P-Mosfet Total Dose Dosimeter

The present invention relates to MOSFET dosimetry and more specifically to an improved dosimeter providing increased accuracy in reading total radiation exposure by eliminating temperature effects.

The space radiation effects of concern to modern microcircuits are Single-Event Upsets (SEUs) and Total Ionizing Dose (TD). In the SEU effect, cosmic rays and high energy protons, that undergo electronic reactions with the integrated circuits, deposit sufficient charge in memory cells and latches to flip bits and corrupt data. Such events do not induce any physical damage and thus are nondestructive. On the other hand, in the TD effect, gate oxides are charged which is a permanent change. This shifts transistor threshold voltages and reduces the channel mobility. A change in these parameters degrades the performance of CMOS integrated circuits (ICs) by changing the propagation delay.

Radiation sensitive field effect transistors have application as integrating dosimeters providing a measurement of the amount of dose absorbed from various radiation sources. Field effect transistor (FET) dosimeters are advantageous as compared to more conventional dosimeters because FET dosimeters are small and they provide continuous readout. They operate on the principle that ionizing radiation causes a shift in threshold voltage due to the accumulation of trapped charge in the gate oxide and at the oxide interface. Unfortunately, FET dosimeters also suffer from a number of shortcomings. In particular, their sensitivity to radiation is low and their accuracy is diminished by changes in threshold voltage caused by temperature variations. Thus it would be highly desireable to have an FET dosimeter which is relatively insensitive to variations in temperature.

The present invention comprises a p-MOSFET total dose dosimeter where the gate voltage is proportional to the incident radiation dose. It is configured in an n-WELL of a p-BODY substrate. It is operated in the saturation region which is ensured by connecting the gate to the drain. The n-well is connected to zero bias. Current flow from source to drain, rather than from peripheral leakage, is ensured by configuring the device as an edgeless MOSFET where the source completely surrounds the drain. The drain junction is the only junction not connected to zero bias. The MOSFET is connected as part of the feedback loop of an operational amplifier. The operational amplifier holds the drain current fixed at a level which minimizes temperature dependence and also fixes the drain voltage. The sensitivity to radiation is made maximum by operating the MOSFET in the OFF state during radiation soak.

It is therefore a principal object of the present invention to provide a p-MOSFET total dose dosimeter with increased radiation sensitivity and decreased temperature sensitivity.

It is another object of the invention to provide an improved dosimeter yielding more accurate readings of total radiation exposure by eliminating temperature effects.

It is another object of the invention to provide more accurate dosimeter measurement by employing a p-MOSFET dosimeter in which the MOSFET is edgeless to eliminate leakage current and in which an operational amplifier virtual ground fixes the drain current at a magnitude selected to substantially eliminate temperature sensitivity.

The aforementioned objects and advantages of the present invention, as well as additional objects and advantages thereof, will be more fully understood hereinafter as a result of a detailed description of a preferred embodiment of the invention when taken in conjunction with the following drawings in which:

FIG. 1 is a schematic diagram of the dosimeter of the present invention; and

FIG. 2 is a graph of the temperature dependence of the dosimeter of the invention showing the drain current point at which the dosimeter operation is temperature independent.

Dosimetry via MOSFET threshold voltage shifts is influenced by at least two second order effects. These must be compensated out by design or included in the data analysis in order to obtain accurate dose measurements. The second order effects are (a) the radiation sensitivity of the transconductance factor, KP, via radiation-induced mobility, .mu., degradation and (b) the temperature sensitivity of both of these effects. In addition the total operation of the device must be considered in order to obtain accurate dose measurements.

The p-MOSFET is operated in the saturation region which is ensured by connecting the gate to the drain as seen in FIG. 1. For this case the drain current is given by: ##EQU1## the above square-law relationship is written as a VG-expression:

The radiation sensitivity is:

The temperature dependence is determined by expanding VT and KP in a Taylor Series:

where T is the absolute temperature, T.sub.o is room temperature, and VT.sub.T =.vertline.dVT/dT.vertline.. The temperature dependence of KP is:

and the Taylor Series expansion is:

where KP.sub.T =.vertline.dKP/dT.vertline.=n.multidot.KP.sub.o /T.sub.o and n is between 1.5 and 2.0.

Substituting the above two equations into the VG-expression yields:

VG=VG.sub.o +{-VT.sub.T +.sqroot.[ID.multidot.L.multidot.KP.sub.T.sup.2 /(2.multidot.KP.sub.o.sup.3 W)]}(T-T.sub.o)

where VG.sub.o =VT.sub.o +.sqroot.[2ID.multidot.L/(KP.sub.o .multidot.W)]. The above derivation used the approximation 1/.sqroot.(1-x)=1+x/2 for small values of x. Thus VG will be independent of T for:

where ID=19.2 .mu.A for a MOSFET length of 1.6 .mu.m and width of 3.2 .mu.m and for the parameters given in FIG. 2 where the temperature dependence is shown explicitly.

The total dose, D, is evaluated is a similar manner. The Taylor series expansions for VT and KP are:

where the VT damage coefficient is VT.sub.D =dVT/dD and

where the KP damage coefficient is KP.sub.D =dKP/dD. Substituting the above VT and KP expressions into the VG-expression leads to:

where again the approximation of 1/.sqroot.(1-x) was used.

The above dose expression requires the evaluation of VT.sub.D and KP.sub.D which is done using Co-60 irradiation. In this evaluation dose rate effects must be evaluated and then extrapolated to operating conditions. The temperature coefficients, VT.sub.T and KP.sub.T, are determined from temperature measurements.

The p-MOSFET dosimeter operating modes include: MEASure, OFF (soak/power down), CALibrate 1 and CALibrate 2. For maximum sensitivity to radiation, p-MOSFETs should be operated in the OFF state during the soak state. In the OFF state the electric field is directed from the gate toward to silicon. This means that during the radiation damage process when positive oxide charge is mobile, it will be distributed more toward the silicon which makes the MOSFET harder to turn on. This can be achieved in the implementation shown in FIG. 1 by applying zero bias to the gate. This requirement has the advantage that the soak and power down state will be identical and means the MOSFET will be operated in a consistent bias state providing a well-known biasing scenario. The key to this bias requirement is that the n-well be connected to a zero potential and not be biased to VDD as is the normal cas in integrated circuit design.

The design of the dosimeter requires current flow only from source to drain and not via a peripheral leakage path. Thus the device is designed as an edgeless MOSFET where the source completely surrounds the drain. In addition, all junctions are held at zero bias except for the drain junction. The operational amplifier holds the source at zero bias providing a drain current ID=Vl/R. The drain voltage is fixed by the amplifier at VG which, for the case of the p-MOSFET, is negative.