METHOD FOR THERMAL ACTIVATION OF A PASSIVATION LAYER

Abstract

METHOD FOR THERMALLY ACTIVATING A PASSIVATION LAYER One aspect of the invention relates to a method for thermally activating a passivation layer (10) disposed on a photovoltaic cell (20). The photovoltaic cell (20) comprises a first face (20a), a second face (20b) opposite the first face, and lateral surfaces (20c) connecting the first and second faces. The passivation layer (10) covers at least one of the lateral surfaces (20c) of the photovoltaic cell (20). The method includes a step of exposing the first face (20a) to electromagnetic radiation (30) emitted by a radiation source (40). The electromagnetic radiation (30) is applied to the first face (20a) along a line. The line sweeps at least part of the first face (20a) and is oriented with respect to the first face (20a) so as to obtain a superheating zone encompassing at least part of the passivation layer (10).Figure to be published with the abbreviation: Figure 1.

Description

Title of the invention: METHOD FOR THERMALLY ACTIVATING A PASSIVATION LAYER technical field

[0001] The present invention relates to the manufacture of photovoltaic modules. More particularly, the invention relates to a method for thermally activating a passivation layer disposed on one or more lateral surfaces of a photovoltaic cell. This thermal activation method finds a particularly advantageous application with photovoltaic subcells obtained by cutting whole photovoltaic cells. STATE OF THE ART

[0002] A photovoltaic module comprises a multitude of identical photovoltaic cells connected in series and / or parallel. A common module format comprises 60 square (or "pseudo-square") cells, 156 mm on each side, arranged in six strings of ten cells connected in series. The six strings of photovoltaic cells are also connected in series. The open-circuit voltage across the module is then equal to 60 times the threshold voltage of a photovoltaic cell. The electric current of the module corresponds approximately to the current supplied by each photovoltaic cell (in practice, photovoltaic cells do not have exactly the same performance, and the electric current is limited by the least efficient cell in the module).

[0003] With the latest photovoltaic cell technologies, particularly TopCON (for "tunneling oxide passivated contact"), PERC (for "passivated emitter and rear cell"), and SHJ ("Silicon heterojunction") technologies, the current of a monofacial cell with a surface area of ​​156 mm x 156 mm reaches high values, on the order of 9 A for a solar irradiance of 1000 W / m². These current values ​​are increased by approximately 20% when using a bifacial cell, due to the diffuse solar radiation captured on the rear face of the cell. This high electrical current flows through the interconnecting elements between the modules' cells and causes significant resistive losses.

[0004] In order to reduce these resistive losses, one solution is to assemble modules with photovoltaic cells of smaller surface area, and therefore lower current. These smaller surface area cells are generally fractions of a photovoltaic cell commonly called "sub-cells" and obtained by cutting full-size photovoltaic cells (e.g. 156 mm x 156 mm).

[0005] Cutting a photovoltaic cell creates new edges which, unlike The front and back surfaces of the cell are not covered by passivation layers. Furthermore, the cutting process (laser cutting, for example) can create defects and introduce impurities near the cutting plane. These defects and impurities reduce the lifetime of free charge carriers by acting as electron-hole pair recombination centers, resulting in decreased cell efficiency. This phenomenon is particularly pronounced in heterojunction photovoltaic cells, which inherently have very few surface defects, and where the creation of even a few localized defects is enough to significantly reduce the cell's electrical performance.

[0006] To optimize the performance of photovoltaic subcells, it is therefore necessary to also passivate the newly created edges of these subcells.

[0007] One technique consists of forming an alumina (Al₂O₃) passivation layer by atomic layer deposition (ALD) on one or more side walls of photovoltaic subcells. International patent application WO2020 / 127896 further describes the formation of a passivation layer by space-based atomic layer deposition (SALD) simultaneously on several photovoltaic subcells.

[0008] Once the passivation layer is deposited, a so-called "activation" step of the passivation is generally carried out by thermal annealing to significantly improve the passivation effect. Indeed, this annealing forms fixed charges which create a field effect and cause hydrogen atoms to diffuse, filling dangling bonds at the interfaces.

[0009] International application WO2020 / 220079 describes cutting a photovoltaic cell into several sub-cells called "shingles" (because they are intended to be interconnected by slightly overlapping them, like tiles or shingles on a roof) and depositing a passivation layer (for example, of Al2O3) on the cut edges of the sub-cells. The deposition is followed by activation annealing at 400 °C.

[0010] Finally, the article [“Low-thermal budget flash light annealing for A12O3 surface passivation”, DK Simon et al., Physica status solidi (RRL) - Rapid Research Letters, Vol. 9, No. 11, pp. 631-635, 2015] describes the deposition of an A12O3 passivation layer on each of the two main faces of a silicon substrate and flash light annealing to thermally activate the two passivation layers.

[0011] Annealing is accomplished by exposing the two main faces of the substrate to flashes of light (of duration equal to 1.8 ms) produced by a xenon lamp. During exposure, the substrate is placed in a dihydrogen (H2) atmosphere heated to a temperature of 200 °C.

[0012] Generally speaking, the thermal activation of a passivation layer deposited on a photovoltaic cell is all the more effective as the annealing temperature is high. However, the photovoltaic cell can be degraded if its temperature exceeds a threshold value, typically around 250 °C for a silicon heterojunction (SHJ) cell and around 500 °C for a silicon homojunction cell (for example, TopCON or PERC type). Summary of the invention

[0013] There is a need to thermally activate a passivation layer disposed on at least one lateral surface of a photovoltaic cell, while reducing the risk of degrading the photovoltaic cell.

[0014] According to a first aspect of the invention, this need is met by providing a method for thermally activating a passivation layer disposed on a photovoltaic cell, the photovoltaic cell comprising a first face, a second face opposite the first face and lateral surfaces connecting the first and second faces, the passivation layer covering at least one of the lateral surfaces of the photovoltaic cell, the method comprising a step of exposing the first face to electromagnetic radiation emitted by a radiation source, the thermal activation method being notable in that the electromagnetic radiation is applied to the first face along a line, in that the line sweeps at least part of the first face and in that the line is oriented with respect to the first face so as to obtain a superheated zone encompassing at least part of the passivation layer.

[0015] By "thermal activation" we mean in particular the formation of fixed charges in the passivation layer (these fixed charges creating a field effect which repels one of the two types of free charge carriers towards the substrate of the photovoltaic cell) and / or the saturation of the pendant bonds on the surface of the crystalline silicon substrate of the photovoltaic cell (which are responsible for a large number of electron-hole pair recombinations).

[0016] Preferably, the first side comprises: • a first edge called the leading edge; • a second edge called the trailing edge, opposite the first edge; and • the third and fourth edges connecting the first and second edges.

