Method for determining the position of at least one defect on semiconductor substrates, manufacturing method and associated device

Abstract

Method for determining the position of at least one defect on semiconductor substrates, manufacturing method and associated device. One aspect of the invention relates to a method for determining the position of at least one defect on semiconductor substrates in a production line, the method being implemented using at least one test substrate having a first face and a second face, the method comprising, for each test substrate: a photoluminescence characterization step of each face of the test substrate at at least one point in the production line, this characterization step comprising, for at least one defect present on the test substrate; a substep of measuring the photoluminescent signal of the defect when the first face of the test substrate is illuminated so as to determine a first amplitude associated with the measured defect;a substep of measuring the photoluminescent signal of the defect when the second face of the test substrate is illuminated so as to determine a second amplitude associated with the measured defect; the lifetime of the charge carriers during characterization and / or the wavelength used during characterization being adjusted so that the measurement contrast is greater than a threshold value, the method further comprising, from the value of the amplitudes determined previously: a step of determining the position, surface of the test substrate, on which each defect considered during the characterization step is located; from the position associated with each defect present on each test substrate, a step of determining the position associated with each defect.

Description

Description Title of the invention: Method for determining the position of at least one defect on semiconductor substrates, method manufacturing and associated device TECHNICAL FIELD OF THE INVENTION

[0001] = The technical field of the invention is that of the detection of defects on the substrates (e.g., silicon wafers) of a production line.

[0002] The present invention relates to a method for detecting the position of a de- effectiveness and in particular a process in which, during detection, the lifespans charge carriers are weak in the substrate. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0003] = During the production of an electronic device, the substrate used for this fa- Brication will be moved to carry out the different stages of the manufacturing process. During its transport, the substrate (for example, a silicon wafer) is in contact with handling elements such as belts, vacuum grippers, trays, etc. These contacts can damage the substrate surface by abrading the surface or by depositing (or even embedding) contaminants on this surface (particles, various deposits, etc.). These observations also apply to substrates after passivation which can also undergo deterioration due to ma- manipulations and transport (scratches, pyramid clipping or texturing, etc.). In the following, the defects resulting from this surface degradation are encompassed under the term “defect”.

[0004] = However, the surface of the substrate (before or after passivation) is degraded at this level defect such that any step carried out on the areas affected by these defects effectiveness will be of lower quality. For example, a nanometric layer (for example filed by PECVD — Plasma Enhanced Chemical Vapor Deposition (in English) having The aim of passivating the surface deposited on these areas will be of lower quality (contaminated, poor adhesion, locally absent deposit, etc.) compared to a passivation layer deposited on a defect-free area.

[0005] It is therefore necessary to eliminate or reduce this defect by optimizing certain manufacturing steps. This optimization first requires knowing the position specifies the defect(s) on the substrate in order to identify the step(s) responsible for their creation.

