The thickness of a sample is commonly measured by a capacitive thickness gauge. The measurement technique follows directly from the formula for capacitance. The capacitance of two parallel plates, separated by air, is directly proportional to the area of the plates and inversely proportional to the distance between the plates. When measuring the thickness of a silicon wafer, as in the RT-110, wafer acts as one plate of the capacitor, and a measurment electrode, held a fixed distance above the bottom surface of the wafer, acts as the second plate. Because the silicon in conductive, the thicker the wafer, the smaller the spacing between the “lower” plate and the upper plate.
In this case the plates are not of equal size. The effective size of the plates is the area of the smaller plate, plus some fringing effect.
|
The wafer rests on a large metal surface. If there is ohmic contact between the large metal surface and the wafer, it is easy to see that the wafer can be one side of the capacitor. If there is no ohmic contact between the large metal surface and the wafer, then there is a large capacitance between them, and the measurement still works. In theory, this capacitive coupling creates an error approximately equal to the ratio of the capacitance between the small top plate and the wafer divided by the capacitance between the wafer and the large metal surface. This error is extremely small compared to the accuracy of the measurement.
The capacitive thickness measurement is calibrated by measuring metal plates of varying thickness and creating a calibration curve.
Semilab In-Line systems (WLT and WMT) measure thickness of wafers while moving on a conveyor belt. In this case the wafer is not grounded, and the one-sided measurement won’t work. Instead, Semilab uses a proprietary double-sided extension of the one-sided measureent technique.
|