[0017] In a first embodiment of the activation method, the passivation layer covers the lateral surface bordered by the trailing edge and the line includes a straight segment oriented perpendicularly to a median of the first face, the median passing through the leading edge and the trailing edge.

[0018] According to one embodiment, the passivation layer comprises a first portion covering the lateral surface bordered by the trailing edge and a second portion covering the lateral surface bordered by the leading edge. The line sweeps the first face in one direction during a first phase and in a second opposite direction during a second phase.

[0019] In a second embodiment, the passivation layer comprises a first portion covering the lateral surface bounded by the trailing edge and a second portion covering the lateral surface bounded by one of the third and fourth edges. The line comprises a straight segment inclined with respect to a median of the first face at an angle strictly less than 90° in absolute value, the median passing through the leading edge and the trailing edge.

[0020] In a third embodiment, the passivation layer comprises a first portion covering the lateral surface bordered by the third edge and a second portion covering the lateral surface bordered by the fourth edge. The line is a broken line comprising a first straight segment inclined with respect to a median of the first face at a first positive angle and a second straight segment inclined with respect to the median at a second negative angle, the median passing through the leading edge and the trailing edge, the first angle and the second angle being less than 90° in absolute value, preferably less than 45° in absolute value.

[0021] According to a development of this third embodiment, the first angle and the second angle are between 10° and 20° in absolute value.

[0022] According to another development, the first angle and the second angle are equal in absolute value.

[0023] The activation method according to the invention may also have one or more of the following characteristics, considered individually or in all technically possible combinations: • the photovoltaic cell and the radiation source are in relative translational motion; • the line sweeps across the entire surface of the first face; • the surface energy density received at any point on the first face is between 1 kJ / m2 and 1000 kJ / m2, advantageously between 10 kJ / m2 and 100 kJ / m2; • the electromagnetic radiation has an irradiance greater than or equal to 1 kW / m2, preferably between 100 kW / m2 and 5000 kW / m2; • the photovoltaic cell and the radiation source are in relative translational motion at a speed between 1 cm / s and 100 cm / s, preferably greater than or equal to 2 cm / s; and • the line has a width between 0.1 mm and 100 mm, preferably between 1 mm and 10 mm.

[0024] A second aspect of the invention relates to a method for manufacturing passivated photovoltaic subcells, comprising the following steps: • cutting photovoltaic cells to form a plurality of photovoltaic subcells, each photovoltaic subcell comprising a first face, a second face opposite the first face and lateral surfaces connecting the first and second faces, at least one of the lateral surfaces of each photovoltaic subcell, called additional, resulting from the cutting of a photovoltaic cell; • deposit a passivation layer on at least one additional lateral surface of the photovoltaic subcells; • thermally activate the passivation layer of the photovoltaic subcells by following a thermal activation process according to the first aspect of the invention.

[0025] Preferably, the photovoltaic cells are cut into half-cells or subunits with a surface area smaller than that of a half-cell. BRIEF DESCRIPTION OF THE FIGURES

[0026] Other features and advantages of the invention will become clear from the description given below, by way of example and not limitation, with reference to the following figures.

[0027] [Fig.1] schematically represents a method of thermal activation of a passivation layer according to a first aspect of the invention, this method comprising scanning a photovoltaic cell with an electromagnetic radiation line.

[0028] [Fig.2] represents a first mode of implementation of the scanning activation method according to the invention.

[0029] [Fig.3] shows the maximum temperature reached in the thickness of a photovoltaic cell for each point of a face exposed to electromagnetic radiation, during an activation process according to [Fig.3].

[0030] [Fig.4] represents a second implementation method of the scanning activation process according to the invention.

[0031] [Fig.5] shows the maximum temperature reached in the thickness of a photovoltaic cell for each point of the exposed face, during an activation process according to [Fig.4].

[0032] [Fig.6] represents a third mode of implementation of the scanning activation method according to the invention.

[0033] [Fig.7] shows the maximum temperature reached in the thickness of the photovoltaic cell for each point of the exposed face, during an activation process according to [Fig.6].

[0034] [Fig.8A] represents the amplitude of the overheating in a portion adjacent to a trailing edge of the photovoltaic cell, as a function of the scanning speed of the electromagnetic radiation.

[0035] [Fig.8B] represents the width of the superheated portion as a function of the scanning speed of the electromagnetic radiation.

[0036] [Fig.9] represents the temperatures at two points of the photovoltaic cell and the temperature difference between these two points, as a function of the surface energy density supplied to the photovoltaic cell, when the face of the photovoltaic cell is completely swept by electromagnetic radiation.

[0037] [Fig. 10] represents the width of the superheated portion as a function of the scanning speed of the electromagnetic radiation, for different series of calculations.

[0038] [Fig. 11 A], [Fig. 1 IB] and [Fig. 1 IC] represent steps in a manufacturing process of photovoltaic sub-cells according to a second aspect of the invention.

[0039] For clarity, identical or similar elements are identified by identical reference signs throughout the figures. DETAILED DESCRIPTION

[0040] Fig. 1 schematically represents a thermal activation process for a passivation layer 10 arranged on a photovoltaic cell 20.

[0041] The photovoltaic cell 20 comprises a first face 20a, a second face 20b opposite the first face 20 and lateral surfaces 20c connecting the first and second faces 20a-20b.

[0042] The first face 20a and the second face 20b can be described as The "main" surfaces are defined by their significantly larger surface area compared to the lateral surfaces 20c. Preferably, they are of the same dimensions and extend along parallel planes. They may also have a surface texture, as illustrated in [Fig. 1]. One of the first and second faces 20a-20b constitutes the front face of the photovoltaic cell 20, i.e., the face intended to be exposed to incident solar radiation, while the other constitutes the rear face of the photovoltaic cell 20.

[0043] The first and second faces 20a-20b of the photovoltaic cell 20 preferably have a rectangular or generally rectangular shape, for example square or pseudo-square. By "generally rectangular," it is understood that one or more corners of the first and second faces 20a-20b may be truncated or rounded. For example, in a photovoltaic cell 20 with a pseudo-square shape, the four corners of the first and second faces 20a-20b of the photovoltaic cell 20 are truncated or rounded.

[0044] Thus, the photovoltaic cell 20 is generally rectangular in shape. The thickness of the photovoltaic cell 20 (measured perpendicular to the first and second faces 20a-20b) is typically between 50 µm and 200 µm pm.

[0045] The passivation layer 10 covers one or more lateral surfaces 20c of the photovoltaic cell 20. Its purpose is to increase the lifetime of the free charge carriers in the photovoltaic cell 20 by reducing the number of electron-hole pair recombinations on the surface of the photovoltaic cell 20, and thus to increase the conversion efficiency of the photovoltaic cell 20. The passivation layer 10 can cover all the lateral surfaces 20c. It can also cover part of the first face 20a and / or part of the second face 20b (typically strips adjacent to the covered lateral surfaces).

