WO2003012468A1

WO2003012468A1 - Method and device for examining heat sources within a sample that comprises an electroconductive material

Info

Publication number: WO2003012468A1
Application number: PCT/EP2002/006830
Authority: WO
Inventors: Frank Altmann; Otwin Breitenstein
Original assignee: Fraunhofer Gesellschaft zur Förderung der angewandten Forschung e.V.
Priority date: 2001-07-27
Filing date: 2002-06-20
Publication date: 2003-02-13
Also published as: DE10136774C1

Abstract

The invention relates to a method and a device for examining heat sources within a sample that comprises an electroconductive material and that is impinged with a voltage for the purpose of heat development. The inventive method is further characterized in that the sample is impinged with a fluorescent substance on at least one sample surface, the fluorescent properties of said substance being temperature-dependent. The sample surface so prepared is subjected to an energy flow that energizes the fluorescent material, and while and/or once the sample is supplied with a voltage, images of the fluorescent radiation emitted by the prepared sample surface are recorded by means of an image recording device and are evaluated by means of an evaluation unit. The invention is specifically characterized in that the voltage is impinged with a modulation frequency fP and that the image recording device is coupled in phase to the modulation frequency fP according to a lock-in principle.

Description

Method and device for examining heat sources within a sample comprising electrically conductive material

Technical field

The invention relates to a method and an apparatus for examining heat sources within an electrically conductive material sample which is subjected to an electrical voltage for the purpose of heat development, in which the sample is exposed to at least one sample surface with a fluorescent agent, the Fluorescence property is temperature-dependent, the sample surface prepared in this way is exposed to an energy flow that energizes the fluorescent agent before, during and / or after the sample is supplied with the electrical voltage and by means of an image recording device images of fluorescent radiation emerging from the prepared sample surface are recorded and evaluated by an evaluation unit ,

State of the art

The function of electronic components such as integrated circuits (ICs) is based on internal current flows. Local heat sources are always formed. These heat sources can be caused by various mechanisms, such as Joule heat when a current flows through an electrical resistor, recombination heat when recombining non-equilibrium charge carriers in a semiconductor, relaxation heat when thermalizing hot charge carriers, or Peltier heat when the current flows at the contact between different ones conductor materials. That's why it is It is obvious to use the thermographic mapping of local temperature differences in semiconductor components for the functional test of these components.

Previously common methods for the thermographic examination of electronic components are the examination with an infrared (IR) thermal camera (see, for example, J. McDonald, L. Optics, G. Albright, "Microthermal imaging in the infrared", Electronics Cooling 3 (1997) pp .}, the investigation with temperature-sensitive liquid crystals (see e.g. DJ Farina, "High resolution thermal mapping of integrated circuits using nematic and thermochromic liquid crystals", Temptronic Corp. publication), and Fluorescent Microthermal Imaging (abbreviated FMI, see e.g. P. Kolodner , JA Tyson, "Microscopic fluorescent imaging of surface temperature profiles with 0.01 ° C resolution", Appl. Phys. Lett. 40 (1982) pp. 782-784).

In particular, the latter FMI method has become established in the past few years in addition to the liquid crystal method as a standard method for thermal functional testing of electronic components and circuits, since its sensitivity is about an order of magnitude better than that of the other two methods. Furthermore, the FMI method has a nominal spatial resolution down to the sub-micrometer range.

In the FMI process, a thin layer of a special dye is applied to the surface of the sample, the luminescence of which is strongly temperature-dependent in the visible spectral range when UV-illuminated. In fluorescence microscopy, heated areas therefore appear darker than unheated areas.

Since the stimulating UV intensity cannot be chosen to be arbitrarily high because of the degradation of the dye used, an image integration time of usually a few seconds to minutes must be selected when measuring the luminescence intensity in order to obtain an image that is as noise-free as possible. This is neither possible with conventional CCD cameras nor is it useful because of the detector noise. For this reason, the FMI method usually uses a cooled, high-resolution slow scan CCD camera, which usually has an image integration time of many seconds. In order to suppress the influence of the surface topography and to achieve a "thermal" signal that is as pure as possible, methods of image storage and image processing are usually used to subtract the luminescent image from the component under no load and to display only the difference image.

