductor is repeatedly measured using the mercury electrode, the coefficient of variation CV can be made 0.5% or less, so that the electrical characteristics guaranteeing a standard center of less than ± 15%. It becomes possible to apply to the measurement of.

1. Semiconductor wafer (silicon single crystal wafer),
10, 20 ... Semiconductor wafer characteristic measuring device (mercury prober),
11, 21 ... stage, 12, 22 ... mercury reservoir, 13,23 ... conduit,
14, 24 ... power supply, 15, 25 ... capacity meter, 16, 26 ... voltmeter, 31 ... mercury,
32 ... Depletion layer formed by static electricity, 33 ... Depletion layer by reverse bias voltage,
40 ... LCR meter, 50 ... PC,
61, 62 ... measurement cable, 63 ... GPIB cable,
71, 73, 81, 83 ... conductive rubber sheet, 72, 74, 82, 84 ... ground wire.

Claims (4)

A device for measuring characteristics of a semiconductor wafer for contacting the semiconductor wafer with mercury as an electrode and measuring the characteristics of the semiconductor wafer,
A mercury reservoir for storing the mercury;
A conduit for guiding the mercury from the mercury reservoir to a contact portion with the semiconductor wafer;
An apparatus for measuring characteristics of a semiconductor wafer, wherein both the mercury reservoir and the conduit are grounded. The apparatus for measuring characteristics of a semiconductor wafer according to claim 1, wherein the mercury reservoir and the conduit are grounded through a conductive rubber sheet. The semiconductor wafer is a silicon single crystal wafer;
3. The semiconductor wafer characteristic measuring apparatus according to claim 1, wherein the characteristic of the semiconductor wafer is resistivity. 4. The semiconductor wafer according to claim 3, wherein, in the resistivity measurement of the semiconductor wafer, a coefficient of variation CV of repeated measurement defined by standard deviation σ / average value x−bar is 0.5% or less. 5. Characteristic measuring device.

Images