[0046] The passivation layer 10 is, for example, made of alumina (Al₂O₃), silicon dioxide (SiO₂), silicon nitride (Si₃N₄), or hydrogenated amorphous silicon nitride (α-SixNy:H). It can be formed by atomic layer deposition (ALD) or by chemical vapor deposition (CVD), with or without plasma assistance (PEALD or PECVD). It can also be obtained by space-based atomic layer deposition (SALD). The space-based atomic layer deposition technique is a variant of the conventional (so-called "time-based") ALD technique. In this variant, the different precursor gases used to depose the material are separated in space rather than in time.

[0047] The passivation layer 10 preferably comprises hydrogen atoms. The hydrogen can be supplied during the deposition of the layer or after deposition by means of an ion implantation step (low temperature process).

[0048] With reference to [Fig. 1], the activation process includes a step of exposing the first face 20a of the photovoltaic cell 20 to electromagnetic radiation 30 emitted by a radiation source 40. The electromagnetic radiation 30 is preferably directed perpendicularly to the first face 20a.

[0049] The electromagnetic radiation 30 is, at least in part, absorbed by the photovoltaic cell 20 and converted into heat, which thermally activates the passivation layer 10.

[0050] During the exposure stage, also called the irradiation stage, the photovoltaic cell 20 and the radiation source 40 are advantageously set in relative translational motion (in other words, one moves relative to the other). For example, the photovoltaic cell 20 can be placed on a support 50 that moves in translation, while the source 40 can be fixed. Conversely, the support 50 of the photovoltaic cell 20 can be fixed and the source 40 movable. The photovoltaic cell 20 and the source 40 can also both be movable, but not in the same direction or at the same speed.

[0051] The support 50 can in particular be a substrate carrier or a conveyor capable of transporting several photovoltaic cells 20 simultaneously.

[0052] Such a translational movement makes it possible to irradiate a large part of the first face 20a of the photovoltaic cell 20 using a radiation source 40 having an irradiation area smaller than the area of ​​the first face 20a (as in the case of a laser system, for example). It also makes it possible to irradiate several photovoltaic cells 20 successively, these photovoltaic cells being, for example, arranged on the same support 50. In this case, it can be considered that the photovoltaic cells 20 are moving past the radiation source 40.

[0053] The electromagnetic radiation 30 has an irradiance E, also called irradiance energy or surface power density, which is advantageously greater than or equal to 1 kW / m², preferably between 100 kW / m² and 5000 kW / m². The irradiance E represents the power of the electromagnetic radiation received by a unit area, this unit area being oriented perpendicular to the direction of the electromagnetic radiation 30. Such an irradiance makes it possible to rapidly heat the photovoltaic cell 20 to a temperature that allows the thermal activation of the passivation layer 10.

[0054] In order to thermally activate the passivation layer 10 disposed on the photovoltaic cell 20, it is not necessary to expose the second face 20b or the lateral surfaces 20c of the photovoltaic cell 20. On the other hand, the first face 20a can be irradiated entirely. Preferably, all exposed regions of the first face 20a receive electromagnetic radiation 30 for the same treatment time t.

[0055] The radiation source 40 can be a laser system comprising, for example, one or more laser arrays, an array of light-emitting diodes or any other device capable of emitting electromagnetic radiation with an irradiance greater than 1 kW / m2.

[0056] The electromagnetic radiation 30 can be monochromatic, i.e., have only one wavelength, or polychromatic, i.e., have several (monochromatic) components of different wavelengths. More precisely, the electromagnetic radiation 30 has at least one wavelength between 300 nm and 1200 nm (in this range, the radiation is absorbed at least partially in the photovoltaic cell 20).

[0057] Figures 2, 4 and 6 represent different modes of implementation of the activation process of the passivation layer 10.

[0058] The first face 20a (defined as the face exposed to electromagnetic radiation 30) comprises: • a first edge 21 called the leading edge and defined as the edge to pass first in front of the radiation source 40; • a second edge 22 called the trailing edge, opposite the first edge 21 and defined as the edge that must pass last in front of the radiation source 40; and • the third and fourth edges 23-24 called transverse (lower and upper edges in the figures) which connect the first and second edges 21-22.

[0059] As previously stated, the first face 20a of the photovoltaic cell 20 preferably has a rectangular or generally rectangular shape (truncated or rounded corners). The leading edge 21 and the trailing edge 22 are thus parallel to each other, as are the transverse edges 23-24.

[0060] Common to all embodiments, the radiation source 40 is configured so that the electromagnetic radiation 30 is applied to the first face 20a along a line 60. In other words, the surface of the photovoltaic cell 20 irradiated by the radiation source 40 (called the irradiation surface) at a given instant of the activation process is line-shaped. The line 60 has a width L which can be between 0.1 mm and 100 mm, preferably between 1 mm and 10 mm.

[0061] Line 60 may comprise one or more straight segments 61, 61' 62a-62b. The straight segments may be connected, thus forming a broken line, or conversely, be disjoint. Line 60 is thus the intersection of a beam of electromagnetic waves generated by the radiation source 40 with the surface of the photovoltaic cell 20, the rays of the beam being contained in one or more planes (intersecting or not).

[0062] The direction D of the relative displacement between the photovoltaic cell 20 and the radiation source 40 is such that the line 60 sweeps at least part of the first face 20a of the photovoltaic cell 20 during the exposure step. The sweep by the line 60 can begin at the leading edge 21 and end at the trailing edge 22. It can also begin at a position between the leading edge 21 and the trailing edge 22 (that is, after the leading edge 21 has passed in front of the radiation source 40) and end at the trailing edge 22.

[0063] The direction of movement D of the radiation source 40 relative to the photovoltaic cell 20, also called the scanning direction D, is for example perpendicular to the leading edge 21 and the trailing edge 22, as shown in Figures 2, 4 and 6.

[0064] The velocity V of the relative displacement between the photovoltaic cell 20 and the radiation source 40 is preferably chosen so as to process a photovoltaic cell 20 by scanning the electromagnetic radiation in a short time t, typically between 0.1 s and 30 s. The displacement velocity V is advantageously between 1 cm / s and 100 cm / s, and preferably greater than or equal to 2 cm / s.

[0065] This scanning activation method is simple and quick to implement. In By continuously passing the 20 photovoltaic cells in front of the 40 radiation source, the activation process can be accomplished at a high rate, compatible with the industrial production requirements of photovoltaic cells.

[0066] By way of example, considering a displacement speed V between 1 cm / s and 100 cm / s and photovoltaic cells 20 whose main faces 20a-20b measure 156 mm by 156 mm, the processing time t of a photovoltaic cell 20 is between 0.156 s and 15.6 s. The number of photovoltaic cells 20 processed per hour can be between 200 and 25000.

