Optics: measuring and testing – Inspection of flaws or impurities – Surface condition
Reexamination Certificate
2002-06-24
2004-12-21
Rosenberger, Richard A. (Department: 2877)
Optics: measuring and testing
Inspection of flaws or impurities
Surface condition
C250S2140AG
Reexamination Certificate
active
06833913
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates generally to inspection systems. More specifically, it relates to light collection mechanisms for inspecting semiconductor wafers and other types of patterned samples.
Conventional darkfield optical inspection tools locate defects on patterned wafers by scanning the surface of the wafer with a tightly focused laser spot and measuring the amount of light scattered by the illuminated spot on the wafer. Dissimilarities in the scattering intensity between similar locations in adjacent dies are recorded as potential defect sites.
The dynamic range of this optical scattering is typically substantial. Changes in scattering intensity of more than a million to one within a single die are not uncommon. This high dynamic range is intrinsic to the optical configuration of the instrument and the scattering properties of the wafers and defects of interest. Because this dynamic range is substantially greater than the reliable measurement range of existing instruments, inspection operators are forced to accept an unpleasant compromise between inspecting with too low a sensitivity in some portions of the die, and temporarily overloading the instrument's detection electronics in other regions.
In general, scanning the wafer with the smallest possible laser spot size maximizes sensitivity to defects by maximizing the spatial resolution of the scattering image. However, this increased resolution generally correlates with an increased pixel density within the light collectors or detectors to properly sample the image. The detectors typically include a sensor for detecting the scattered light and generating an analog signal based on such detected light and an analog-to-digital converter (ADC) for converting the analog detected signal into a digital detected signal. The digital detected signal may then be analyzed for defects. Since all the pixels are measured serially, and only a limited amount of time is available to scan each wafer, there is a fundamental relationship between the speed of the measurement electronics and the maximizing of the spatial resolution of the scattering is image. To enable high spatial resolution, higher bandwidth analog electronics and faster ADC's are often utilized.
In addition to maximizing the speed of the measurement electronics to thereby maximize spatial resolution, it is desirable to maximize the dynamic range of the light that is discernable by the measurement electronics. However, there is a fundamental trade-off in ADC's between speed and dynamic range. That is, dynamic range is typically limited by noise and offset errors, both of which tend to increase with speed.
Accordingly, there is a need for improved inspection mechanisms that are capable of quickly detecting light having a relatively high dynamic range.
SUMMARY OF THE INVENTION
Accordingly, mechanisms are provided for detecting a relatively wide dynamic range of intensity values from a beam (e.g., scattered light, reflected light, or secondary electrons) originating from a sample, such as a semiconductor wafer. In other words, the inspection system provides detected output signals having wide dynamic ranges. The detected output signals may then be analyzed to determine whether defects are present on the sample. For example, the intensity values from a target die are compared to the intensity values from a corresponding portion of a reference die, where a significant intensity difference may be defined as a defect.
In a specific embodiment, an inspection system for detecting defects on a sample is disclosed. The system includes a beam generator for directing an incident beam towards a sample surface and a detector positioned to detect a detected beam originating from the sample surface in response to the incident beam. The detector has a sensor for detecting the detected beam and generating a detected signal based on the detected beam and a non-linear component coupled to the sensor. The non-linear component is arranged to generate a non-linear detected signal based on the detected signal. The detector further includes a first analog-to-digital converter (ADC) coupled to the non-linear component, and the first ADC is arranged to digitize the non-linear detected signal into a first digitized detected signal. The system further includes a data processor for determining whether there is a defect present on the sample surface based on the first digitized detected signal.
In a further aspect, the system also includes a transformation mechanism for transforming the first digitized detected signal into a second digitized detected signal that compensates for noise variation associated with different intensity levels of the first detected output signal. The data processor is further arranged to receive the second digitized signal and the step of determining whether there is a defect is based indirectly on the first digitized detected signal by being based directly on the second digitized detected signal. In a further implementation, the transformation mechanism operates to cause a derivative of the second digitized detected signal to be equal to a normalization function, which is an estimate of the inverse of the noise level or uncertainty in the measurement. One such normalization function may be computed by dividing an average of an envelope function by the envelope function itself, the envelope function being calculated based on an observed repeatability of measurements of the first digitized detected signal.
In one aspect, the sensor is a photomultiplier tube (PMT). In other implementations, the sensor is an electron multiplier tube, a micro-channel plate PMT, an avalanche photodiode, a metal channel dynode PMT, a wire mesh dynode PMT, a PMT with explicit gate or grid electrodes, or an imaging array with programmable integration time.
In one aspect, the non-linear component is a logarithmic amplifier. In a further aspect, the detector further includes a first feed back circuit for automatically adjusting a sensor gain of the sensor based on the non-linear detected signal or the detected signal. In one embodiment, the first feed back circuit has a variable voltage supply component coupled to the non-linear component and arranged to adjust a voltage level of the sensor gain based on non-linear detected signal or the detected signal, a voltage reference signal, and one or more control signal(s). The first feed back circuit further includes an amplifier coupled to the variable voltage supply and arranged to amplify the sensor gain signal prior to it being input to the sensor.
In a further embodiment, the system further includes a second ADC for receiving the sensor gain, digitizing the sensor gain and outputting it as a digitized sensor gain signal. The system also has a first transformation mechanism for calibrating the digitized detected signal into a calibrated detected signal and a second transformation mechanisms for calibrating the digitized sensor gain signal into a calibrated gain signal. The system further includes an arithmetic logic unit (ALU) arranged to subtract the calibrated gain signal from the calibrated detected signal to form a first detected output signal. Preferably, the first and second transformation mechanisms take the form of a look-up table embodied within a memory device, but they may also be implemented as a mathematical equation which is evaluated by a digital computer, digital signal processor, or programmable logic device. The data processor is further arranged to receive the first detected output signal and the step of determining whether there is a defect is based indirectly on the first digitized detected signal by being based directly on the first detected output signal.
In yet a further aspect, the system includes an offset mechanism arranged to receive a user-selected sensor gain and offset the first detected output signal by a log of the user-selected gain to thereby emulate a programmable sensor gain, where the sensor gain is not altered. In a further aspect, the first and second transformation mechan
Benitez Eva L.
Chen Dongsheng (Don)
Greene John D.
Sullivan Jamie M.
Vella Eric N.
Beyer Weaver & Thomas LLP.
KLA-Tencor Technologies Corporation
Olynick, Esq. Mary Ramos
Rosenberger Richard A.
LandOfFree
Apparatus and methods for optically inspecting a sample for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and methods for optically inspecting a sample for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and methods for optically inspecting a sample for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3305452