Working process end point real time determination method

A working process end point real time determination method, comprising:

• a first step of averaging a working process measurement signal obtained from a working process measuring instrument and indicating the progress of a working process at intervals equal to an integral number of times a predetermined period to discretely calculate an average value for each interval equal to the integral number of times the predetermined period as averaged data;
• a second step of successively comparing a working process elapsed time up to the present with a predetermined time set in advance and waiting that the working process elapsed time up to the present exceeds the predetermined time set in advance;
• a third step of waiting that the averaged data of the working process measurement signal calculated in the first step reaches a predetermined value or a predetermined multiple;
• a fourth step of calculating an average gradient over a plurality of ones of the averaged data of the working process measurement signal calculated in the first step which belong to a predetermined period in the past including the value of the averaged data at the present measurement point of time;
• an eighth step of waiting that an absolute value of the average gradient of the averaged data of the working process measurement signal calculated in the fourth step becomes equal to or higher than a first threshold value set in advance;
• and
• a ninth step of calculating a short time gradient of the averaged data of the working process measurement signal only from the value at the present measurement point of time and the value in the nearest past of the averaged data of the working process measurement signal calculated in the first step after the absolute value of the average gradient of the averaged data of the working process measurement signal calculated in the fourth step becomes equal to or higher than the first threshold value and determining that the working process has come to an end if the short time gradient reaches a value within a second threshold value set in advance successively more than a predetermined number of times, if the short time gradient assumes a value within the second threshold value totally by more than the predetermined number of times after the absolute value of the average gradient becomes equal to or higher than the first threshold value or if the ratio at which the short time gradient assumes a value within the second threshold value is equal to or higher than a predetermined ratio.

A working process end point real time determination method, comprising:

• a first step of averaging a working process measurement signal obtained from a working process measuring instrument and indicating the progress of a working process at intervals equal to an integral number of times a predetermined period to discretely calculate an average value for each interval equal to the integral number of times the predetermined period as averaged data;
• a second step of successively comparing a working process elapsed time up to the present with a predetermined time set in advance and waiting that the working process elapsed time up to the present exceeds the predetermined time set in advance;
• a third step of waiting that the averaged data of the working process measurement signal calculated in the first step reaches a predetermined value or a predetermined multiple;
• a fourth step of calculating an average gradient over a plurality of ones of the averaged data of the working process measurement signal calculated in the first step which belong to a predetermined period in the past including the value of the averaged data at the present measurement point of time;
• a 23rd step of multiplying a maximum value, a minimum value or an average value of the averaged data obtained for a time between a point of time when a fixed time for excepting a signal variation in an initial stage of the working process elapses and another point of time when the working process measurement signal begins a great variation by different predetermined values to calculate relative values as first and second threshold values;
• a 24th step of joining the average gradient value at the present measurement point of time and another one of the average gradient values in the near past after a point of time when an absolute value of the average gradient of the averaged data becomes equal to or higher than the first threshold value to perform extrapolation to the future to calculate a time at which the average gradient becomes equal to zero in the future as an estimated value of a working process end time;
• a 25th step of subtracting a delay time caused by use of the data in the past from the estimated value of the working process end time and temporarily determining that the working process has come to an end if the time obtained by the subtraction indicates a time prior to the present measurement point of time;
• and a 26th step of calculating a short time gradient of the averaged data only from a value of the averaged data at the present measurement point of time and another value of the averaged data in the nearest past and determining that the working process has come to an end from a logical AND between when the short time gradient assumes a value within the second threshold value successively more than a predetermined number of times, that the short time gradient assumes a value within the second threshold value totally by more than the predetermined number of times after the absolute value of the average gradient assumes a value equal to or higher than the first threshold value or that the ratio at which the short time gradient assumes a value within the second threshold value exceeds a predetermined ratio and the determination result in the 25th step.

A working process end point real time determination method, comprising:

• a first step of averaging a working process measurement signal obtained from a working process measuring instrument and indicating the progress of a working process at intervals equal to an integral number of times a predetermined period to discretely calculate an average value for each interval equal to the integral number of times the predetermined period as averaged data;
• a second step of successively comparing a working process elapsed time up to the present with a predetermined time set in advance and waiting that the working process elapsed time up to the present exceeds the predetermined time set in advance;
• a third step of waiting that the averaged data of the working process measurement signal calculated in the first step reaches a predetermined value or a predetermined multiple;
• a fourth step of calculating an average gradient over a plurality of ones of the averaged data of the working process measurement signal calculated in the first step which belong to a predetermined period in the past including the value of the averaged data at the present measurement point of time;
• a 23rd step of multiplying a maximum value, a minimum value or an average value of the averaged data obtained for a time between a point of time when a fixed time for excepting a signal variation in an initial stage of the working process elapses and another point of time when the working process measurement signal begins a great variation by different predetermined values to calculate relative values as first and second threshold values;
• an eighth step of waiting that an absolute value of the average gradient of the averaged data becomes equal to or higher than the first threshold value;
• and
• a ninth step of calculating a short time gradient of the averaged data of the working process measurement signal only from the value at the present measurement point of time and the value in the nearest past of the averaged data of the working process measurement signal calculated in the first step after the absolute value of the average gradient of the averaged data of the working process measurement signal calculated in the fourth step becomes equal to or higher than the first threshold value and determining that the working process has come to an end if the short time gradient reaches a value within the second threshold value successively more than a predetermined number of times, if the short time gradient assumes a value within the second threshold value totally by more than the predetermined number of times after the absolute value of the average gradient becomes equal to or higher than the first threshold value or if the ratio at which the short time gradient assumes a value within the second threshold value is equal to or higher than a predetermined ratio.