[0067] As shown later by means of thermal simulations, scanning with an electromagnetic radiation line heats the photovoltaic cell 20 non-uniformly, unlike a traditional annealing furnace or simultaneous exposure of the entire first face 20a of the photovoltaic cell (so-called "full-plate" exposure). More specifically, the temperature of the photovoltaic cell 20 increases very sharply in a localized area near one or more edges of the first face 20a.

[0068] The activation method according to the invention exploits this local overheating phenomenon to activate the passivation layer 10 disposed on one or more lateral surfaces 20c of the photovoltaic cell 20. The line 60 is oriented with respect to the first face 20a so as to obtain an overheating zone which encompasses at least part of the passivation layer 10, and preferably the entire passivation layer 10. The passivation effect of the passivation layer 10 can thus be greatly improved while reducing the risk of degrading the photovoltaic cell 20.

[0069] The activation process can be characterized by the following parameters: • E: the irradiance of electromagnetic radiation 30 (in kW / m2); • V: the relative speed of movement between the photovoltaic cell 20 and the radiation source 40, also called electromagnetic radiation scanning speed (in cm / s); • L: the width of line 60 (in mm); and • e: the thickness of the photovoltaic cell 20 (in pm).

[0070] The surface energy density Q (in kJ / m2), or surface energy, received at any point on the face of the photovoltaic cell by electromagnetic radiation is defined by the following relation:

[0071] [Math.l] Q=(E*L)I(IO*V)

[0072] The parameters E, V, and L are preferably chosen to obtain a surface energy density Q between 1 kJ / m² and 1000 kJ / m², advantageously between 10 kJ / m² and 100 kJ / m². The ranges of values ​​given above are for the irradiance E of the electromagnetic radiation 30, the scanning speed V, and the line width L. 60 are wide enough to achieve this.

[0073] Thermal simulations have shown that the temperatures at different points of the photovoltaic cell during the exposure stage increase with the surface energy density Q. Furthermore, for a given combination of the parameters E, V, and L, other thermal simulations have shown that these temperatures depend on the thickness e of the photovoltaic cell. The smaller the thickness e of the photovoltaic cell, the higher the temperatures, because the volumetric energy density absorbed increases. All the simulation results presented below were obtained with a cell thickness e of 160 pm.

[0074] In a first embodiment illustrated by [Fig. 2], the line 60 comprises a straight segment 61 oriented perpendicularly to a median M of the first face 20a of the photovoltaic cell 20. The median M is the straight segment that connects the midpoints of the leading edge 21 and the trailing edge 22. It is preferably perpendicular to the leading edge 21 and the trailing edge 22. The radiation source 40 is advantageously dimensioned so that the line 60 irradiates the entire width of the photovoltaic cell 20.

[0075] Figure 3 is a map of the maximum temperature reached in a photovoltaic cell 20 (whose main faces are pseudo-square: 156 mm x 156 mm) during an activation process according to Figure 2. It was obtained by means of a thermal simulation of the scanning activation process. In this first simulation, as in the subsequent ones, the irradiance E of the electromagnetic radiation is equal to 216 kW / m², the scanning speed V is equal to 2.48 cm / s, and the width of line 60 is equal to 6 mm. The dashed curve represents the isotherm T = 200 °C.

[0076] This mapping shows that the maximum temperature of the photovoltaic cell 20 during exposure is lowest near the leading edge 21 (first exposed to electromagnetic radiation 30), that it gradually increases towards the trailing edge 22 to reach (and remain at) a so-called plateau temperature of approximately 195 °C, and then increases further to exceed 200 °C in a portion 71 adjacent to the trailing edge 22 (i.e., the edge last exposed to electromagnetic radiation 30). Outside this so-called "superheated" portion 71, the maximum temperature of the photovoltaic cell 20 is therefore below 200 °C.

[0077] The local overheating near the trailing edge 22 is explained by the formation of a thermal boundary layer at the front of the beam. This thermal boundary layer is linked to heat diffusion and propagates in the photovoltaic cell from the leading edge 21 to the trailing edge 22. As it approaches the trailing edge 22, the heat can no longer diffuse towards the front of the beam, which causes overheating.

[0078] Fig. 3 also shows that the temperature of the photovoltaic cell is equal to the plateau temperature in a major part of the photovoltaic cell, this part being referred to here as "central" because it is located between the leading edge 21 and the trailing edge 22 and contains the point located at the center of the exposed face 20a, in (x,y)=(0,0).

[0079] The activation process according to the first embodiment is accomplished when the passivation layer 10 covers the lateral surface 20c bordered by the trailing edge 22 of the first face 20a, in other words when the passivation layer 10 runs along the trailing edge 22 (in practice, the cell is oriented with respect to the direction of movement D so that the lateral surface 20c covered by the passivation layer 10 corresponds to the trailing edge 22).

[0080] According to a development of the first embodiment (illustrated by a dashed arrow in [Fig. 2]), the line 60 sweeps the first face 20a from the leading edge 21 towards the trailing edge 22 during a first phase of the exposure step and from the trailing edge 22 towards the leading edge 21 during a second phase of the exposure step. Sweeping the line 60 in both opposite directions of the movement direction D results in a superheated zone comprising two (disjoint) portions: one adjacent to the trailing edge 22 (the superheated portion 71 in [Fig. 3]) and the other adjacent to the leading edge 21.

[0081] This embodiment variant allows the thermal activation of a passivation layer 10 comprising a first portion covering the lateral surface 20c bordered by the trailing edge 22 and a second portion covering the lateral surface 20c bordered by the leading edge 21 (in other words when the passivation layer 10 runs along the leading edge 21 and the trailing edge 22).

[0082] In a second embodiment represented by [Fig. 4], the line 60 comprises a straight segment 61' inclined with respect to the median M at an angle α (non-zero) strictly less than 90° in absolute value, and preferably between 30° and 60° in absolute value. The angle α denotes the smallest measurable angle between the line 60 and the median M. It may be positive or negative.

[0083] The inclination of the line 60 with respect to the median M has the effect of displacing at least part of the overheating towards one of the two transverse edges 23-24 of the first face 20a.

[0084] Fig. 5 is another map of the maximum temperature reached in the photovoltaic cell, obtained by thermal simulation of the activation process according to Fig. 4, taking as an example a positive angle α equal to 45°.

[0085] In this example, overheating (T>200 °C) occurs in the portion 71 adjacent to the second edge 22 (trailing edge) and in a portion 72 adjacent to the third edge 23 (lower transverse edge). The line 60 forms with each of these two edges 22, 23 an angle [3i, [32 equal to 45° (cf. [Fig.4]).