[0006] To do so, the document P.Tomanek et al., “Detection and Localization of Defects in Monocrystalline Silicon Solar Cell, Advances in Optical Technologies” (2010) proposes a method for characterizing defects by coupling measurements Electrical and optical. Although effective, this method can only be performed after establishing electrical contact with the substrate to be characterized (using a metal grid, for example), which limits its scope. Furthermore, it requires additional manipulation of the substrate, which could introduce further defects. Finally, its implementation in an industrial production line would require significant investment, as this method does not rely on any characterization techniques commonly used in industry. Alternatively, the paper Chen et al., “Solar cell surface defect inspection based on multispectral convolutional neural network,” J. Intell. Manuf. (2020), proposes a characterization technique using multispectral analysis of the substrate surface. This technique has the advantage of requiring minimal handling and therefore does not introduce defects. Furthermore, it is fast, making it compatible with industrial production rates. However, this characterization only allows the identification of defects if they are on the substrate surface and therefore cannot identify those related to the presence of contaminants located beneath the passivation layers, for example. Conversely, photoluminescence is a proven defect characterization technique that allows the identification of all types of defects. Furthermore, it requires minimal handling and is compatible with industrial production rates. As an illustration, [Fig. 1] shows a typical photoluminescence image of a substrate that has undergone degradation by the transport system and has been subsequently passivated by a layer of approximately 10 nm of hydrogenated amorphous silicon (a-Si:H — PECVD coating). In this figure, the dark areas (associated with low photoluminescence, denoted PL in [Fig. 1]) correspond to the degraded areas, while the light areas (associated with high photoluminescence) correspond to areas where the surface condition has been preserved and the passivation is of good quality. Typically, the wavelength used for electron excitation in the substrate being characterized is in the near-infrared (approximately 900 nm). This wavelength has a penetration depth in silicon of around 80 µm (penetration depth is a statistical value representing the depth at which 63% of the photons have been absorbed). Therefore, in this material, although photogeneration occurs primarily near the surface exposed to light, a significant portion takes place deeper within the cell, as 37% of the photons are absorbed beyond the penetration depth. Furthermore, when the material used for manufacturing is of very high quality (as is the case, in particular, for the manufacture of photovoltaic devices), the diffusion length of the photogenerated carriers is generally much greater than the thickness of the substrate used. Consequently, the photogenerated carriers during Light excitations statistically have an equal chance of recombining on a defect located on the illuminated face as on a defect located on the opposite face. Furthermore, the implementation of the photoluminescence technique makes it difficult to identify the front or back surface of the substrate on which the defects are located. This difficulty is illustrated in [Fig. 2], which shows a comparison of an image of a front-facing defect (left) and an image of a back-facing defect (right). Moreover, this difficulty is expected to increase with technological advancements, which anticipate an increase in the lifetime of charge carriers in materials as well as a thinning of the substrates used to manufacture semiconductor devices, such as photovoltaic cells. There is therefore a need for a method compatible with industrial manufacturing and enabling the identification of the position (including the face) of one or more defects on a substrate in a manufacturing line. Summary of the invention The invention offers a solution to the problems mentioned above, by modulating, during characterization, the wavelength used and / or the lifetime (and therefore the diffusion length) of the charge carriers in the reference substrate used so as to be able to differentiate the signal of a defect depending on whether it is located on one or the other face of the reference substrate. One aspect of the invention relates to a method for determining the position of at least one defect on semiconductor substrates in a production line, the method being implemented using at least one test substrate having a first face and a second face, the method comprising, for each test substrate: a step of characterizing each face of the substrate by photoluminescence witnessed at least at one point in the production chain, this characterization step- sterization including, for at least one defect present on the control substrate; a substep of measuring the photoluminescent signal of the defect when the The first face of the control substrate is illuminated so as to determine a first amplitude associated with the measured defect; a substep of measuring the photoluminescent signal of the defect when the the second side of the control substrate is illuminated so as to determine a second amplitude associated with the measured defect. Amplitude is defined as the ratio between the signal measured over a defect-free region and the signal measured over the defect. Furthermore, in the process according to a first aspect of the invention, the lifespan of the charge carriers during characterization and / or the wavelength used during characterization are adjusted so that the ratio between the highest amplitude and the lowest amplitude, called measurement contrast, is