EP2984494A1

EP2984494A1 - Method for studying the temperature reliability of en electronic component

Info

Publication number: EP2984494A1
Application number: EP14712244.4A
Authority: EP
Inventors: Florent Miller; Sébastien MORAND; Florian Moliere; Thomas Santini
Original assignee: Airbus Group SAS
Current assignee: Airbus SAS
Priority date: 2013-04-09
Filing date: 2014-03-17
Publication date: 2016-02-17
Also published as: FR3004262B1; WO2014166701A1; FR3004262A1

Abstract

The invention relates to a method for analysing the temperature reliability of an electronic component (10) comprising an electronic chip (12) mounted in a housing (11), said electronic chip (12) consisting of a plurality of layers (13-15) of materials. The invention is characterised in that the method comprises the following steps: thermally coupling a thermally conductive coupling body (25) to the electronic chip (12); determining a wavelength according to the layers (13-15) of material to be crossed and the absorption rate of a layer of material to be used; and thermally stressing an area of interest (20) by means of a laser source (32) emitting the pre-determined wavelength.

Description

METHOD OF STUDYING THE RELIABILITY IN TEMPERATURE OF A

ELECTRONIC COMPONENT

Field of the invention

The present invention belongs to the field of the analysis of electronic components, and more particularly to the procedures for analyzing the operation of electronic components subjected to thermal stresses. The electronic components, before their implementation, are generally subjected to analysis procedures during which these components are notably subject to constraints representative of the operational constraints, that is to say the constraints to which they are susceptible of to be confronted in their final environment. These analysis procedures are especially important, especially for components that are required to operate in highly constrained environments, such as components to be used in space and / or military missions, aircraft, power plants, energy, etc. State of the art

During a procedure for analyzing the reliability of an electronic component, it is common to subject the component in operation to a thermal stress, which corresponds, for example, to an operational thermal stress (it is then referred to as a functional test) or a thermal stress applied to accelerate the aging of said integrated circuit (it is called aging test).

The aging tests make it possible to reveal the failure mechanisms such as the breakdown of the gate oxide of the transistors, the instability of the threshold voltage in temperature as well as the phenomenon of electromigration present at the level of the metallizations. These tests aging also make it possible to anticipate the thermomechanical fatigue of the component present at the level of the connectors, the metallizations and the barrier layer between the etching and oxide layers of a component.

To carry out the functional or aging tests, one method consists of carrying out tests in a climatic chamber or under a packing bell. Thus, the component and its support are solicited, statically or by cycle, in temperature. However, these tests require a large number of elements such as the component chip, the housing, the solders, the connection pins, the connection wires, thus limiting the possibilities of interpretation of the failure mechanisms.

Methods, by physical contact with thermally conductive elements, allow a localization of the temperature stress, particularly at the chip. A specific preparation of the component is then necessary to make possible a direct physical access to the area of interest (metallization / semiconductor / ...). Among these methods, it will be possible to mention that of making a localized thermal stress using a Peltier effect module coupled to a thermally conductive element which makes contact with the constrained material (semiconductor, metal, etc.). ) described in the French patent application No. FR 2 959 018. Thus, the thermal stress is much more localized than in the case of an oven test for example. However, for mechanical reasons, this method does not reduce the size of the element brought into contact with the area of interest below the mm ² . However, a smaller solicitation area would limit heating zones around that of interest. Moreover, this method requires physical contact between the area of interest and the thermally conductive element which, in some cases, may be limiting. In other cases, it is interesting to be able to solicit on the contrary on much larger surfaces (the mechanical contact can hardly be assured on large surfaces, especially in the presence of asperities or different layers). Presentation of the invention

The present invention intends to overcome the drawbacks of the prior art by proposing a solution which makes it possible to locate thermal stresses on an electronic component by means of a laser beam coupled to a cooling device of the component.

For this purpose, the present invention relates, in its most general sense, to a method for analyzing the temperature reliability of an electronic component comprising an electronic chip mounted in a housing, said electronic chip being composed of several layers of materials. said method comprising the following steps:

- Thermal coupling of a coupling member, thermally conductive, with the electronic chip;

determination of a wavelength as a function of the layers of material to be crossed and of the absorption rate of a layer of material to be stressed; and

- Thermal stressing of an area of interest by means of a laser source emitting the predetermined wavelength.