A working process end point real time determination method for a CMP apparatus for performing chemical and mechanical polishing of a semiconductor wafer, comprising:

• a tenth step of calculating, as measured light amount averaged data, an average value after each predetermined interval of time of a reflected light amount measured by a polished condition monitoring apparatus of the reflected light amount measurement type mounted on said CMP apparatus for irradiating inspection light upon a polished face of the semiconductor wafer and monitoring a polished condition from a variation of the reflected light amount obtained then;
• an eleventh step of calculating, as average gradient data, an average rate of change over a plurality of measured light amount averaged data which belong to a predetermined period in the past including the value at the present measurement point of time of the measured light amount averaged data calculated in the tenth step;
• a twelfth step of joining, after an absolute value of the average gradient of the measured light amount averaged data calculated in the eleventh step assumes a value equal to or higher than a predetermined value, the average gradient at the present measurement point of time and another one of the average gradients in the near past to perform extrapolation to the future to estimate and calculate a time at which the relationship that the average gradient is equal to 0 is reached;
• a thirteenth step of subtracting a polishing time till the present since the oldest point of time in the past used for calculation of the average gradient in the eleventh step from the time at which the relationship that the average gradient is equal to 0 is reached to calculate an end point determination time;
• and
• a fourteenth step of determining a polishing end point based on the end point determination time calculated in the thirteenth step and the present polishing time.

A working process end point real time determination method, comprising:

• a tenth step of calculating, as measured light amount averaged data, an average value after each predetermined interval of time of a reflected light amount measured by a polished condition monitoring apparatus mounted on a CMP apparatus for performing chemical and mechanical polishing of a semiconductor wafer;
• an eleventh step of calculating, as average gradient data, an average rate of change over a plurality of measured light amount averaged data which belong to a predetermined period in the past including the value at the present measurement point of time of the measured light amount averaged data calculated in the tenth step;
• a twelfth step of joining, after an absolute value of the average gradient of the measured light amount averaged data calculated in the eleventh step assumes a value equal to or higher than a predetermined value, the average gradient at the present measurement point of time and another one of the average gradients in the near past to perform extrapolation to the future to estimate and calculate a time at which the relationship that the average gradient is equal to 0 is reached;
• a thirteenth step of subtracting a polishing time till the present since the oldest point of time in the past used for calculation of the average gradient in the eleventh step from the time at which the relationship that the average gradient is equal to 0 is reached to calculate an end point determination time;
• a 31st step of multiplying a maximum value or an average value of the measured light amount averaged data obtained for a time after a fixed time for excepting a signal variation in an initial stage of the polishing elapses until the measured light amount averaged data begins to drop a great amount by a predetermined value to calculate a relative value as a threshold value;
• a 32nd step of calculating a first average value of a plurality of ones of the measured light amount averaged data which belong to a predetermined period in the past including the value of the measured light amount averaged data at the present measurement point of time and a second average value of another plurality of ones of the measured light amount averaged data which belong to another predetermined period in the further past and calculating a variation amount per unit time between the first and second average values as an average gradient in a short time;
• a 33rd step of detecting that the average gradient data calculated in the eleventh step assumes a minimum value;
• a 34th step of counting an accumulated value of the number of times by which the average gradient data in the short time calculated in the 32nd step assumes a value equal to or higher than the counter threshold value after it is determined that the average gradient data calculated in the eleventh step assumes a minimum value;
• a 35th step of resetting, if the average gradient data calculated in the eleventh step assumes a maximum value after it is determined in the 33rd step that the average gradient data assumes a minimum value, the accumulated value counted in the 34th step and repeating the minimum value detection in the 33rd step;
• and
• a 36th step of comparing the end point determination time and the present polishing time with each other and determining that the polishing end time comes if the end point determination time is equal to or smaller than the present polishing time and the number of times by which the average gradient in the short time counted becomes within the predetermined threshold value is equal to or greater than a predetermined number.