RU2254587C1 - Method of selecting integral microcircuits for radiation stability and reliability - Google Patents

Abstract

FIELD: electronic engineering.

SUBSTANCE: set of microcurcuits is irradiated step by step (number of steps equals two or more) with small dose of ionizing radiation having power from several kilorad to several tens of kilorad. Standard electric characteristics of microcircuits are measured as well as minimal supply voltage of each circuit at which voltage a microcircuit is still capable of working. Dosage dependences are built; the dependences describe changes in standard characteristics and minimal voltage of operation under effect of radiation. Dose of failure is predicted for any microcircuit by means of the dependences, at which dose at least one standard characteristic reaches its limit value or minimal supply voltage reaches nominal value of supply voltage of microcircuit. Reliability of microcircuit is defined after subjecting irradiated microcircuits to burning from deviation in value of one or several standard characteristic or minimal supply voltage from their initial values taken before they were subject to radiation. Method allows to predict radiation stability of microcircuits and selection of microcircuits having increased levels of radiation stability and reliability and (or) rejection of microcircuits having abnormally low levels of radiation stability and reliability.

EFFECT: reduced costs; ability to keep radiation characteristics of systems; high degree of adaptability to manufacture.

1dwg, 2 tbl

Description

The invention relates to the field of electronic engineering, in particular, is intended for the separation of integrated circuits by radiation resistance and reliability.

A known method of selecting electronic products for durability or reliability [1], including the formation of a representative "training" sample for each type of sorting from the test batch of products, irradiating it with a small dose, causing a significant change in at least two parameters specified in the technical conditions or additional constructing the distribution of sample products according to the degree of change in parameters caused by irradiation, annealing until parameters are restored, and testing the sample for radiation resistance until n less than 50% of failures - in the case of presorting the test batch to the radiation resistance; testing the sample for a median resource either during the required time between failures or during the time necessary to determine the required resource - in the case of the rejection of the test batch for reliability; reliability testing of the sample under the combined action of radiation, heat and electric loads until at least 50% of failures are received - in the case of the test party being sorted by the mean time between failures under the combined action of destabilizing factors.

This method, along with obvious advantages, such as the classification of products by radiation resistance and (or) reliability groups under various operating conditions according to the results of tests of the “training” sample, also has drawbacks consisting in the need for radiation resistance tests and reliability of a representative sample and in the impossibility Indication of the maximum dose for each product.

There is also a method of controlling MOS of semiconductor devices and integrated circuits on wafers [2], which proposes the inclusion of the operation "irradiation - annealing" between operations of selective and 100% control of products, which allows to identify potentially reliable products and reject potentially unreliable products.

The advantage of this method is the individual control of the reliability of each product on the plate. However, this method does not allow the determination of the radiation resistance of each product due to the impossibility of conducting irradiation in the electric mode.

The closest analogue - the prototype - of the invention is a method for selecting radiation-resistant electronic equipment products [3], which includes irradiating a batch of products intended for installation in on-board equipment with a relatively small dose of gamma-quanta or electrons, followed by selection and exclusion from the batch of devices with the largest changes in parameters . It is also possible to irradiate with a full dose equivalent to the expected absorbed dose of radiation in real operating conditions, and restore the initial parameters after irradiation by annealing at elevated temperature.

The disadvantage of this method in terms of separation of products for changing parameters during low-dose irradiation is the inability to determine the dose of failure, and in terms of irradiating products to failure, the inability to completely restore the parameters during long-term low-temperature annealing to the original values of the product parameters due to the accumulation of radiation defects during irradiation a large dose at which a failure has occurred.

In the inventive method for separating integrated circuits by resistance to ionizing radiation and reliability, including irradiating a batch of microcircuits with a small dose of ionizing radiation (from several krad to several tens of krad), measuring their electrical parameters and subsequent annealing to stabilize the parameters, the irradiation is carried out in stages with the number of stages not less than two, in addition to standard parameters, the minimum voltage of each microcircuit is measured, at which its functioning is maintained, dependencies describing the change in standard parameters and the minimum operating voltage under the action of radiation, and with their help they predict for each microcircuit a failure dose at which at least one standard parameter reaches its limit value or the minimum supply voltage reaches the nominal value of the microcircuit's supply voltage, and reliability microcircuits are determined after their annealing by the deviation of the value of one or more standard parameters or the minimum supply voltage from their initial values before exposure.

When developing the method, it was assumed that the radiation resistance and reliability of integrated circuits depends on the density of defects in the films of the dielectric and at the interface with the semiconductor. When irradiated with a small dose (from a few cradles to several tens of cradles), charge accumulates on technological defects formed during the manufacture of the microcircuit, without significant formation of new (radiation) defects, which causes a change in the standard parameters and the minimum supply voltage at which the functioning of the microcircuit is maintained. Standard parameters are determined by the elements at the inputs and outputs of the microcircuit, and the minimum supply voltage characterizes the performance of all elements of the microcircuit. The higher the density of defects, the stronger the parameters of the microcircuit change and the lower the dose, their limit value will be reached. The dose dependence, the parameters of which are found for each sample of the microcircuit according to the measurement results after each stage of irradiation with standard parameters and the minimum supply voltage at which operation is maintained, allows us to predict the dose of failure for each product in the batch when standard parameters reach the limit values and (or) minimum voltage power rating. During low-temperature annealing, a thermal discharge of a charge from point defects occurs and it is retained in large defects that are in microcircuit samples with an anomalously low reliability [4]. Therefore, the non-restoration of parameters during annealing makes it possible to identify microcircuit samples with abnormally low reliability and, conversely, a complete restoration or even improvement of parameters indicates a low density or absence of large defects and a high level of reliability.