There are methods for analyzing component failure by laser heating to locate areas of a degraded semiconductor component. The invention makes it possible to implement, for example, the thermally induced voltage modification (TIVA) method or the thermally induced resistance modification method "OBIRCH" (for "Optical induced voltage alteration"). beam induced resistance change ").

In the case of the method for modifying the thermally induced resistance "OBIRCH", it is in particular to identify degradations in the metallizations of the electronic component. This method is based on a local modification of the properties (resistivity for example) of the metallization depending on whether the impacted area is degraded or not. The temperature variations generated by the laser beam are however in this case adjusted (in energy, frequency or duration) so as not to contribute to degrade or age the component. The temperature variations produced are typically of the order of ten degrees.

The invention thus makes it possible to simulate, on the basis of the physical laser heating mechanisms, much larger temperature ranges making it possible to reproduce, in a localized manner (at the scale of one or a few elementary structures in the component) the constraints environmental and / or functional aspects of an electronic component. The invention makes it possible, for example, to impose temperatures varying between -55 ° C. and + 125 ° C., depending on the characteristics of the laser source and the regulation source.

According to one implementation, the wavelength is determined so that the energy of the photons of the laser beam is less than the energy of the forbidden band of the electronic chip. This implementation prevents the photoelectric interactions in the component. In the case of a silicon component, the wavelength must thus be less than 1, 1 μιτι.

According to one implementation, said method comprises two recurrent steps: a first step of activating the laser source and a second step of deactivating the laser source. This implementation creates a cycle phenomenon in the component that accelerates its aging.

According to one implementation, said method comprises a step of adapting the laser source to an area of interest of the electronic component. This implementation makes it possible to target an area of interest whose size is adjustable according to the needs. Preferably, said method comprises a step of focusing or adjusting the size of a collimated beam of the laser source on one or more structures of interest of the electronic component. This implementation makes it possible, for example, to target with a high precision, through the use of a high-magnitude objective, a transistor or a logical function of a component for which a sensitive mechanism can intervene, or if the beam is less focused and / or collimated, to solicit a component in whole. According to one embodiment, the thermal coupling member is a passive cooling device. A passive cooling device includes climatic chambers, cryostats and natural convection devices. It keeps the component at a substantially constant temperature.

According to one embodiment, the thermal coupling member is an active cooling device. An active cooling device includes thermoelectric modules, water cooling devices and forced convention devices. It maintains, more effectively than a passive device, the component at a substantially constant temperature.

According to one implementation, said electronic component comprises a front face and a rear face and the coupling member may be applied on the same face as the thermal stressing or on opposite faces. This implementation makes it possible to very effectively adapt the device for implementing the method as a function of the test carried out.

According to one implementation, the method comprises a radiation stressing step of the electronic chip through an opening, previously formed in the housing, the coupling member preserving optical access to said opening. This implementation makes it possible to limit the energy losses of the laser to pass through the housing.

According to one embodiment, said method further comprises:

a second step of determining a wavelength according to the layers of material to be crossed and the absorption rate of the layer of material to be solicited and

a second step of thermally stressing the zone of interest by means of a laser source emitting the predetermined wavelength. This implementation makes it possible to solicit the electronic component by two laser sources simultaneously having the effect of increasing the possibilities of energy transfer. Preferably, a first Laser source is directed to one side while the other source is directed to the other side.

Brief description of the drawings

The invention will be better understood by means of the description, given below purely for explanatory purposes, of the embodiments of the invention, with reference to the figures in which:

FIG. 1 illustrates a device for analyzing the temperature reliability of an electronic component according to a first embodiment, in which the laser beam and the coupling member are directed on the front face;

FIG. 2 illustrates an analysis device according to a second embodiment, in which the laser beam is collimated on the front face and the coupling member is in contact with the rear face;

FIG. 3 illustrates an analysis device according to a third embodiment, in which the laser beam is focused on the rear face and the coupling member is in contact with the rear face;

FIG. 4 illustrates an analysis device according to a fourth embodiment, in which the laser beam is focused on the front face and the coupling member corresponds to a climatic chamber;