[0086] The activation process according to the second embodiment is accomplished when the passivation layer 10 comprises a first portion covering the lateral surface 20c bordered by the trailing edge 22 and a second portion covering the lateral surface 20c bordered by one of the transverse edges 23-24 (in other words when the passivation layer 10 runs along the trailing edge 22 and one of the transverse edges 23-24).

[0087] Figure 6 represents a third embodiment of the activation method in which line 60 is a broken line. Line 60 comprises a first straight segment 62a inclined with respect to the median M by a first (strictly) positive angle α, and a second straight segment 62b inclined with respect to the median M by a second (strictly) negative angle α2. The first angle α1 and the second angle α2 are non-zero angles less than 90° in absolute value, preferably less than 45° in absolute value. Line 60 then has an arrow (or chevron) shape.

[0088] An arrow-shaped line 60 makes it possible to move at least in part the overheating of the trailing edge 22 towards the two transverse edges 23-24 of the first face 20a, instead of only one of these edges in the second embodiment (Figs.4-5).

[0089] Each of the angles ai and a2 is preferably between 10° and 20° in absolute value. The overheating then occurs almost exclusively along the two transverse edges 23-24.

[0090] Fig. 7 shows, by way of example, the maximum temperature in the photovoltaic cell 20 obtained by thermal simulation of the activation process according to Fig. 6, when the angles ai and a2 are both 20°.

[0091] The activation process according to the third embodiment is accomplished when the passivation layer 10 comprises a first portion covering the lateral surface 20c bordered by the third edge 23 and a second portion covering the lateral surface 20c bordered by the fourth edge 24 (in other words when the passivation layer 10 runs along the two transverse edges 23-24).

[0092] The first angle ai and the second angle a2 may not be equal in absolute value. Nevertheless, when they are equal, the heating of the photovoltaic cell 20 is symmetrical, as illustrated by the temperature map in [Fig. 7]. This allows both portions of the passivation layer 10 to be activated in the same way.

[0093] Figure 8A is a graph representing the amplitude AT of the superheating as a function of the scanning speed V of the electromagnetic radiation. The amplitude AT of the superheating corresponds to the temperature difference between the hottest point in the superheated portion 71 and the point located at the center of the exposed face, at (x,y)=(0,0).

[0094] Figure 8B is a graph representing the width θ of the superheated portion 71 as a function of the scanning speed V of the electromagnetic radiation. The width θ of the superheated portion 71 is measured from the trailing edge 22 of the photovoltaic cell 20, perpendicular to this edge and on the median M of the first face 20a (at y=0).

[0095] These two figures show that, for a given surface energy density (here 60 kJ / m2), the amplitude AT of the superheating and the width ô of the superheated portion 71 decrease when the scanning speed increases (from a scanning speed of about 2 cm / s for the amplitude AT).

[0096] The thermal simulation results, particularly those in [Fig. 8B], show that the width θ of the superheated portion 71 is mainly governed by the sweep speed V. The width θ of the superheated portion 71 can be expressed by a power law of the following form:

[0097] [Math.2] ô = aVb

[0098] where V is the scanning speed and a and b are first and second coefficients (and more particularly real numbers).

[0099] Figure 9 shows the results of further thermal simulations of the activation process according to Figure 2, carried out with the same photovoltaic cell but varying the irradiance E of the electromagnetic radiation between 155 kW / m² and 237 kW / m² and the scanning speed V between 0.92 cm / s and 4.82 cm / s. Line 60 scans the entire surface of the first face 20a of the photovoltaic cell. The plateau temperature Tpiateau at the center of the photovoltaic cell (at x = y = 0), the maximum temperature Tmax reached at the trailing edge 22, and the temperature difference AT between the maximum temperature Tmax and the plateau temperature Tpiateau (AT = Tmax - Tpiateau) are represented as a function of the surface energy density Q resulting from the chosen simulation parameters.

[0100] It is observed that the plateau temperature Tpiateau at the center of the photovoltaic cell (at x = y = 0) and the maximum temperature Tmax increase with the surface energy density Q supplied to the photovoltaic cell 20.

[0101] In the scanning activation method according to the invention, it is preferable to choose the highest possible surface energy density Q while maintaining a plateau temperature Tpiateau less than or equal to the threshold temperature at which the photovoltaic cell is likely to degrade. This threshold temperature is hereinafter referred to as the "photovoltaic cell degradation temperature" and can be between 200 °C and 500 °C depending on the type and composition of the photovoltaic cell. To determine the surface energy density Q, the thickness e and the environment of the photovoltaic cell are advantageously taken into account. tovoltaic (ambient temperature, convective exchanges...).

[0102] Once the surface energy density Q is determined, the parameters E, V, and L are determined to ensure optimal implementation of the activation process. A high scanning speed V, for example, makes it possible to reduce the processing time t and the width ô of the superheated portion 71. Generally, the aim is to obtain the smallest possible width ô (to limit potential degradation of the photovoltaic cell) but still sufficient to encompass the passivation layer 10, while guaranteeing a superheat amplitude AT that allows for efficient activation of the passivation layer 10. The desired width ô is preferably between the thickness of the passivation layer 10 and 3 mm. The scanning speed V is advantageously chosen to be greater than or equal to 2 cm / s.

[0103] A thermal simulation campaign was conducted to determine the first coefficient a and the second coefficient b of the relation ô(V) (see Math. 2). The simulation campaign comprises several series of calculations performed using different values ​​for the parameters Q, E, V, L, and e of the activation process. These values ​​are given in Table 1 below for each of the series.

[0104] [Tables 1] Q Series (kJ / m²) V (cm / s) E (kW / m²) L (mm) e (pm) Q1 60 from 0.1 to 100 from 10 to 10000 6 160 Q2 10 from 0.1 to 100 from 1.67 to 1670 6 160 Q3 200 from 0.1 to 100 from 33.3 to 33333 6 160 L1 60 from 0.1 to 100 from 3 to 3000 20 160 L2 60 from 0.1 to 100 from 60 to 60000 1 160 el 52.3 2.48 216 6 from 60 to 160 Ml from 19.3 to 155 from 0.92 to 4.82 from 155 to 237 6 160

[0105] In each of the series 'Q1', 'Q2', 'Q3', 'LU' and 'L2', the scanning speed V is varied between 0.1 cm / s and 100 cm / s. The irradiance E is adjusted as a function of the speed V according to relation Math. 1 above in order to obtain the indicated density value Q. For these five series, different values ​​of coefficients a and b were therefore obtained by linear regression (in a logarithmic space).

[0106] Figure 10 shows the results of these five series of calculations and the associated linear regressions. The corresponding values ​​of the coefficients a and b, as well as the maximum relative error between the values ​​of δ observed in the simulations and those predicted by linear regression, are given in Table 2 below.

