Method of detecting and controlling in-situ faults in rapid...

Data processing: generic control systems or specific application – Specific application – apparatus or process – Chemical process control or monitoring system

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C219S638000, C219S490000, C219S502000, C165S096000, C438S011000, C438S795000, C505S310000

Reexamination Certificate

active

06577926

ABSTRACT:

FIELD OF THE INVENTION
The present invention broadly relates to rapid thermal processing systems, especially of the type employed in semiconductor manufacturing operations, and deals more particularly with a method of detecting and controlling in-situ faults related to failure of the system to operate within specified temperature limits.
BACKGROUND OF THE INVENTION
Rapid thermal processing (RTP) systems are used in semiconductor manufacturing operations to carry out several different processes, including rapid thermal oxidation and rapid thermal annealing. RTP systems have several advantages over conventional furnace systems. For example, one limitation of furnace systems employed to carry out oxidation processes is its inertia to temperature transition, resulting in a higher thermal budget. The thermal budget can be reduced considerably by increasing the duration of these transitions through the use of RTP. In the case of annealing processes wherein annealing removes defects introduced by ion implantation, the use of RTP systems provides a higher level of dopant activation and annealing effect, compared to conventional furnace systems.
In an RTP system, a semiconductor wafer is rapidly heated from a low temperature to a high processing temperature. It is held at this elevated temperature for a short time and then brought back rapidly to a low temperature. Typical temperature transition rates range from 10°/sec to 350°/sec, compared to about 0.1°/sec for furnace processing. RTP durations at high processing temperatures vary from one second to five minutes.
In order to control process parameters and particularly temperature, a temperature sensing device, such as an optical pyrometer is used to measure the temperature on the backside of the wafer, typically at a wave length, for example, of 0.95 &mgr;m.
RTP processing is important where precise thermal control and short high temperature process times are critical. When used to grow thin oxide films, the process is referred to as rapid thermal oxidation (RTO). Oxide layers with film thicknesses from 0.004 &mgr;m to 0.4 &mgr;m can be grown in pure oxygen at 900° C. to 1150° C. for a duration of only 15 to 180 seconds. In addition, the electrical characteristics of rapid thermal oxides have been found to be equivalent to or better than furnace-grown thermal oxides. RTO grown films on polycrystalline silicon exhibit electrical breakdown fields that approach those of oxides grown on single-silicon. RTP is also used to grow other insulating films, to activate implanted ions and form shallow junctions, to alloy contacts and form conducting fields, such as titanium-silicide and titanium-nitrides, and to reflow glass.
Rapid thermal annealing allows the removal of defects introduced by ion implantation, and activation of species with little movement of the diopants. While conventional furnaces are capable of supplying high temperature, the slow insertion and removal of the wafers and the heat capacity of the system require ramp up and ramp down times in the range of minutes, thus resulting in excessive movement of dopants. In contrast, during rapid thermal annealing, the entire wafer is heated uniformly in seconds and, after annealing, the wafer is cooled in seconds.
Typical RTP systems are integrated into semiconductor processing stations, such as CVD or PVD chambers where heating is provided by a multiplicity of heating elements, typically thermal generating lamps, spatially arranged into a plurality of heating zones. The heating elements are normally arranged so as to face both the front side and backside of the wafer. One or more temperature sensing devices, such as pyrometers or emissometers, are strategically placed so as to sense the temperature at one or more locations within the chamber. The temperature sensors are typically connected by optical fibers to a temperature recording system which records the temperature at multiple locations in the chamber throughout each processing cycle in which one or more wafers are processed.
Precise temperature control in RTP systems over the entire processing cycle may be critical to achieving acceptable processing results. This means that the temperature at each monitored location in the chamber must be maintained within certain limits during both temperature ramp up and ramp down sequences. When the temperature at one or more monitored locations is not maintained within desired limits, a “fault” occurs which may have a material effect on quality of process, and thus on the quality of the wafer.
In the past, it was not always possible to detect a fault during an RTP cycle. Consequently, the out-of-limit condition was not be detected, and appropriate steps could not be taken to correct the condition, until after one or more batches of wafers had been processed. Accordingly, the inability to detect and dynamically correct a temperature fault condition during an RTP cycle resulted in scrap rates, and therefore, reduced yields.
The task of accurately monitoring the temperature at multiple locations in the processing chamber is complicated by several factors. First, the emissivity of heat radiation from the wafer is significantly affected by the nature of the surface on the wafer backside. For example d-poly film grown by HTF on the wafer backside can form Newtonian ring-like concentric circles which have a constructive or destructive effect on the radiation waves impinging the wafer surface, thereby increasing or decreasing the magnitude of measured radiation. Second, inaccurate radiation measurements sometimes occur as a result of the presence of small broken fragments and residues of wafers remaining in the processing chamber from processing previous batches of wafers. The presence of these residual wafer fragments in the chamber alters emissivity measurements, thus giving rise to inaccurate readings which may prevent the detection of temperature control faults, and the initiation of dynamic controls needed to bring temperature parameters within proper limits.
Yet another factor affecting temperature control relates to degradation of certain components of the RTP system. Emissometers, optical fibers and pyrometers functionally decay in performance over their service lives, thus affecting their performance and accuracy. The decay in performance of these components is not easily detected, further complicating the task of detecting temperature faults and dynamically correcting control parameters.
Accordingly, there is a clear need in the art for a method of detecting in-situ, temperature control faults, and dynamically correcting temperature control parameters in RTP systems of the type described above.
SUMMARY OF THE INVENTION
According to one aspect of the present invention, a method is provided for detecting and controlling in-situ faults occurring in a rapid thermal processing system. The method comprises the steps of: (A) generating a set of data relating to the distribution of power applied to the heating elements in the zones; (B) converting the data in the data set into a data format representing the current collective state of the power distribution for all of the zones; (C) comparing the converted data set with a set of statistical data representing a standard: and, (D) changing at least one process control parameter affecting the thermal process based on the comparison.
The data conversion is preferably performed by determining the total power applied to all the lamps in all of the zones and producing a plurality of fractions by dividing the numerical values of the power applied to the lamps in each of the zones by the total power. The fractions are then arranged in sequence from smallest to largest numerical value following which a calculation is made of the slope of the sequentialized fractional values. The slope is preferably calculated by the least squares straight-line method. The data conversion step preferably includes weighting at least certain of the fractions based on data obtained from a statistical process control system. The method further includes the step of storing the s

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Method of detecting and controlling in-situ faults in rapid... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method of detecting and controlling in-situ faults in rapid..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of detecting and controlling in-situ faults in rapid... will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3137932

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.