FIG. 5 illustrates an analysis device according to a fifth embodiment, in which the laser beam is collimated on the front face and the coupling member corresponds to a climatic chamber;

FIG. 6 illustrates an analysis device according to a sixth embodiment, in which the laser beam is focused on a large surface of the front face and the coupling member corresponds to a climatic chamber; FIG. 7 illustrates the evolution of the absorption coefficient per centimeter in boron-doped silicon as a function of photons of different energies;

FIG. 8 illustrates the evolution of the absorption rate of several materials as a function of the wavelength of the laser; and

• Figure 9 illustrates the evolution of the temperature in the component as a function of the distance to the area of interest.

DETAILED DESCRIPTION OF THE EMBODIMENTS OF THE INVENTION

FIG. 1 illustrates a device for analyzing the temperature reliability of an electronic component 10. Said component 10 is composed of a housing 11 surrounding a chip 12, the chip 12 is composed of several layers 13-15 of materials . For example, the chip 12 of FIG. 1 may be composed of an aluminum metallization layer 13, a silicon substrate 14 and a copper lower layer 15. The chip 12 has two opposite faces 17-18, a rear face 18 and a front face 17. In the context of the invention, the term "housing" 1 1 any element preventing mechanical access to the electronic chip 12. 1 1 housing may for example be formed by a protective shell, an encapsulation resin ...

On the front face 17, the device imposes a thermal stressing by means of a laser beam 33 emanating from a laser source 32 and a thermal regulation by means of a thermal coupling member 25.

In the example of Figure 1, the thermal coupling member 25 comprises a source 24 of thermal stress, which can be of any known type. Preferably, the source 24 of thermal stress comprises a thermoelectric module 29, such as a Peltier module. A Peltier module generally has two opposite faces. When it is supplied with current, one of its faces, called "hot face", will warm up, while the other face, called "cold face", will cool down. Preferably, the source 24 of thermal stress also comprises a thermoregulator module 23, which can be of any known type. This is for example a module thermoregulator circulating heat transfer fluid (water, oil, etc.), implemented to regulate the temperature of a face of the Peltier module, the cold face if it is desired to apply a heat stress to the electronic component 10. The Coupling member 25 also has an end 26 to be thermally coupled to source 24 of thermal stress. The end 26 of the coupling member 25 is wholly or partly made of thermally conductive material, for example copper or copper alloy, aluminum or aluminum alloy, etc. The end 26 of the coupling member 25 is adapted, by its geometry, to be introduced into an opening 19 previously formed in the housing 1 1 of the electronic component 10. By "geometry" of the end 26 is meant its external volume, that is, its shape and dimensions. The opening 19 previously formed in the housing 1 1 provides mechanical access to all or part of the front face 17 of the electronic chip 12, with which a coupling face of the end 26 is thermally coupled at the time of entry. stress of the electronic component 10.

The coupling member 25 comprises an optical window 30 allowing the laser beam 33 to pass to reach the front face 17 of the electronic chip 12. The optical window 30 can take any known shape such as a cylindrical recess.

Figures 2 to 4 show other embodiments of the device for analyzing the temperature reliability of an electronic component 10.

The thermal coupling member 25 may take different forms other than that of Figures 1 and 3 comprising a thermoelectric module. The

2 shows a forced convection system in which the coupling member 25 is in contact on the entire rear face 18 of the electronic chip 12. The coupling member 25 can thus be traversed by a hydraulic circuit whose temperature of the water is adjustable. Figures 4, 5 and 6 illustrate a coupling member 25 formed by a climatic chamber 37 whose temperature is fully regulated. The laser beam 33 enters the climatic chamber by an optical access window having a glazed surface. The coupling member 25 has the effect of limiting as much as possible the thermal stress experienced by the whole of the electronic component 10 to the zone of interest 20. In fact, the localized stress can be effective only if localized heating is coupled to a dissipation phase to evacuate the heat generated locally or if a second source is used to force a thermal gradient in the area of interest.

The coupling member 25 can be positioned on the opposite side to the laser excitation (in the case of FIG. 2) or on the same side (in the case of FIGS. 1 and 3) or can encompass the electronic component 10 (as in FIG. 4). . The function of the coupling member 25 is to amplify the thermal stresses and to constrain the electronic component 10 more locally.