[0107] [Tables2] Series ab Error (%) Ql 16.2 -0.393 31 Q2 19.6 -0.343 25 Q3 13.9 -0.418 27 L1 22.8 -0.388 23 L2 14.2 -0.406 26

[0108] The values ​​of the coefficients a and b are given for V in cm / s and ô in mm.

[0109] For the five series, the first coefficient a varies between approximately 13 and 23, the second coefficient b varies between approximately -0.34 and -0.42 and the maximum relative error is less than 33%.

[0110] The adjustment of the coefficients a and b as a function of the parameters Q and L can be carried out empirically or using other numerical simulations.

[0111] In general, the first coefficient a can be between 0 and 1000, preferably between 5 and 50, and the second coefficient b can be between -10 and 0, preferably between -1 and -0.1.

[0112] The series of calculations 'el' shows the influence of the thickness e of the photovoltaic cell 20 in a reference activation process. For a thickness e ranging from 60 pm to 160 pm, the width <5 of the superheated portion varies by 10%. Equation Math. 2 defined for a given value of e can therefore reasonably be used for another thickness within the range of 60 pm to 160 pm. Alternatively, the coefficients a and b can be adjusted according to the thickness e of the photovoltaic cell.

[0113] Finally, the 'Ml' series groups together simulations corresponding to different experimentally tested configurations of the activation process. In these tests, the energy density Q ranges from 19.3 kJ / m² to 155 kJ / m². The relative error between the width values ​​δ simulated in the 'Ml' series and those calculated by linear regression using the coefficients α and β obtained for the 'Ql' series (Q = 60 kJ / m²) is less than 15%. This result confirms that, for the considered values ​​of energy density Q and sweep speed V, adjusting the coefficients α and β as a function of the energy density Q can only provide a small gain in determining the width <5 of the superheated portion 71.

[0114] The activation method described above is applicable regardless of the type, shape, and dimensions of the photovoltaic cell 20. In particular, it is applicable to full-size photovoltaic cells (for example, square or pseudo-square 156 mm x 156 mm), whose edges have not been (or not sufficiently) activated. fisamment) passivated, than to fractions (or pieces) of photovoltaic cells, hereafter called sub-cells and obtained by cutting.

[0115] The photovoltaic cell 20 can be a silicon heterojunction (SHJ) cell, a so-called "tandem" cell comprising a silicon heterojunction cell and a perovskite cell stacked one on top of the other or a silicon homojunction cell, for example of type PERT (for "passivated emitter and rear totally diffused" in English) or TOPCon (for "tunnel oxide passivated contact" in English).

[0116] An SHJ cell comprises amorphous silicon, which is known to degrade beyond a threshold temperature. This threshold temperature, known as the amorphous silicon degradation temperature, can range from 190 °C to 320 °C, depending in particular on the amorphous silicon deposition method (these are referred to as "low-temperature" cells). For example, it is 220 °C.

[0117] In the case of an SHJ cell, the E, V, and L parameters of the activation process are preferably chosen such that the plateau temperature Tpiateau (reached in a major part of the first face 20a and in particular at the center of the first face 20a in Figures 3 and 5 or at the center of the upper and lower halves of the first face 20a in [Fig. 7]) is less than or equal to a degradation temperature of the SHJ cell, and preferably less than or equal to 220 °C. The degradation temperature of the SHJ cell is preferably equal to the degradation temperature of amorphous silicon. However, another material could limit the degradation temperature of the SHJ cell.

[0118] The parameters E, V, and L are further advantageously chosen so that the maximum temperature Tmax (reached at the edge where the passivation layer 10 is located) is between 200 °C and 350 °C. A portion of the SHJ cell can withstand a maximum temperature Tmax exceeding the degradation temperature only for a given exposure time. For example, at 280 °C, the SHJ cell is not degraded if the exposure time is less than 30 s.

[0119] A silicon homojunction cell (PERC, TOPCon, etc.) can withstand higher temperatures than an SHJ cell (it is referred to as a "high-temperature" cell). The degradation temperature of a homojunction cell is generally between 400 °C and 500 °C. The E, V, and L parameters of the activation process are then preferably chosen so that the plateau temperature Tpiateau is less than or equal to this degradation temperature, and preferably less than or equal to 400 °C. The E, V, and L parameters are further advantageously chosen so that the maximum temperature Tmax is between 200 °C and 800 °C. Again, each maximum temperature Tmax value above the degradation temperature is associated with a maximum exposure time value.

[0120] Conversely, a tandem cell can withstand lower temperatures than an SHJ cell. The degradation temperature of a tandem cell is approximately 120 °C. The E, V, and L parameters of the activation process are then preferentially chosen so that the platform temperature Tpiateau is less than or equal to the degradation temperature of the tandem cell, i.e., less than or equal to 120 °C. The E, V, and L parameters are further advantageously chosen so that the maximum temperature Tmax is between 100 °C and 200 °C. Again, each maximum temperature value Tmax above the degradation temperature (here 120 °C) is associated with a maximum exposure time (for example, a few seconds at 200 °C).

[0121] Regardless of the type of photovoltaic cell, the electromagnetic radiation exposure step can include several successive scanning phases, for example by passing the photovoltaic cell 20 several times in front of the same radiation source 40 (notably through a loop), by increasing the number of electromagnetic radiation sources 40, or by making several round trips with the same electromagnetic radiation line 60. Several successive scanning phases allow the exposure time of the passivation layer to the maximum temperature Tmax (also called the thermal activation temperature) to be accumulated, and thus the passivation effect to be maximized. Two successive scanning phases are advantageously separated (in time) by a cooling phase of the photovoltaic cell. The number of scanning phases depends on the desired total exposure time.

[0122] Since the activation temperature for an SHJ cell or a tandem cell is lower than that of a homojunction cell, the exposure time for the SHJ cell or the tandem cell should be longer than that for the homojunction cell. The SHJ cell and the tandem cell are therefore more likely to be exposed multiple times in succession.

[0123] Figures 1 IA to 1 IC illustrate a preferred embodiment of a method for manufacturing photovoltaic subcells according to a second aspect of the invention. A photovoltaic subcell here refers to a fraction or piece of a full-size photovoltaic cell, also called a "whole" photovoltaic cell. Photovoltaic subcells are intended, for example, for the manufacture of photovoltaic modules with low resistive losses compared to those of conventional photovoltaic modules (composed of whole photovoltaic cells).

[0124] This manufacturing process comprises a step S1 consisting of cutting full-size photovoltaic cells 100 into a plurality of subcells 200 (see Figs. IIA-IIB) and a step S2 of passivating the subcells 200 (see Fig. IIC). For clarity, only a photovoltaic cell 100 and a subcell 200 were represented (in cross-sectional view) in figures 11A and 1 IC respectively.

[0125] The photovoltaic cells 100 were previously manufactured from semiconductor substrates, for example crystalline silicon. These substrates were initially cut from a silicon ingot, then subjected to several manufacturing steps (for example surface structuring, doping, annealing, passivation, screen printing...), but no other cutting step.