Although the FMI process has become established in many places, it still has some disadvantages. The main disadvantage concerns the temperature resolution of 0.01 ° C (= 10 iτiK), which is inadequate for many applications. Especially for the investigation of so-called "low power" components, it would be desirable to improve the sensitivity by a further 1 to 2 orders of magnitude.

Furthermore, in the investigation of silicon components, the strong thermal conductivity of silicon has a disturbing effect, as a result of which the heat generated in a specific area is generated relatively quickly, i.e. can spread laterally within a few milliseconds, whereby the lateral temperature contrasts "smear" after a short time. This effect occurs especially in macroscopic applications when the heat sources are not small compared to the dimensions of the sample. As a result, the macroscopic spatial resolution of the FMI method is significantly worse than the microscopic one.

An analogous problem also arises with infrared (IR) thermography if it is operated as a stationary process, i.e. if the heat input is constant over time due to the application of an electrical voltage to the sample. In addition, IR thermography has the disadvantage that the spatial resolution of the method is limited to a range of 3-5 μm due to the wavelength. As a result, IR thermography is in principle not suitable for examining sub-micrometer structures.

A better use of the potential of IR thermography can be achieved by the so-called lock-in thermography. With lock-in thermography the heat input is periodically modulated and phase-coupled individual thermal images are recorded. The numerical correlation of the recorded thermal images ensures that the result reflects the local periodic temperature modulation. Lock-in thermography using infrared cameras is known per se and is used, for example, for non-destructive material testing (eg G. Busse, D. Wu, W. Karpen, "Thermal wave imaging with phase sensitive modulated thermography", J. Appl. Phys. 71 (1992) pp. 3962-3965).

It has been found that lock-in thermography can significantly improve both the temperature resolution and the macroscopic spatial resolution compared to stationary IR thermography (O. Breitenstein, M. Langenkamp, F. Altmann, D. Katzer, A. Lindner , H. Eggers, "Microscopic lock-in thermography investigation of leakage sites in integrated circuits", Rev. Sci. Instr. 71 (2000) pp. 4155-4160). However, the microscopic spatial resolution of this method is limited to 3-5 μm by the wavelength used, which means that it cannot be used to examine sub-micrometer structures either.

US Pat. No. 5,653,539 describes a method and a device for temperature detection of a surface on which a chemiluminescent layer is applied, which is capable of emitting temperature-dependent luminescent radiation, which is received by means of a suitable detector and is used via an evaluation unit to determine the surface temperature.

Finally, EP 1006346 A1 shows a measuring method for determining the surface temperature which is based on the principle of phophoresis measurement. In this case, a layer containing phosphorus or phosphorus is applied to a surface to be examined, the temperature-dependent luminescent radiation of which is received by a detector system and evaluated accordingly. Even with the two methods mentioned above, only a rough spatially resolved temperature measurement can be achieved.

Presentation of the invention

It is therefore an object of the present invention to develop a method and a device for examining heat sources within an electrically conductive material sample which is subjected to an electrical voltage for the purpose of developing heat in such a way that the disadvantages mentioned above with respect to the prior art are avoided. In particular, the method should serve for the thermal functional diagnosis of electronic components and should have a better temperature resolution and a better macroscopic spatial resolution than the FMI method. In addition, the process should not be influenced by temperature drift. Finally, it should also be possible to carry out the method using inexpensive uncooled CCD cameras.

The solution to the problem on which the invention is based is specified in claim 1. The subject matter of claim 18 is a device designed according to the invention, with which the method can be carried out. Features which advantageously further develop the inventive concept can be found in the subclaims and in the description and the figure.

The approach according to the invention is based on the FMI method, which is known per se, in which an electrical voltage is applied to the sample for the purpose of examining heat sources within a sample comprising electrically conductive material, for the purpose of developing heat. The sample is provided on at least one of its sample surfaces with a fluorescent agent, the fluorescence property of which is temperature-dependent, and is exposed to an energy flow that energizes the fluorescent agent before, during and / or after the sample is supplied with the electrical voltage. With the help of an image recording device, images of the fluorescence radiation emerging from the prepared sample surface are recorded and correspondingly evaluated by means of an evaluation unit. According to the invention, a modulation frequency f _P is now applied to the electrical voltage and the image recording device is phase-coupled to the modulation frequency f _P by means of a lock-in principle.