Moreover, the laser beam 33 makes it possible to create a localized heating in a precise layer 13-15 of the electronic chip 12. For example, in the case of FIG. 2, on a stack of three layers of aluminum 13, of silicon 14 and copper 15, the regionally heated area being aluminum, the energy of the laser source 30 is adjusted so that the temperature at the surface of the aluminum layer 13 is +200 ° C. The copper layer 15 serves as an interface with the coupling member 25 imposing a temperature of -20 ° C. The presence of active cooling makes it possible to greatly reduce the temperature at the silicon layer 14 of the electronic chip 12 while maintaining a high level of temperature at the level of the aluminum layer 13 on which the area of interest is placed. 20. The laser beam 33 can also be focused by means of optical lenses 34 of different magnitudes as shown in FIGS. 3 and 6. The focused laser beam 33 allows, according to the objective used, to urge one or more elementary structures of the electronic component 10 or an entire component. In the case of Figure 6, the focused laser beam can solicit a component in whole. In contrast, the laser beam 33 may also be collimated, as shown in Figures 2 and 5, to adjust the size of the thermally biased area of interest. Moreover, the device can implement several laser beams 33 to increase the stress. For example, two laser beams 33 may be arranged on either side of the electronic component 10 to bias an elementary structure independently of the position of the coupling member 25. A first laser source is then arranged to bias the component 10 by the front face 17 and a second laser source is arranged to bias the component 10 by the rear face 18. The two laser sources must emit beams of different wavelengths determined according to the layers 13-15 of materials that the laser beams must cross to solicit the same area of interest 20.

When the laser source 32 emits, it very locally heats the material. Due to the small heated surface, very rapidly, when the laser source 32 no longer emits, the zone of interest 20 returns to ambient temperature or to the operating temperature. The laser beam 33 can thus be used to simulate aging of the electronic component 10 by alternating laser excitation phases and non-excitation phases. In addition, the laser source 32 may also emit a pulsed or non-pulsed laser beam 33.

The wavelength λ of the laser source 32 is determined according to three criteria:

to limit the other physical phenomena disturbing the normal operation of the component 10, in particular in the case for which the component 10 is under tension,

allow the laser beam 33 to cross, if necessary, the layers of materials to reach the zone of interest 20, and

- maximize the absorption of laser energy in the area of interest 20.

To limit the other physical phenomena disturbing the normal operation of the component 10, the energy of the photons must be less than the forbidden band energy of the chip 12 to prevent photoelectric interactions in the materials. For example, in the case of a 10 silicon electronic component, the wavelength λ should be greater than 1, 1 μιτι if the beam is to cross this area before reaching the area of interest.

To allow the laser beam 33 to pass through different layers of materials, it is a question of minimizing the absorption mechanisms so that the bulk of the energy of the laser beam 33 can be brought into the area of interest 20. At this point, Indeed, Figure 7 illustrates the absorption rate τ (in cm-1) in a layer of boron-doped silicon material (the dopings are given in 1 e18 / cm3) as a function of different photon energies δ and for different amounts of doping. The energy δ of the photons is directly connected to the wavelength λ by the Planck constant and the celerity of the light. The photon energy δ has a decay phase followed by a growth phase. Despite the logarithmic representation of the absorption rate τ which can be misleading, the evolutions of the absorption rate τ are more pronounced when the doping of silicon increases. For example, for a doping of 2.8e18 / cm3, the absorption rate τ changes between 30 and 10 then between 10 and 300 whereas for a doping of 120e18 / cm3, the absorption rate τ changes between 1800 and 800 then between 800 and 2000. The minimum energy δ differs depending on the amount of doping, this minimum is substantially located to 1 .12 electronvolts (eV) at 300 K. Thus, for boron doped silicon layers, a length a wavelength λ close to a value greater than 1 .1 μιτι corresponding to an energy δ of 1 .12 eV is preferably used, the silicon being transparent for these wavelengths with respect to the linear absorption mechanisms. This analysis differs according to the material of the layer or layers to be crossed.