[0126] The photovoltaic cells 100 each comprise a first face 100a, a second face 100b opposite the first face 100a and lateral surfaces or walls 100c connecting the first face 100a and the second face 100b.

[0127] Preferably, the photovoltaic cells 100 are ready to be interconnected in a string of cells. They are provided on the first face 100a and / or on the second face 100b with one or more metallizations 110 intended to collect the photogenerated charge carriers and to receive interconnecting elements, for example, wires or metallic ribbons. The metallizations 110 are preferably electrically conductive tracks called "busbars". The busbars 110 can electrically connect collecting fingers (not shown in [Fig. 11 A]) distributed over the entire surface of the first face 100a and / or the second face 100b. One of the first and second faces 100a-100b (the rear face) of the photovoltaic cells 100 can also be entirely metallized. In one implementation variant, the 100 photovoltaic cells are devoid of busbars 110 but only have collecting fingers.

[0128] The first face 100a and the second face 100b of each photovoltaic cell 100 advantageously have a passivation layer 120. This passivation layer 120 renders the surface defects of the photovoltaic cell 100 inactive and improves the lifetime of the photogenerated charge carriers. Preferably, the passivation layer 120 also covers the lateral surfaces 100c of the photovoltaic cell 100.

[0129] In this preferred embodiment, the cutting of each photovoltaic cell 100 is carried out in two successive operations Fl and F2, illustrated respectively by figures 11A and 1 IB.

[0130] The Fl operation is a so-called cutting initiation operation which consists of exposing one of the first and second faces of the photovoltaic cell 100 (the first face 100a in the example of [Fig.1 IA]) to a laser 130 in order to form a trench 140. The depth of the trench 140 is strictly less than the thickness of the photovoltaic cell 100. The depth of the trench 140 is preferably between 50 pm and 150 pm, while the thickness of the photovoltaic cell 100 is typically between 150 pm and 200 pm.

[0131] To limit damage to the cell on both sides of trench 140, the laser 130 is advantageously pulsed, the pulses having a duration between 10 6 s and 10 15 s.

[0132] The formation of the trench 140 by means of the laser 130 creates a zone of weakness which facilitates the mechanical cleavage of the cell during operation F2 (see [Fig. 1 IB]). The mechanical cleavage takes place in the plane of the trench 140, starting from it, preferably by applying identical forces on both sides of the trench 140.

[0133] The photovoltaic cells 100 are preferably cut into two sub-cells 200 of equal area (see [Fig. IIB]), then called "half-cells", or into subunits of lesser area than a half-cell, typically into three, four, five or six sub-cells 200 of equal area (the sub-cells 200 are then thirds, quarters, fifths or sixths of a cell). The cutting of the photovoltaic cells 100 can be fully automated.

[0134] Each subcell 200 comprises a first face 200a and a second face 200b, corresponding respectively to a portion of the first face 100a and a portion of the second face 100b of the photovoltaic cell 100 from which the subcell 200 originates. Each subcell 200 may further have one or more lateral surfaces 200c, each corresponding to all or part of a lateral surface 100c of the photovoltaic cell 100. The first face 200a, the second face 200b, and the lateral surfaces 200c of the subcells 200 are thus advantageously covered with the passivation layer 120.

[0135] Like the photovoltaic cells 100, the sub-cells 200 are ready to be interconnected. A portion of the metallizations 110 of the photovoltaic cell 100 is present on the first face and / or on the second face of each sub-cell 200.

[0136] Each subcell 200 also includes one or more additional lateral surfaces 200c' resulting from the cutting of the photovoltaic cell 100. These additional lateral surfaces 200c' constitute areas where the semiconductor material (e.g., silicon) has been exposed. In other words, these additional lateral surfaces 200c' lack a passivation layer, unlike the first face 200a, the second face 200b, and any other lateral surfaces 200c of the subcell 200. For example, when a photovoltaic cell 100 is cut into four parallel cell strips, two cell strips have two parallel, non-passivated edges, and two other cell strips have only one non-passivated edge.

[0137] With reference to [Fig. 1 IC], the manufacturing process then includes a step S2 of passivating at least one cut edge of the subcells 200, and preferably all the cut edges. This step aims to neutralize the defects of the additional lateral surface 200c' corresponding to said cut edge, and preferably all the additional lateral surfaces 200c'. Thus, it is possible to limit the reduction of the photovoltaic efficiency of sub-cells 200 which is linked to the generation of new edge(s) by cutting.

[0138] The S2 passivation step includes a substep of depositing a passivation layer 210 on said at least one additional lateral surface 200c' of the subcells 200 and a substep of activating the passivation layer 210 of the subcells 200.

[0139] The passivation layer 210 is, for example, composed of alumina (Al₂O₃), silicon dioxide (SiO₂), silicon nitride (Si₃N₄), or hydrogenated amorphous silicon nitride (α-SixNy:H). It can be formed by atomic layer deposition (ALD) or by chemical vapor deposition (CVD), with or without plasma assistance (PEALD or PECVD). It can also be obtained by space-based atomic layer deposition (SALD).

[0140] The deposition of the passivation layer 210 can be carried out on several sub-cells 200 simultaneously, thanks in particular to the passivation process described in international application WO2020 / 127896.

[0141] The S2 passivation step may also include a substep of hydrogen ion implantation in the passivation layer 210, after its deposition and before its activation.

[0142] The activation of the passivation layer 210 of the subcells 200 is accomplished by means of the activation process described above, for example by passing the subcells 200 under the radiation source 40.

[0143] When the subcells 200 are half-cells, each subcell 200 comprises only one edge to be passivated. The thermal activation method according to [Fig. 2] is advantageously employed (by orienting each subcell 200 so that the edge to be passivated corresponds to the trailing edge).

[0144] When the subcells 200 are quarter cells, obtained by cutting a photovoltaic cell 100 along a first direction and a second perpendicular direction, each subcell 200 comprises two adjacent edges to be passivated. The thermal activation method according to [Fig. 4] is advantageously employed.

[0145] When the subcells 200 are strips of cells, obtained by cutting a photovoltaic cell 100 along parallel directions, each subcell 200 comprises one edge to be passivated or two parallel edges to be passivated. The thermal activation method according to [Fig. 2] (single sweep or back-and-forth type) or according to [Fig. 6] is advantageously used.

[0146] As illustrated in [Fig.1 IC], the deposition conditions can be chosen so that the passivation layer 210 also covers a band 200d of the face impacted by the laser (the first face 200a in the example of Figures 11A-1 IC). This band is contiguous to the additional lateral surface 200c' covered by the passivation layer 210. Thus, the defects created by the laser in the first face or the second face of the subcells 200, in the vicinity of the cut edges, can be neutralized in the same way as the defects present on the additional lateral surface 200c'.