greater than a threshold value, itself greater than or equal to 1. The method according to a first aspect of the invention then comprises: based on the previously determined amplitude values, a step of determination of the position, on the control substrate, of each defect considered during the characterization stage; and from the face associated with each defect present on each control substrate, a step of determining the position associated with each defect. In addition to the characteristics mentioned in the preceding paragraph, the process according to a first aspect of the invention may have one or more complementary characteristics from among the following, considered individually or according to all technically possible combinations. In one embodiment, a light intensity greater than 20 sun equivalents is used during the characterization step so as to induce, during this step, a decrease in the lifetime of charge carriers in the control substrate by increasing the recombination rate by Auger effect. In one embodiment, the control substrate is a substrate having a short charge carrier duration. In one embodiment, the diffusion length of the charge carriers of the control substrate during the characterization step is less than the thickness of the control substrate. In one embodiment, during the characterization step, the adjustment of the lifetime of the charge carriers and the wavelength used is carried out in two phases: a first phase during which the adjustment is only made on the load carrier lifespan; and when adjusting the lifespan of load carriers is no longer if possible and the contrast is below the threshold value, a second phase during which the adjustment is made to the wavelength used. A second aspect of the invention relates to a method for manufacturing a semiconductor device using a manufacturing line (CP), the manufacturing method comprising, at least at one point of the manufacturing line (CP), a step of implementing the method (100) for determining the position of at least one defect according to a first aspect of the invention. In addition to the characteristics mentioned in the preceding paragraph, the process according to a second aspect of the invention may have one or more ca- complementary characteristics from the following, considered individually or according to all technically possible combinations. In one embodiment, the semiconductor device is manufactured from a substrate with a long charge carrier lifetime and the step of implementing the process according to a first aspect of the invention is preceded by a step of introducing, into the manufacturing chain, at least one control substrate with a short charge carrier lifetime. In one embodiment, the semiconductor device is a photovoltaic device. A third aspect of the invention relates to a device for determining, by photoluminescence, the position of at least one defect on semiconductor substrates of a manufacturing line, comprising at least one illumination means, at least one acquisition means, and means configured to implement a method according to a first or second aspect of the invention. A fourth aspect of the invention relates to a computer program comprising instructions which lead the device according to a third aspect of the invention to execute the steps of the process according to a first aspect of the invention. A fifth aspect of the invention relates to a computer-readable medium on which the computer program is recorded according to a fourth aspect of the invention. The invention and its various applications will be better understood by reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES The figures are presented for illustrative purposes only and are in no way limiting to the invention. Figure [Fig. 1] shows a photoluminescence image of defects on a wafer. Figure 2 illustrates the absence of difference between the photoluminescent signal of a front panel defect and the same signal for a rear panel defect using prior art techniques. Figure 3 shows a flowchart of a process according to the invention. Figure 4 shows a schematic representation of a production line in which there are test substrates for the implementation of a process according to the invention. Figures [Fig.5A] and [Fig.5B] illustrate the measurement of the attenuation for a given defect for each face of a reference substrate. Figures [Fig.6A] and [Fig.6B] show the evolution of the contrast as a function of the lifetime of the carriers ([Fig.6A], for a wavelength of 700 nm) and as a function of the wavelength ([Fig.6B], for a lifetime of charge carriers of 10us). Figures [Fig.7A] to [Fig.7B] illustrate the influence of charge carrier lifetime, wavelength or luminous flux on the contrast measured for a given defect. Figures [Fig.8A] to [Fig.8B] illustrate the contrast between the photoluminescent signal of a front panel defect ([Fig.8A]) and the same signal for a rear panel defect ([Fig.8B]) with the method according to the invention. Figure [Fig. 9A] shows a schematic representation of a device according to a first embodiment. Figure [Fig. 9B] shows a schematic representation of a device according to a first embodiment. DETAILED DESCRIPTION The figures are shown for illustrative purposes only and are not intended to limit the invention. Unless otherwise specified, the same element appearing in different figures has a unique reference numeral. Method for determining the position of at least one defect A first aspect of the invention, illustrated in [Fig. 3] and [Fig. 4], concerns a method 100 for determining the position of at least one defect (i.e., a defect as described in the introduction) on semiconductor substrates SB in a CP production line. The term "position" refers to the face of the substrate SB on which the defect is located, as well as its location on that