A pulsed heat input is generated within the sample by the electrical voltage applied to the sample with a predeterminable modulation frequency f _P, which is recorded in phase lock-in principle by the image recording device, which records images with a likewise predeterminable refresh rate f _w becomes. This will combine the good microscopic spatial resolution of the FMI method with the improvement of the local temperature sensitivity and the macroscopic spatial resolution of the lock-in IR thermography method. This means that the electrical voltage is no longer applied stationary to the sample to be examined, which is preferably an electronic component, as was previously the case with the IR thermography method, but rather that the component is operated, for example, with a preferably rectangularly pulsed electrical voltage.

The energy flow which excites the fluorescent agent on the sample or component surface is preferably UV radiation, which in the simplest case is stationary, i.e. with a constant intensity over time, is irradiated onto the surface. The luminescent surface is recorded in a light microscope with a sensitive camera as an image recording device, digitized and fed to a computer system as an evaluation unit.

The camera is operated at a fixed refresh rate f _w , which, depending on the camera type, is between under a heart and about 50 Hz or above. With the aid of the lock-in principle, the modulation of the electrical voltage applied to the component is correlated with the image repetition frequency f _w by means of a phase coupling. The lock-in correlation is preferably carried out the images of the camera fed to the evaluation unit according to a method customary in lock-in IR thermography. An example is the simplest 4-point correlation method (G. Busse, D. Wu, W. Karpen, "Thermal wave imaging with phase sensitive modulated thermography", J. Appl. Phys. 71 (1992) pp. 3962-3965 ) referenced, in which four images are measured at equal intervals in each lock-in period, ie while the component is being supplied with electrical voltage, and the temperature modulation amplitude for each pixel from these four values S1, S2, S3 and S4 A and the phase position Φ of the temperature modulation can be calculated according to the following rule:

A = _Λ / (S1-S3) ² + (S2 -S4) ²

The lock-in method is characterized in that the measurement can be averaged or integrated over many lock-in periods, which improves the signal-to-noise ratio. In terms of noise, the lock-in integration time is equivalent to the image integration time in the conventional, stationary FMI method. In the case of the lock-in FMI method according to the invention, however, this integration time is distributed over many individual modulation periods. The advantage of the method is therefore that the use of expensive cooled slow-scan CCD cameras is not necessary, but that cheaper CCD cameras can be used which operate at a refresh rate of 50 Hz or even higher. The reason for this is that slow signal changes are not included in the result anyway when using the lock-in method.

The maximum possible lock-in frequency, which initially corresponds to the modulation frequency f _P with which the component is supplied with electrical voltage, is f | for this type of method implementation _0ck . _in = f _P = f _w / 4, is therefore still significantly below the refresh rate f _{w of} the camera. However, this is a clear one, especially when using a slow-scan CCD camera O

In contrast, it is known from lock-in IR thermography that the higher the lock-in frequency, the better the macroscopic spatial resolution of this method. Lock-in thermography with very low frequencies of a few Hz or even below therefore does not improve the macroscopic spatial resolution compared to stationary thermography. For this reason, frequencies in the 100 Hz range up to the kHz or even MHz range should be aimed for in lock-in thermography. Since this is not possible with the slow-scan CCD cameras customary for the FMI process, the combination of lock-in process and FMI process is not obvious.

In a special embodiment of the method according to the invention, however, the modulation frequency f _P and thus also the effective lock-in frequency can also be significantly above the refresh rate f _w . For this purpose, the amplitude and thus the intensity of the exciting UV light is not kept constant, but is modulated with a frequency f _m0C ι, which preferably differs somewhat from f _P. An optimal UV modulation frequency is f _m0 d = (fp + fw / 4).

If the relaxation time of the luminescence of the dye used is sufficiently short, then these test conditions lead to the luminescence light being amplitude-modulated with a frequency of f _w / 4. This means that the four-point correlation can be carried out here, as described above, with four successive camera images, even though the pulse repetition frequency used for the heat input is significantly higher.

