Temperature abnormality detection method and semiconductor manufacturing apparatus

[0023] In the present invention, the management range corresponding to the parameter behavior in the transient gradient state, specifically, the management range corresponding to the temperature variation of the hot plate at placement of the article to be processed above the hot plate is calculated based on the temperature data of the hot plate, the process recipe data, and the elevation timing data for the article to be processed, and apparatus abnormality is detected with the use of the management range. Thus, detailed evaluation is enabled as to whether the temperature variation of the hot plate detected by the temperature control section in the placement of the article to be processed above the hot plate is different from the behavior of the reference temperature data (for example, temperature data measured in the past) or not. When the detected temperature variation of the hot plate is different from the behavior of the reference temperature data, for example, when the detected temperature of the hot plate does not vary so much compared with the behavior of the reference temperature data, it is judged as occurrence of apparatus abnormality, for example, abnormality that the article to be processed conveyed above the hot plate is placed at the predetermined wafer placing position inaccurately. Hence, by setting the management range per unit period for the parameter behavior showing a transient gradient, the detection, judgment, and classification of temperature abnormality in the semiconductor manufacturing apparatus can be performed more precisely than the conventional case where the “threshold value” is set for the peak value of the temperature variation of the hot plate.

[0024] As described above, the present invention relates to a method for detecting temperature abnormality in manufacturing a semiconductor device, and exhibits an effect that detection, judgment, and classification of temperature abnormality can be performed with high precision in a case applied to, for example, temperature management of a hot plate used in a photolithography process and the like, which is very effective.

However, when it is desired to detect abnormality of a heat treatment apparatus more precisely, it is insufficient to merely set the “threshold value” for the peak value of the temperature variation of the hot plate as in the conventional technique.

In detail, the method disclosed in Patent Document 1 attains not so precise abnormality detection in a heat treatment apparatus, and it is difficult to precisely manage the apparatus.