face. As already mentioned, the prior art generally allows the location of a defect to be determined without specifying which face this location refers to (in other words, the prior art allows the location of defects to be determined in a plane, but does not allow the determination of which face of the substrate this plane is relative to). The process 100 according to the invention is implemented using at least one SBT control substrate having a first FA face and a second FR face. As will be detailed below, an SBT control substrate can be a substrate chosen from among the SB substrates already present on the production line and intended to produce a final device (for example, a photovoltaic device) or a substrate inserted into the production line for the sole purpose of implementing the process 100 according to the invention.In the latter case, the SBT control substrate is a substrate whose charge carrier lifetime is significantly shorter than the charge carrier lifetime of the SB substrates already present on the CP production line and intended for the manufacture of a final device (for example, at least ten times shorter, at least one hundred times shorter, or even at least one thousand times shorter — for example, for a production line on which substrates whose charge carrier lifetime is . With charge carrier lifetimes of a few milliseconds, the lifetime of the charge carriers in the control substrate could be on the order of ten microseconds (measured at an injection rate of An = 5 x 10⁻¹⁰ cm⁻¹). For example, the lifetime of the charge carriers in the SBT control substrate(s) could be less than 0.1 ms, while the lifetime of other SB substrates already present in the CP production line is greater than or equal to 1 ms, for example, between 1 ms and 10 ms. Such an SBT control substrate can, for example, be obtained by activating thermal donors or by introducing metallic contaminants during the fabrication of the silicon ingot. An illustration of a CP production line containing conventional SB substrates (i.e., not involved in the implementation of the invention) and SBT control substrates is shown in [Fig. 4]. Again, it is understood that, in some embodiments of the invention, the SBT control substrates are identical to the conventional SB substrates except that they undergo a photoluminescence characterization step. Conversely, in other embodiments, when the lifetime of the charge carriers in the SBT control substrates is short, the SBT control substrates are different from the conventional SB substrates present on the production line and serve the sole purpose of implementing process 100 according to the invention. characterization of ch L in by photolumi The process 100 according to the invention comprises, for each SBT control substrate, a step 11 of photoluminescence characterization of each face of the SBT control substrate at at least one point in the CP production line. Preferably, this characterization is carried out at several points in the CP production line in order to more easily determine the manipulations that introduced the various defects determined by means of the process 100 according to the invention. More specifically, the characterization step El includes, for at least one DE defect present on the SBT reference substrate, a substep E11 measuring the photoluminescent signal of the DE defect when the first face FA of the SBT reference substrate is illuminated, so as to determine a first amplitude Ara associated with the measured DE defect. Preferably, this substep El is implemented for a plurality of DE defects present on the SBT reference substrate. The characterization step El also includes, for the DE defect(s) considered in the previous substep E11, a substep E12 of measurement of the photoluminescent signal of the DE defect when the second face FR of the SBT test substrate is illuminated so as to determine a second amplitude Arr associated with the measured DE defect. The measurement of each face can be performed by flipping the SBT control substrate between the measurement of the first face (FA) and the measurement of the second face (FR). In this In this case, the device used to implement the invention generally includes a manipulation means configured to perform this reversal. However, this operation can also be performed by an operator. Alternatively, the measurement of each FA / FR face can be performed using two illumination means and two MA acquisition means. In this case, the device used to implement the invention comprises a first illumination means configured to illuminate the first face and a first acquisition means configured to acquire the signal from this first face. It also comprises a second illumination means configured to illuminate the second face and a second acquisition means configured to acquire the signal from this second face. Alternatively, the measurement of each FA / FR face can be carried out using one illumination means and two MA acquisition means. In this case, the device used for the implementation of the invention comprises an illumination means configured to illuminate the first face, a first acquisition means configured to acquire the signal from this first face and a second acquisition means configured to acquire the signal from the second face, the acquisition then being carried out simultaneously on the first face and the second face. As illustrated in [Fig.5A] and [Fig.5B], the amplitude mentioned in the two preceding substeps 1E11-1E12 is defined as the ratio between the intensity of the signal measured on a defect-free region and the intensity of the signal measured on the defect. Furthermore, during the El characterization step, the lifetime of the charge carriers and / or the wavelength of the light used for photoluminescence are adjusted so that the ratio between the highest