Contactless LSI junction leakage testing method

Abstract

An inductively coupled oscillator method for inducing eddy currents in a semiconductor pn junction wafer while irradiating said wafer with pulsed light of selected intensity. The oscillator loading due to the pulsed light modulated eddy current losses is monitored and displayed on an oscilloscope in the form of a decay time plot of voltage amplitude, the plot being a function of the pulsed light intensity and the recombination rate of light-induced electrons and holes on each side of the junctions. The leakage characteristics of the junctions which are desired to be measured are one of the factors determining said rate. Leakage characteristic is made the predominent factor by setting the intensity of the pulsed light to a value which produces a nearly straight line decay time plot on the oscilloscope display. The slope of the line then is a measure of the leakage characteristic.

Description

BACKGROUND OF THE INVENTION

The invention generally relates to methods for testing for the presence of imperfections in semiconductor specimens and, more particularly to a contactless method for measuring PN junction leakage.

DESCRIPTION OF THE PRIOR ART

Contactless methods for measuring the decay time of bulk semiconductor materials are known in the art. For example, the paper "Contactless Measurement of Resistivity of Slices of Semiconductor Materials" by Nobuo Miyamoto et al, Review of Scientific Instruments, Vol. 38, No. 3, March 1967, page 360, discloses a high frequency capacitively coupled technique whereas the paper "Simple Contactless Method for Measuring Decay Time of Photoconductivity in Silicon" by R. M. Lichtenstein et al, Review of Scientific Instruments, Vol. 38, No. 1, January 1967, page 133, deals with a high frequency inductively coupled technique. Both techniques monitor the amplitude of oscillation of a high frequency carrier which is capacitively or inductively coupled, respectively, to the semiconductor sample while the sample is irradiated with pulsed light. Each pulse of light excites electrical carriers which temporarily increase the loading on the high frequency oscillator and cause a corresponding temporary decrease in the amplitude of oscillations. When the light pulse terminates, the amplitude of oscillations returns to its steady state value at a rate determined by the carrier recombination rate of the irradiated sample.

A more complicated situation exists when one or more junctions are present in the sample under test. High frequency oscillations which are capacitively coupled into the irradiated PN junction-containing specimen are substantially uneffected by the presence of the junctions. That is, the presence of the junctions does not significantly change the loading on the capacitively coupled high frequency oscillator. On the other hand, an inductively coupled high frequency oscillator experiences significant change in loading due to the presence of junctions in the semiconductor specimen under test. However, if the intensity of the pulsed light irradiating the specimen is maintained at relatively high values, consistent with prior art intensity levels, the loading of the inductively coupled oscillator attributable to the leakage associated with the PN junctions is not accurately detectable.

SUMMARY OF THE INVENTION

PN junction leakage in a semiconductor specimen including large scale integration specimens is measured by inductively coupling high frequency oscillations to the specimen while the specimen is subjected to pulsed light of selected intensity. The capacitance of the junction or junctions in the sample is charged to an amount dependent on the intensity of the pulsed light. When each light pulse terminates, the junction capacitance discharges at a rate determined by the existing discharge path impedance.

Each junction can be equivalently represented by a leakage resistor and a rectifying diode connected in shunt across a junction capacitor. The impedance of the diode is small relative to the leakage impedance when the diode is conducting. The impedance of the diode is high relative to the leakage impedance when the diode is not conducting, i.e., when the forward voltage across the diode is insufficient to overcome the conduction threshold.

It has been found that when the intensity of the pulsed light irradiating the junction specimen is set at a value insufficient to charge the junction capacitance to an amount causing forward conduction of the junction diode, the discharge of the junction capacitance following the termination of the pulse light is determined substantially solely by the leakage impedance of the junction. Such a setting of the light intensity is achieved in accordance with the present invention by observing a decay time plot of the discharge of the junction capacitance and reducing the amplitude of the light pulse until the ratio of the initial slope of the decay time plot relative to the terminal slope of the decay time plot portion attributable to leakage discharge is less than about 2:1.

BRIEF DESCRIPTION OF THE DRAWING

FIG. 1 is a simplified schematic circuit diagram of a preferred embodiment of the invention;

FIG. 2 is a cross-sectional view of a PN junction semiconductor specimen irradiated by pulsed light;

FIG. 3 is a representative superimposed series of decay time plots produced by the apparatus of FIG. 1;

FIG. 4 shows one of the plots of FIG. 3 in its entirety; 258 28 wafers containing PN junctions. Each wafer was subjected to a junction breakdown voltage test and to the decay time trace test of the present invention. Each dot in the plot of FIG. 5 represents the breakdown voltage observed on a given wafer and the time designated "lifetime" for the decay trace obtained from the same wafer to fall to 37 1/2% of its initial (peak) amplitude following the cessation of a light pulse. The experimental results of FIG. 5 can be more easily appreciated by reference to the plot of FIG. 6 which summarizes the data represented in FIG. 5.

Each point represented by an X in FIG. 6 represents the average value of the points of FIG. 5 surrounding a respective "lifetime" value. For example, the 10 points plotted around the "lifetime" value of 100 microseconds are averaged and plotted on FIG. 6 as point 40. Similarly, the median value of the same ten dots of FIG. 5 are represented by the single circle 41 of FIG. 6. The other points on FIG. 6 are plotted in a similar manner. In the case of point 42 and 43, the median and average values coincide. The number of samples represented by each of the points on FIG. 6 is shown below the respective "lifetime" value. Measurements on a total of 258 wafers are represented in the plots of FIGS. 5 and 6. The plotted data shows that the junction leakage measurement provided in accordance with the present invention correlates well with premature junction breakdown voltage, the latter of which is recognized in the art as an indicator of "pipe" defects in transistor wafers.

It will be observed that the technique of the present invention averages the junction leakage behavior of all junctions on a given LSI wafer without requiring any physical contact to the wafer. The non-destructive test nature of the technique allows its use on product wafers at many different times during their fabrication as desired.

While the invention has been particularly shown and described with reference to the preferred embodiment thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

Claims

1. A method for testing pn junctions in a wafer comprising:

irradiating said wafer with pulsed light of selected intensity to charge the capacitance of the junction in a direction tending to render said junction forwardly conductive,

said junction having a forward conduction threshold,

inductively coupling to said wafer high frequency oscillations for inducing eddy currents in said wafer,

said oscillations becoming amplitude modulated each time said wafer receives a pulse of said light, and

monitoring said amplitude modulation while varying the intensity of said pulsed light to determine a value of said pulsed light which charges said capacitance to a value beneath said threshold.

2. The method defined in claim 1 wherein said amplitude modulation is monitored while varying the intensity of said pulsed light to determine the maximum value thereof which charges said capacitance to a value beneath said threshold.

3. The method defined in claim 1 wherein said amplitude modulation is monitored by producing an oscilloscope decay time trace related to said amplitude modulation while varying the intensity of said pulsed light to determine a value thereof which causes the ratio of the initial slope of said trace to the terminal slope of said trace to be no greater than about 2:1.

4. The method defined in claim 1 wherein said amplitude modulation is monitored by producing an oscilloscope decay time trace related to said amplitude modulation while varying the intensity of said pulsed light to determine the maximum value thereof which causes the ratio of the initial slope of said trace to the terminal slope of said trace to be no greater than about 2:1.

5. A pn junction wafer manufacturing process wherein the method defined in claim 1 is utilized at a plurality of different times during said process.