In particular, the UV modulation frequency can be 100 Hz and can be synchronized with the mains frequency. This corresponds to the use of a conventional AC-operated mercury vapor lamp for generating the UV lighting, as is also used in conventional FMI systems, and which "flickers" in time with the current load. If voltage pulses with a frequency of f _p = 100 Hz - f _w / 4 are then applied to the sample, then the luminescent light is also amplitude-modulated here with fw / 4, which means that again with 4 pictures the 4-point correlation can be carried out, although the pulse repetition frequency used for heat input is much higher at 100 Hz.

Another particularly advantageous method variant provides that the electrical voltage of the component under investigation is not pulsed to modulate the heat input into the sample, but that a nominal operating voltage is applied permanently and only certain control inputs of the component are appropriately controlled with control voltages. Depending on these control signals, certain heat sources arise and disappear in the component, which can provide information about the function of the component. Through a complex timing of different control signals and clock signals, it can be achieved that only the elements to be examined are specifically activated in such a way that they generate a periodically modulated heat which is then detected by the lock-in FMI method. This enables a detailed investigation of the function of complex structures that goes far beyond the summary mapping of internal heat sources in the component in question.

Brief description of the invention

The invention is described below by way of example without limitation of the general inventive concept using an exemplary embodiment with reference to the drawing. It shows:

Fig. 1 Schematic circuit diagram of an apparatus for performing the method according to the invention.

Ways of carrying out the Invention, Industrial Usability

FIG. 1 shows the functional diagram for a typical exemplary embodiment for carrying out the lock-in FMI method according to the invention. The device shown in FIG. 1 shows an electronic component 3, on the component surface of which is coated with a fluorescent agent, UV radiation from a UV lighting device is directed. The UV radiation causes fluorescence radiation, which is imaged onto a CCD camera 6 via a microscope 5. The images recorded by the CCD camera are then evaluated in a computer.

The component itself is supplied with electrical voltage by means of a frequency generator 1 and a multiplexer 2 in the context of digital frequency processing, as described below.

The digital frequency processing consists in modulating the UV illumination while recording each individual image with the same frequency with which the heat generation in the sample is also modulated (f _m od = fp) - however, between two successive image recordings there is a phase UV modulation shifted by 90 ° compared to sample heating. As a result, the total phase shift after taking 4 pictures is 360 °, which corresponds to an average frequency difference of 1/4 of the refresh rate.

For this purpose, the frequency generator 1 used for the modulation has four outputs, at which the set modulation frequency appears with a phase position of 0 ° (reference), 90 °, 180 ° and 270 °. The multiplexer 2 selects one of these four signals and uses it to pulse the heat generation in the sample 3 under investigation. This is irradiated by a UV lighting device 4, the intensity of which is modulated by the 0 ° (reference) signal of the signal generator 1. The luminescence image generated is recorded via the microscope 5 by the CCD camera 6, which is operated at a fixed refresh rate f _w . The image trigger signal taken from the camera control is used to switch the multiplexer 2 cyclically, whereby the phase of the sample control is increased by 90 ° after each captured image. The images of the camera transmitted to the computer system 7 are correlated with one another there according to the known 4-point method, if necessary averaged over several periods, and according to the Take at least 4 pictures as shown in lock-in thermography.

LIST OF REFERENCE NUMBERS

frequency generator

multiplexer

Sample, component

lighting means

microscope

Camera computer, evaluation unit

Claims

claims

1. A method for examining heat sources within a sample (3) having an electrically conductive material, which is subjected to an electrical voltage for the purpose of heat development, in which the sample (3) is exposed to at least one sample surface with a fluorescent agent whose fluorescent property is temperature-dependent, the sample surface thus prepared is exposed to an energy flow that energizes the fluorescent agent before, during and / or after the sample is supplied with the electrical voltage, and an image recording device (6) is used to record images of fluorescent radiation emerging from the prepared sample surface and by means of an evaluation unit (7) are evaluated, characterized in that the electrical voltage is subjected to a modulation frequency f _P and the image recording device (6) is phase-coupled with the modulation frequency f _P by means of a lock-in principle lt will.