and lowest amplitudes for a given defect, hereafter contrast Cav,ar, Est greater than a threshold value, this threshold value being greater than or equal to 1, preferably greater than or equal to 1.05 or even greater than or equal to 1.06 (this value generally allowing visual identification), or even greater than or equal to 1.1. Also, the illumination means of the device used to implement the method according to the invention are generally configured to modulate their flux and / or the wavelength used to illuminate the SBT reference substrate. In one embodiment, the wavelength used is greater than or equal to 400 nm. Indeed, for some cells, below 400 nm, photons are mainly absorbed by the passivating layer(s) and not by the SBT reference substrate to be characterized. It is, however, worth noting that when the SBT reference substrate has a short charge carrier lifetime, then the modulation of the illumination parameters, although desirable because it provides an additional degree of freedom, is not necessary for the implementation of the invention. Figures 6A and 6B show the evolution of contrast as a function of charge carrier lifetime (Fig. 6A, for a wavelength of 700 nm) and wavelength (Fig. 6B, for a charge carrier lifetime of 10 µs). These two figures clearly demonstrate how adjusting the values ​​of these parameters allows us to obtain the desired contrast. The adjustment of the parameters mentioned previously (charge carrier lifetime and wavelength) can, for example, be achieved by repeating substeps 1E11 and 1E12 described earlier, varying the charge carrier lifetime and / or wavelength between each iteration until a contrast greater than the threshold value is obtained. Alternatively, the parameters required to obtain the desired contrast can be determined by simulation (and then verified experimentally, with parameter adjustments if necessary). In one embodiment, to further improve contrast, the diffusion length of the charge carriers of the SBT control substrate during the characterization step 1E1 is less than the thickness of the SBT control substrate. This diffusion length can be adjusted by selecting an SBT control substrate with such a diffusion length or by adjusting the flux used during illumination. In other words, in this embodiment, the lifetime of the charge carriers is reduced so that the diffusion length of the charge carriers of the SBT control substrate during the characterization step 1E1 is less than the thickness of the SBT control substrate. In one embodiment, the process also includes a step of processing the images acquired during the characterization in order, for example, to facilitate comparison between the different characterizations. Determining the position (including the face) of each discard pile At the end of the characterization step 1E1, the amplitude measured by illuminating each FA / FR face of the SBT reference substrate is known. Furthermore, the lifetime of the charge carriers is chosen to establish a contrast between the signals measured for each face. Thus, as shown in [Fig. SA] and [Fig. 5B], the highest amplitude is obtained when the DE defect is located on the illuminated face. Therefore, the method 100 according to the invention comprises, starting from the amplitude values ​​determined during the characterization step 1E1, a step 1E2 for determining the position of each considered DE defect. The considered defect is understood to be the defect(s) for which the amplitude and contrast were determined during the characterization step 1E1. Determining the position associated with © the effect As mentioned in the introduction, in the prior art, photoluminescence imaging allows the location of a defect to be determined in a plane without... It is necessary to determine the face on which the defect is present. However, the present invention overcomes this limitation. It therefore makes it possible to determine the face associated with each defect and the location of the defect on that face. Assuming that the defects, for example recurring defects, observed on the SBT control substrate(s) are defects, that is, defects due to the handling of the SBT control substrate(s), the position of the defects thus detected corresponds to the position of the defects. Therefore, the method 100 also includes, starting from the position associated with each DE defect present on each SBT control substrate, a step 1E3 for determining the position associated with each defect. Adjustment of the lifetime of the control substrates and the wavelength As mentioned previously, during the El characterization step, the lifetime of the charge carriers and / or the wavelength are adjusted so that the ratio between the highest attenuation and the lowest attenuation for a given defect, also noted as Cav / ar contrast, is greater than a threshold value, this threshold value being greater than or equal to 1, preferably greater than or equal to 1.05, or even greater than or equal to 1.06, or even greater than or equal to 1.1. Regarding charge carrier lifetimes, this adjustment can be made by adopting one of the following three solutions: choosing a substrate with an inherently suitable charge carrier lifetime; modifying the charge carrier lifetimes during the characterization step to produce the desired contrast; or a combination of the two preceding solutions. With regard to light, this adjustment generally results in a reduction of the wavelength used. Preferably, the adjustment is performed in two steps: first, the adjustment is made only to the charge carrier lifetime (by changing the reference substrate and / or increasing the luminous flux). Then, when adjusting the charge carrier lifetime is no longer possible (for example, the maximum flux is reached or no reference substrate with a lower lifetime is available) and