FR2699323A1 - Reed switch with 3D metallic microstructure - Google Patents

Abstract

The switch (1) has a base (2) supporting two rods (19,21) each placed on a foot (15,17) and thus defining two electrodes (4,6). The base and rods are placed inside a sealed envelope.The switch is manufactured on a silicon wafer which is processed in a batch with other wafers and then cut out when the processing is finished. The base is covered by a silicon dioxide layer and presents two conducting areas (12,13).

Description

"REED" CONTACTOR AND MANUFACTURING METHOD
OF THREE-DIMENSIONAL METALLIC MICROSTRUCTURES
SUSPENDED
The present invention relates to a contactor called "rod" or "reed", that is to say a contactor comprising a closed enclosure inside which are mounted two rods or conductive beams respectively connected to two electrical connection means accessible from outside said enclosure, and return means for, in the absence of a magnetic field, return the rods to a rest position in which their distal parts are separated from one another by a space. The rods or beams being also at least partially made of a material having a high magnetic susceptibility so that in the presence of an external magnetic field of sufficient intensity, their distal parts are brought into contact with one another. other thus establishing an electrical contact between said two connection means.

 The invention also relates to a method of manufacturing by galvanic method three-dimensional metallic microstructures suspended above a substrate, and in particular a method of manufacturing the above-mentioned "reed" contactor.

 Contactors of the type described above are already known, "reed" or "pin" contactors are in fact common electrical components available on the market. They are most often intended to be used in relays in association with a coil capable of producing a magnetic field. Such a relay consisting of a "reed" contactor and a coil is called a "reed" relay.

 The known contactors most often consist of two ferromagnetic steel rods of small diameter arranged in the extension of one another and held integral with one another by fixing each in one of the two ends of a generally glass hollow glass bulb.

 The two steel rods pass through the two walls in which they are respectively fixed and their ends emerge from the interior wall and extend opposite one another in overhang inside the glass enclosure. The ends of the steel rods are further laminated in the form of two flexible blades whose ends intersect. In the rest position, that is to say in particular in the absence of an external magnetic field, a space of a few tenths of a millimeter separates the two blades from one another. In the presence of an external magnetic field of sufficient intensity having a component parallel to the orientation of the two rods, they will magnetize. The two rods being arranged substantially in the extension of one another, they will magnetize in the same direction and the free ends of the two rods will become respectively a pale north and a pale south. This will cause the appearance of an attractive magnetic force between these two ends which will consequently be brought and maintained in contact with each other as long as the external magnetic field remains.

 Although giving full satisfaction in many applications, such contactors have the disadvantage of being too large for certain other applications. Among these applications, mention may be made in particular of limit switch detectors for various microtechnical applications. The external dimensions of such a contactor are in fact typically around 15 mm for the length of the cylindrical glass bulb and from 2 to 3 mm for the diameter of the latter. The two magnetic alloy blades, for their part, typically have a diameter before rolling of 0.5 mm. It is not possible to reduce much the dimensions of a construction like that which has just been described. Indeed, it is in particular necessary that as the dimensions of the contactor are reduced, the precision with which the relative position of the two rods is determined increases in proportion. If this were not the case, the distal ends of the two rods could either touch each other permanently or be separated by a space too large to allow the contact to be closed even in the presence of a high magnetic field. In known "reed" contactors, the two rods are, as we have said, kept integral with one another by fixing in the wall of the enclosure in which they are enclosed and it is this fixing which determines their relative position. The glass bulb serving as enclosure being produced by shaping the glass, the manufacturing tolerances are too large and it is almost impossible to obtain a relative positioning of the two rods whose precision would be better than a few tenths of a millimeter.

 The present invention therefore aims to overcome the drawbacks which have just been described by providing a very small "reed" contactor which is at least as reliable as known "reed" contactors.

 The invention also aims to provide a method for manufacturing three-dimensional metallic microstructures suspended above a substrate, and in particular rods or thin beams of ferromagnetic material suspended above a substrate.

 Thus, the subject of the invention is a contactor comprising two conductive beams connected respectively to electrical connection means, said beams each comprising a distal end, said distal ends being close to one another and movable relative to one another to the other between a first so-called open position in which they are separated from each other by a space, and a second so-called closed position in which they are in contact with each other, said contactor further comprising return means for returning said distal ends to said open position, said beams being further produced at least partially in a magnetizable material so that when they are subjected to a magnetic induction field of sufficient intensity, said distal ends are brought into said closed position, said contactor being characterized in that it comprises a pl base year comprising two separate electrically conductive zones, in that said beams are integral with said base plane at the level of said two conductive zones, and in that at least one of said beams bears on said base plane via support means to which it is suspended.