In order for the absorption of the laser energy by the zone of interest to be maximum, the wavelength λ must be chosen preferentially in a range of values for which the material of the zone of interest has a rate of absorption τ maximum. For this purpose, FIG. 8 illustrates the absorption rate T of the laser for different materials as a function of the wavelength λ of the laser source 30. The materials represented are aluminum (Al), silver (Ag ), copper (Cu), gold (Au), molybdenum (Mo) and iron (Fe). Except for aluminum, for all the other materials presented, the absorption rate τ decreases the longer the wavelength λ increases. As regards aluminum, the absorption rate τ decreases to a wavelength λ of 0.5 μιτι and then has a resonance peak. Thus, to transfer energy (and therefore heat) an aluminum metallization, a wavelength λ between 800 nm and 1 .5μιτι can be advantageously used. By combining this constraint with the preceding constraints, when the laser beam 33 passes through a silicon layer 14 to reach the zone of interest 20, the choice is preferably made on a wavelength λ of 1 .3 μιτι or 1 .5 μιτι.

FIG. 9 represents the evolution of the temperature (in Kelvin) in the thickness (in μιτι) of a chip 12, comprising an aluminum layer 13, a silicon layer 14 and a copper layer 15, solicited at the of the aluminum layer 13. The temperature at the aluminum layer 13 is about 400 K and then it decreases rapidly in the thickness of the silicon layer 14 between 390 K and 345 K. The temperature decreases still in the thickness of the copper layer 15 between 310 K and 260 K. The laser beam 33 thus effectively heats the aluminum layer 13 while the coupling member 25 allows a rapid dissipation of heat to reduce the thermal stress of the other layers 14-15.

The invention makes it possible to solicit an electronic component 10 in a localized manner and on large temperature variations. The invention can be applicable to any type of electronic components (digital, linear, power) and technologies (modulo an adjustment of the wavelength λ of the laser). The invention can also be used on any physical system (solid, liquid) requiring localized stressing.

Claims

1. A method for analyzing the temperature reliability of an electronic component (10) having an electronic chip (12) mounted in a housing (1 1), said electronic chip (12) being composed of a plurality of layers (13-15) of materials, characterized in that the method comprises the following steps:

- Thermal coupling of a coupling member (25), thermally conductive, with the chip (12) electronic;

- determination of a wavelength (λ) as a function of the layers (13-

15) of material to be crossed and the absorption rate (τ) of a layer of material to be stressed; and

- Thermal stressing of an area of interest (20) by means of a laser source (32) emitting the predetermined wavelength (λ).

2. Method according to claim 1, characterized in that the wavelength (λ) is determined so that the energy of the photons of the laser beam is less than the energy of the forbidden band of the chip (12) electronic .

3. Method according to claim 1 or 2, characterized in that it comprises two recurrent steps: a first step of activating the laser source (32) and a second step of deactivating the laser source (32).

4. Method according to one of claims 1 to 3, characterized in that it comprises a step of adapting the laser source (32) on an area of interest (20) of the electronic component (10).

5. Method according to claim 4, characterized in that it comprises a step of adjusting the size of the beam, by focusing or enlarging / reducing the collimated beam of the laser source (32) on an area of interest of the component electronic (10).

6. Method according to one of claims 1 to 5, characterized in that the coupling member (25) thermal is a passive cooling device.

7. Method according to one of claims 1 to 5, characterized in that the coupling member (25) thermal is an active cooling device.

8. Method according to one of claims 1 to 7, characterized in that said electronic component (10) comprises a front face (17) and a rear face (18) and the end (26) of the coupling member (25) can be applied on the same face (17-18) as the laser source (32) or on opposite faces.

9. Method according to one of claims 1 to 8, characterized in that it comprises a radiation stressing step of the electronic chip (12) through an opening (19), previously formed in the housing (1). 1), the coupling member (25) preserving optical access (30) to said opening (19).

10. Method according to one of claims 1 to 9, characterized in that it further comprises:

a second step of determining a wavelength (λ) as a function of the layers (13-15) of material to be crossed and the absorption rate (τ) of a layer of material to be solicited and

a second step of placing the zone of interest (20) under thermal stress by means of a laser source (32) emitting the predetermined wavelength (λ).