the desired contrast is not achieved, the adjustment is made to the wavelength used until the desired contrast is reached. Charge Carrier Lifetime Adjustment In one embodiment, the SBT control substrates have a short charge carrier lifetime. A short lifetime is defined as a lifetime of less than 1 ms, preferably between 1 ms (excluding and including) and 1 µs (including) (a shorter lifetime being difficult to achieve uniformly). In this embodiment, the SBT control substrate(s) must be introduced into the CP production line in addition to the SB substrates intended for device fabrication. The CP production line then includes long-life SB substrates that are not used for the implementation of the invention and one or more short-life SBT control substrates used for the implementation of process 100 according to the invention. In this embodiment, the adjustment should then be understood as the insertion of a control SBT substrate having a reduced charge carrier lifetime compared to the control SBT substrate used in the previous iteration of characterization step 1E1 (during which the desired contrast was not achieved). Adjustment of the charge carrier lifetime modified during characterization In one embodiment, the light source used for the photoluminescent characterization step is a high-flux source. In other words, during the characterization step E1, the SBT control substrate is subjected to high-flux radiation. High flux is defined as a light intensity greater than 20 sun equivalents (the lifetime then being approximately equal to 150 juices), preferably greater than or equal to 100 sun equivalents (the lifetime then being approximately equal to 50 juices—for example, as shown in [Fig. 7A], for this intensity, the contrast is similar to the contrast obtained at 900 nm for a lifetime of 50 µs). The use of a high photon flux reduces the lifetime of the charge carriers through the Auger effect. In this embodiment, the adjustment should then be understood as the modification of the incident photon flux relative to the flux used during the previous iteration of the 1E1 characterization step (during which the desired contrast was not achieved). Preferably, the flux is adjusted so that the lifetime of the charge carriers is less than Ims, preferably between 1 ms (not included) and lus (inclusive). Mixed adjustment of carrier lifespan It is of course possible to implement the two previous options, that is, to use a reference SBT substrate with a shorter lifetime than the other SB substrates in the CP production chain and then subject it to a high photon flux during the characterization step El. The objective remains the same: the lifetime of the charge carriers must be short during the characterization step El. In this embodiment, the adjustment should then be understood as the modification of the reference SBT substrate and the flux relative to the flux or the reference SBT substrate used during the previous iteration of the characterization step 1E1 (during which the desired contrast was not achieved). Adjusting the wavelength used In one embodiment, the wavelength used for characterization is The modified wavelength was used to improve contrast. Indeed, reducing this wavelength decreases the depth of light penetration into the SBT control substrate, thus concentrating photogeneration in an area close to the illuminated FA / FR face of the SBT control substrate. Consequently, charge carriers generated by shorter wavelengths will, on average, have a longer path to travel to reach the unilluminated FR / FA face of the SBT control substrate. Therefore, DE defects located on the unilluminated FR / FA face will statistically have less impact on the measured signal at shorter wavelengths, thus improving contrast. In this embodiment, the adjustment should then be understood as the reduction of the wavelength relative to the wavelength used during the previous iteration of the 1E1 characterization step (and during which the desired contrast was not achieved). Effects of adjusting these parameters on contrast Figures [Fig.7A] to [Fig.7B] illustrate the influence of the parameters mentioned above on the contrast measured for a given DE defect. Figure 7A represents the contrast value (edge ​​value) as a function of the wavelength used and the charge carrier lifetime. The improvement in contrast resulting from a decrease in charge carrier lifetime or wavelength is clearly visible in this figure. For example, if the user wants a contrast ratio of at least 1.5 and only has an 800 nm source available, then they will need to choose a charge carrier lifetime of less than 0.2 ms. Figure 7B represents the contrast value (contour value) as a function of the wavelength used and the luminous power (or luminous flux – sun equivalent) used for photoluminescence characterization. Again, the improvement in contrast resulting from an increase in flux (and therefore a decrease in charge carrier lifetime) and a decrease in the wavelength used is clearly visible in this figure. For example, if the user wants a contrast of at least 1.5 and only has a 700 nm source, then they will need to choose a flux greater than or equal to 300 sun equivalents. It should be noted, however, that for the same wavelength, a flux of 50 sun equivalents already provides a contrast greater than 1.1, normally sufficient to identify the face of the SBT reference substrate on which a DE defect is located. Illustration of the ] according to the invention To illustrate the advantages of a process 100 according to the invention, a silicon SBT test substrate with a very short charge carrier lifetime (-10µs — (scatter length -100 µm) underwent a passivation coating on both faces. Then, using a diamond point, the passivation coating was locally removed (to form the inscription "CEA") on one of the faces so as