[0147] The deposition of the passivation layer 210 on the face impacted by the laser and on the lateral surface 200c' is advantageously carried out in one step (for example using ALD or SALD techniques).

[0148] When the subcell 200 comprises several additional lateral surfaces 200c' and several of them are passivated in step S2, the passivation layer 210 advantageously covers several bands 200d of the face impacted by the laser, each band 200d being contiguous to one of the additional lateral surfaces 200c' passivated.

[0149] The activation method and the subcell manufacturing method according to the invention are not limited to the embodiments described above and many variants and modifications will appear to the person skilled in the art.

[0150] It is possible, in particular, to passivate the four lateral surfaces 20c of the photovoltaic cell 20 by carrying out two successive exposure steps. For example, one exposure step can use a line 60 composed of a straight segment 61 (see [Fig. 2]) to passivate the leading edge 21 and trailing edge 22 (by performing at least one round trip), and the other exposure step can use an arrow-shaped line 60 (see [Fig. 2]) to passivate the two transverse edges 23-24 (upper and lower). Two radiation sources 40 can thus be provided to generate the two types of line.

[0151] Furthermore, the scanning of the photovoltaic cell 20 by the electromagnetic radiation line 60 (at the scanning speed V) can be obtained by means other than a relative translational movement between the photovoltaic cell 20 and the radiation source 40. The radiation source 40 can in particular be pivotable and the photovoltaic cell 20 stationary.

[0152] Finally, at step SI of the subcell manufacturing process (see Figs.1 1A-1 IB), the full-size photovoltaic cells 100 can be cut using a technique other than laser pre-cutting followed by mechanical cleaving, for example by laser alone, with or without laser-induced thermal separation and possible cooling, by mechanical cleaving alone or by mechanical grooving.

Claims

Demands

1. A method for thermally activating a passivation layer (10) disposed on a photovoltaic cell (20), the photovoltaic cell (20) comprising a first face (20a), a second face (20b) opposite the first face and lateral surfaces (20c) connecting the first and second faces, the passivation layer (10) covering at least one of the lateral surfaces (20c) of the photovoltaic cell (20), the method comprising a step of exposing the first face (20a) to electromagnetic radiation (30) emitted by a radiation source (40), characterized in that the electromagnetic radiation (30) is applied to the first face (20a) along a line (60), in that the line (60) sweeps across at least a portion of the first face (20a), resulting in an overheating zone (71, 72) adjacent to one or more edges of the first face (20a),and in that the line (60) is oriented with respect to the first face (20a) such that the overheating zone (71, 72) encompasses at least part of the passivation layer (10).

2. Method according to claim 1, wherein the photovoltaic cell (20) and the radiation source (40) are animated by a relative translational motion.

3. A method according to any one of claims 1 and 2, wherein the first face (20a) comprises: - a first edge (21) called the leading edge; - a second edge (22) called the trailing edge and opposite the first edge (21); and - third and fourth edges (23, 24) connecting the first and second edges (21, 22).

4. Method according to claim 3, wherein: - the passivation layer (10) covers the lateral surface (20c) bounded by the trailing edge (22); and - the line (60) comprises a straight segment (61) oriented perpendicularly to a median (M) of the first face (20a), the median passing through the leading edge (21) and the trailing edge (22).

5. A method according to any one of claims 3 and 4, wherein: - the passivation layer (10) comprises a first portion covering the lateral surface (20c) bordered by the trailing edge (22) and a second portion covering the lateral surface (20c) bordered by the leading edge (21); and - the line (60) sweeps the first face (20a) in a first direction during a first phase and in a second opposite direction during a second phase.

6. A method according to claim 3, wherein: - the passivation layer (10) comprises a first portion covering the lateral surface (20c) bounded by the trailing edge (22) and a second portion covering the lateral surface (20c) bounded by one of the third and fourth edges (23-24); and - the line (60) comprises a straight segment (61') inclined with respect to a median (M) of the first face at an angle (a) strictly less than 90° in absolute value, the median (M) passing through the leading edge (21) and the trailing edge (22).

7. A method according to claim 3, wherein: - the passivation layer (10) comprises a first portion covering the lateral surface (20c) bordered by the third edge (23) and a second portion covering the lateral surface (20c) bordered by the fourth edge (24); and - the line (60) is a broken line comprising a first straight segment (62a) inclined with respect to a median (M) of the first face (20a) by a first positive angle (ai) and a second straight segment (62b) inclined with respect to the median (M) by a second negative angle (a2), the median (M) passing through the leading edge (21) and the trailing edge (22), the first angle (ai) and the second angle (a2) being less than 90° in absolute value, preferably less than 45° in absolute value.

8. A method according to claim 7, wherein the first angle (aj and the second angle (a2) are between 10° and 20° in absolute value.

9. A method according to any one of claims 7 and 8, wherein the first angle (ai) and the second angle (a2) are equal in absolute value.

10. A method according to any one of claims 1 to 9, wherein the line (60) sweeps across the entire surface of the first face (20a).

11. A method according to any one of claims 1 to 10, wherein the surface energy density (Q) received at any point of the first face (20a) is between 1 kJ / m2 and 1000 kJ / m2, advantageously between 10 kJ / m2 and 100 kJ / m2.

12. A method according to any one of claims 1 to 11, wherein the electromagnetic radiation (30) has an irradiance (E) greater than or equal to 1 kW / m2, preferably between 100 kW / m2 and 5000 kW / m2.

13. A method according to any one of claims 1 to 12, wherein the photovoltaic cell (20) and the radiation source (40) are in relative translational motion at a displacement velocity (V) between 1 cm / s and 100 cm / s, preferably greater than or equal to 2 cm / s.

14. A method according to any one of claims 1 to 13, wherein the line (60) has a width (L) between 0.1 mm and 100 mm, preferably between 1 mm and 10 mm.

15. A method for manufacturing passivated photovoltaic subcells (200), comprising the following steps: - cutting (S1) photovoltaic cells (100) to form a plurality of photovoltaic subcells (200), each photovoltaic subcell comprising a first face (200a), a second face (200b) opposite the first face (200a) and lateral surfaces (200c, 200c') connecting the first and second faces, at least one (200c') of the lateral surfaces of each photovoltaic subcell (200), called additional, resulting from the cutting of a photovoltaic cell (100); - depositing (S2) a passivation layer (210) on at least one additional lateral surface (200c') of the photovoltaic subcells (200); - thermally activate (S2) the passivation layer (210) of the photovoltaic subcells (200) by following a process thermal activation according to any one of claims 1 to 14.

16. A manufacturing method according to claim 15, wherein the photovoltaic cells (100) are cut into half-cells (200) or subunits with a surface area less than that of a half-cell.

Images