2. The method according to claim 1, characterized in that the image recording device (6) records images with an image repetition frequency f _w of the prepared sample surface from which fluorescent radiation emerges.

3. The method according to claim 2, characterized in that the reciprocal value of the modulation frequency f _P corresponds to a so-called lock-in period within which 1 / f _w images are recorded by the image recording device (6), and that for a measurement process for examination of heat sources in the sample (3) is averaged over a number of lock-in periods.

4. The method according to any one of claims 1 to 3, characterized in that the time-modulated with the modulation frequency f _P electrical voltage is a square wave voltage and is applied to the sample (3).

5. The method according to any one of claims 1 to 4, characterized in that electromagnetic radiation, preferably UV light, is used as the energy flow.

6. The method according to any one of claims 1 to 5, characterized in that the energy flow acts on the sample surface with constant intensity over time.

7. The method according to any one of claims 1 to 6, characterized in that the f _P and f _{w are} chosen such that:

f _P <f _w .

8. The method according to claim 7, characterized in that the following applies: f _P = fw / 4. and that the evaluation unit (7) evaluates the images recorded by the image recording device (6) by means of a 4-point correlation.

9. The method according to any one of claims 1 to 5, characterized in that the intensity of the energy flow acting on the sample surface is applied with a modulation frequency f _m0d .

10. The method according to claim 9, characterized in that the following applies: f _m0 d not equal to f _P.

11. The method according to claim 9 or 10, characterized in that the following applies: f _m0 d = fp + f / 4

12. The method according to any one of claims 9 to 11, characterized in that the following applies: f _m0 d = 100 Hz, and that f _m o is synchronized with the network frequency.

13. The method according to any one of claims 9 to 12, characterized in that the phase coupling based on the lock-in principle between the image recording device (6) and f _{P takes place} on the basis of the difference frequency from f _P and i.

14. The method according to claim 9, characterized in that the following applies: f _m0 d equals f _P , that f _m od and f _{P have} a phase relationship Φ to one another, and that between two images Φ recorded in immediate succession by the image recording device (6) Grows 90 °.

15. The method according to claim 14, characterized in that Φ assumes the following values: 0 °, 90 °, 180 ° and 360 °.

16. The method according to any one of claims 1 to 15, characterized in that the fluorescent agent has a relaxation time which is small compared to the reciprocal of f _P.

17. The method according to any one of claims 1 to 16, characterized in that an electronic component is used as a sample (3), the electrical conductor structures and / or contact points, preferably solder joints, are checked with regard to their electrical conductivities with regard to their functional behavior.

18. The method according to claim 17, characterized in that the electronic component with the component typical electrical operating and control voltages is applied, by which the component is heated and the fluorescence radiation influenced by the heating is recorded and evaluated.

19.Device for examining heat sources within an electrically conductive material-containing sample (3) which can be subjected to an electrical voltage for the purpose of heat development and which is exposed to at least one sample surface with a fluorescent agent, the fluorescence property of which is temperature-dependent, with an illuminant ( 4), which exposes the sample surface thus prepared to an energy flow energizing the fluorescent agent, an image recording device (6) for recording images of the fluorescence radiation emerging from the prepared sample surface and an evaluation unit (7) for evaluation, which is generated by the image recording device (6) Images taken according to the lock principle, whereby a frequency generator (1) is provided, which generates the electrical voltage with a predeterminable modulation frequency f _P and in a number of predeterminable phase positions Φ and a multiplexer for Selection of the modulated electrical voltage with a specific phase position, which is applied to the sample (3).

20. The apparatus according to claim 19, characterized in that the lighting means (4) is connected to the frequency generator (1) for modulating the energy flow with the phase position Φ = 0 °.

21. The apparatus of claim 19 or 20, characterized in that the energy flow of the lighting means (4) is amplitude modulated with a fixed frequency of preferably 100 Hz, and that the frequency generator (1) for modulating the energy flow with its phase position Φ = 0 ° to this amplitude modulation is synchronized.

22. Device according to one of claims 19 to 21, characterized in that the lighting means (4) is a UV lighting unit.

23. Device according to one of claims 19 to 22, characterized in that the image recording device (6) is a CCD camera, which is preceded by a microscope.