to induce a defect similar to those generated during the various manipulations undergone by SB substrates on a CP manufacturing line. Subsequently, a photoluminescence image of the face containing the defect ([Fig. 8A]) and of the face opposite the defect ([Fig. 8B]) were acquired. The signal associated with the defect is clearly visible in [Fig. 8A], whereas it is strongly attenuated in [Fig. 8B]. In other words, the method according to the invention makes it possible to obtain high contrast, making it very easy to identify the face on which the possible defect is located and therefore its position. This result should be compared to images obtained with a prior art method (see [Fig. 2] already introduced), from which it is not possible to identify the face associated with the defect in question. Method for manufacturing a semiconductor device A second aspect of the invention relates to a method for manufacturing a semiconductor device comprising a step of implementing method 100 for determining the position of at least one defect according to the invention. In one embodiment, the manufacturing method further comprises, prior to the step of implementing method 100 according to the invention, a step of introducing one or more SBT test substrates into the manufacturing chain, said SBT test substrates having a short charge carrier lifetime and being used for implementing method 100 for determining the position of at least one defect according to the invention. Such a process is particularly suited to the manufacture of photovoltaic devices insofar as the latter use substrates with a long charge carrier lifetime and / or a very thin substrate thickness, making it difficult to identify the face on which a defect is located when using the solutions proposed in the prior art. Device for determining the position of at least one defect according to the invention A third aspect of the invention, illustrated in [Fig. 9A] and [Fig. 9B], relates to a device DI for determining the position of at least one defect. This device comprises means configured to implement method 100 according to a first aspect of the invention. In an embodiment illustrated in [Fig. 9A], the DI device according to the invention comprises an illumination means MI (for example, a laser source, an LED source, or a polychromatic source with a monochromator) configured to illuminate the surface of a control substrate and an acquisition means MA (For example, a CCD sensor, preferably combined with a bandpass filter and / or one or more lenses of different magnifications) configured to acquire the photoluminescent signal from the illuminated surface. In one embodiment, the illumination means MI is configured to allow modulation of the luminous flux and / or the wavelength used. Thus, it is possible during the characterization step to modulate the wavelength and / or the flux with all the advantages described above. Preferably, in this embodiment, the device also includes a manipulation means MM (for example, a suction cup mechanism on a rotating support) configured to flip the SBT control substrate. Thus, both faces of the SBT control substrate can be illuminated successively. Alternatively, the flipping of the SBT control substrate can be performed manually by an operator. In one embodiment, the DI device comprises two illumination means MM1 / MM2 as described above, the two illumination means MM1 / MM2 being arranged such that when the first illumination means MM is configured to illuminate the first face of the SBT control substrate, the second illumination means MM2 is configured to illuminate the second face of the SBT control substrate with parameters identical to those of the first illumination means MM1. In this embodiment, the device also comprises two acquisition means MA1 / MA2, each acquisition means being configured to acquire the photoluminescent signal from one of the illuminated surfaces, different from the other acquisition means. Furthermore, the illumination of the SBT control substrate by the first illumination means MMI and by the second illumination means MM2 does not occur simultaneously. In one embodiment, the acquisition means MA is configured to simultaneously acquire an image of each face of the SBT substrate. In this embodiment, illumination can be performed on one of the two faces and acquisition simultaneously on both faces without requiring the SBT substrate to be flipped. The device further comprises a computing means MC (for example, a processor or an ASIC card) associated with a memory ME (for example, RAM, possibly accompanied by a hard drive) comprising the instructions necessary for implementing a method 100 according to a first aspect of the invention, in particular the processing of the acquired signal so as to determine the amplitude and contrast measured during characterization. The memory ME is also configured to store the data necessary for this implementation. In one embodiment, the computing means MC is configured to control the means illumination M / MI1 / MD, the acquisition means MA / MA1 / MA2 and the manipulation means MM of the control substrate SBT. In one embodiment, the DI device also includes a display means (for example a screen) for displaying, on instruction from the computing means, the image or images from process 100 according to a first aspect of the invention, and in particular from characterization step 1E1. In one embodiment, the DI device also includes an input means (for example a keyboard, a mouse or a touch screen) configured to receive from an operator or transmit to an operator the data necessary for the implementation of a process 100 according to a first aspect of the invention. In one embodiment, the device also includes an alignment means to allow easier comparison of the images acquired for each SBT substrate.