As examples of the implementation of the proposed method, we present the results of a study of integrated circuits of low (type 564IE14) and high (type 28F020) degrees of integration.

Based on the change in the minimum supply voltage, the maximum dose of ionizing radiation of the CMOS integrated circuit type 564IE14 was predicted. The drawing shows the dose dependence of the minimum supply voltage (U _pit ) _{min of} this chip. According to the measurement results (U _pit ) _min at irradiation doses of less than 10 crad, the parameters of the function of the form (U _pit ) _min = AD ⁿ + (u _pit ) _{min0 were determined} , which describes the dose dependence (U _pit ) _min : n = 0.512, A = 0.664 glad ^-0.512 . When (U _pit ) _min = 5 V, the predicted dose of failure turned out to be equal to D _open = 27.13 deg. The experiment showed a chip failure at a dose of 30 krad. Thus, the error of the proposed method for predicting the dose of failure in this case was ~ 9.6%.

The ability to predict the dose of failure by change (U _pit ) _min was also tested on microchips with a high degree of integration of flash memory of type 28F020 with a memory capacity of 2 Mbit. The measurement results (U _pit ) _min after irradiation, the results of the forecast and experimental verification of the dose of failure are presented in table 1. As you can see, the forecast error does not exceed 10%. Thus, the proposed method allows to determine the individual failure dose of each microcircuit and to separate the batch of microcircuits according to their radiation resistance.

The results of annealing flash memory chips at a temperature of 90 ° C for 473 hours after irradiation with a dose of 10 krad are presented in Table 2. As you can see, the batch of chips was divided into 3 groups with high, medium and low densities of large defects, respectively. Microcircuits with a low density of large defects (or their absence) are characterized by an increased level of reliability, and such samples are recommended for use in highly reliable equipment (for example, on-board equipment of spacecraft). Microcircuits with a high density of large defects should be rejected as characterized by an abnormally low level of reliability.

Thus, the proposed method for separating integrated circuits by radiation resistance and reliability is technologically advanced, it allows determining the value of the failure dose and reliability level for each microcircuit and does not impair the operational characteristics of the integrated circuits.

Literature

1. The method of selecting electronic products for durability and reliability (RF Patent No. 2168735, MKI G 01 R 31/26, 31/28, published June 10, 2001, bull. No. 16).

2. A method for monitoring MOS semiconductor devices and integrated circuits on wafers. (RF patent No. 2073254, MKI G 01 R 31/26, published 02/10/97, bull. No. 4).

3. Chernyshev A.A., Vedernikov V.V., Galeev A.I., Goryunov N.N. Radiation rejection of semiconductor devices and integrated circuits. - Foreign electronic technology. 1979. Issue 5. S.3-25.

4. Popov V.D. Post-radiation effect in IP. Non-destructive quality control of electronic electronics: Science, Technology, Business. 2002. No4. S.36-39.

Table 1
The method of separation of integrated circuits by radiation resistance and reliability №№ (U _pit ) _min , V Forecast D _Otk Experiment D = 0 D = 4 steal D = 8 steal D = 10 deg 40 steal 50 steal 1 1.29 1.91 2,37 2,39 38 steal renouncement - 2 1.27 1.89 2,37 2,50 45 steal function renouncement 3 1.32 2.05 2,55 2.60 43 steal function renouncement 4 1.25 1.97 2.44 2,58 50 steal function renouncement 5 1.27 1.92 2,37 2.45 54 steal function renouncement 6 1.32 1.89 2.22 2.29 63 steal function function

table 2 (U _pit ) _min , V 1,60 1,59 1,50 2.95 1,59 1,61 1,66 3.09 1,62 1.35 1.30 Defect Density Level Wed Wed Wed high. Wed Wed Wed high Wed bottom. bottom. Note: Density level of major defects: high. - high, cf. - middle, bottom. - low.

Claims (1)

A method for dividing integrated circuits by resistance to ionizing radiation and reliability, including irradiating a batch of microcircuits with a small dose of ionizing radiation, measuring their electrical parameters and then annealing them to stabilize the parameters, characterized in that the irradiation is carried out in stages with at least two stages; standard parameters, the minimum supply voltage of each microcircuit, at which its functioning is maintained, is built dose dependencies that describe changes the standard parameters and the minimum operating voltage under the action of irradiation, and with their help they predict for each microcircuit a failure dose at which at least one standard parameter reaches its limit value or the minimum supply voltage reaches the nominal value of the microcircuit supply voltage, and the reliability of the microcircuit is determined after annealing by deviation of the value of one or more standard parameters or the minimum supply voltage from their initial values before exposure.