Claims

Demands

1. A method (100) for determining the position of at least one defect on semiconductor substrates (SB) in a production line (CP), the method (100) being implemented using at least one reference substrate (SBT) having a first face (FA) and a second face (FR), the method (100) comprising, for each reference substrate (SBT): a step (El) of characterization by photoluminescence of each face of the control substrate (SBT) at at least one point of the production chain (CP), this stage (El) of character- risation including, for at least one defect (DE) present on the control substrate (SBT); a substep (E11) of photoluminescent signal measurement minescent of the defect (DE) when the first face (FA) of the control substrate (SBT) is illuminated by so as to determine a first amplitude (Aga) associated with the measured defect (DE); a substep (E12) of photoluminescent signal measurement minescent of the defect (DE) when the second face (FR) of the control substrate (SBT) is illuminated by so as to determine a second amplitude (Arr) associated with the measured defect (DE); the amplitude being the ratio between the signal measured on a defect-free region and the signal measured on the defect, the lifetime of the charge carriers during characterization and / or the wavelength used during characterization being adjusted so that the ratio between the highest amplitude and the lowest amplitude, called measurement contrast (C av ar), is greater than a threshold value, itself greater than or equal to | ; based on the values ​​of the amplitudes determined beforehand Previously, a step (E2) of determining the position, on each control substrate (SBT), of each defect (DE) considered during the characterization step (E1); from the position associated with each defect (DE) present on each control substrate (SBT), a step (E3) of determination- positioning associated with each defect.

2. A method (100) according to the preceding claim in which a light intensity greater than 20 equivalent suns is used when of the characterization step (El) so as to induce, during this step (El), a decrease in the lifespan of load carriers in the control substrate (CTS) by increasing the recombination rate by Augier effect.

3. A method (100) according to any one of the preceding claims in which the control substrate (SBT) is a substrate with a short carrier duration dump.

4. A method (100) according to any one of the preceding claims in which the SBT (substrate control beam) charge carrier diffusion length during the characterization step (El) is less than the thickness of the control substrate (SBT).

5. A method (100) according to any one of the preceding claims, wherein, during the characterization step (El), the adjustment of the lifespan charge carriers and the wavelength used is determined by two phases: a first phase during which the adjustment is not carried out that on the lifespan of the load carriers; and when adjusting the lifespan of load carriers is no longer possible and the contrast is less than the value threshold, a second phase during which the adjustment is performed on the wavelength used.

6. Method of manufacturing a semiconductor device using a manufacturing line (CP), the manufacturing process comprising, in minus one point in the manufacturing chain (CP), a step in work of the process (100) for determining the position of at least one defect according to one of the preceding claims.

7. A method according to the preceding claim, wherein the semi- the conductor is made from a long-life substrate load carriers and in which the process implementation stage (100) according to any one of claims 1 to 5 is preceded by a step introduction, into the manufacturing chain, of at least one substrate short-life indicator of load carriers.

8. A method according to either of the two preceding claims, wherein the The semiconductor device is a photovoltaic device.

9. Photoluminescence position determination device (DI) of at least one defect on semiconductor substrates of a manufacturing chain (CP) comprising at least one means illumination (MI / MI1 / MI2), at least one means acquisition (MA / MA1 / MA2) and means configured to put into implements a method (100) according to any one of claims 1 to 6.

10. | Computer program comprising instructions that conduct the device (DI) according to the preceding claim to perform the steps of the method (100) according to any one of claims 1 to 6.

11. Computer-readable medium on which the program is recorded computer according to the preceding claim.