 The two beams or contact blades being, according to the present invention, secured to a base plane and no longer to the wall of a bulb, their relative positioning is much more precise. In addition, the construction in accordance with the invention which has just been described lends itself to manufacture by micro-machining using sacrificial layers and more particularly by a technique using galvanic deposition operations, and in particular the using the method which is the subject of the present invention.

The present invention indeed also relates to a method of manufacturing by galvanic method three-dimensional metallic microstructures suspended from a substrate, characterized in that it comprises the steps of:
a) creating a first layer of photoresist on one face of said substrate;
-b) configuring said first photoresist layer so as to produce in the thickness thereof at least one free growth space revealing said face of the substrate;
-c) growing by galvanic deposition a metal pad inside said free space, until the metal is flush with the surface of the photoresist;
-d) creating a metallization level on the surface of said first photoresist layer;
-e) depositing a new layer of photoresist on said metallization level;
-f) configuring said new photoresist layer so as to produce in the thickness thereof at least one free space for growth revealing said metallization level;
-g) growing by galvanic deposition a metal pad inside said free space produced in the new photoresist layer;
h) eliminating the photoresist layers and the non-functional parts of the metallization levels.

 According to this process, a given photoresist layer first fulfills a "mold" or "mask" function for galvanic growth and then, in the second step, assuming that a new photoresist layer is deposited on the first, it fulfills the role of sacrificial layer. Since the same layer successively fulfills the function of "mold" and that of sacrificial layer, the process lends itself to an iterative repetition for producing structures comprising any number of layers.

However, other characteristics and advantages of the invention will appear on reading the detailed description which follows given by way of example and taken with reference to the appended figures in which:
FIG. 1 is a side view in section of a "reed" contactor according to a first embodiment of the present invention.

 FIG. 2 is a top view in section taken along II-II of FIG. 1.

 FIGS. 3 to 18 are side views in section of the contactor of FIG. 1 at various stages of its manufacture according to a particular embodiment of the method of the present invention.

 FIGS. 19 to 20 are sectional side and top views respectively of a Ureed 'contactor according to a second embodiment of the present invention.

 Referring now to Figures 1 and 2, which represent a contactor "reedu 1 to the invention, we see that it is formed of a base plane 2 on which are supported two beams 19,21. We can see more specifically that the two beams 19, 21 are integral with the base plane 2 by means of two feet referenced respectively 15 and 17. Each of the beams 19, 21 forms with the foot to which it is fixed an electrode structure (referenced 4 and 6 respectively). A cap 8 covers these two electrodes and forms with the base plane an airtight enclosure for them. As will be seen below, the "reed" contactors in accordance with the present embodiment of the invention are preferably produced in batches or "batch" on silicon wafers or "wafers" and are, at the end of the process, separated from each other by cutting. The base plane 2 of the contactor 1 therefore consists of a rectangle of silicon cut in the wafer used to manufacture the batch of contactors. According to another embodiment of the present invention, the silicon wafer could be replaced by a glass plate.

 The base plane 2 comprises a surface layer (the presence of which is indicated by a thickening of the hatching in the drawings, and which is referenced 10) formed of silicon dioxide and therefore electrically insulating. It can also be seen in FIG. 1 that the base plane comprises on its upper face, two separate electrically conductive zones 12 and 13 constituted by metallization areas. As will be explained in more detail below, these metallization zones are, according to the preferred embodiment of the present invention, made up of two distinct metal layers deposited on the substrate successively. We still see in Figure 1 that, as we have already said, the two beams 19,21 are each fixed on a foot 15,17 and that moreover these two feet are fixed on the base plane 2 at the level respectively of the two metallization areas 12 and 13. The two beams 19, 21 extend horizontally in cantilever from the tops of the two feet 15, 17 and constitute therewith two electrode structures 4, 6 produced each in one piece. Still in this present embodiment, the two electrodes are oriented so that the distal parts of the beams extend in the direction of one another, or more precisely that the beams 19, 21 both extend in the vertical plane which contains the feet 15,17 of the electrodes. The electrodes 4,6 are, as we will see later, produced by galvanic growth of a ferromagnetic alloy preferably of iron and nickel. The “reed” contactors according to the invention can be of considerably smaller dimensions than those which have been given above in relation to the contactors of the prior art. In the embodiment described here of a "reed" contactor according to the invention, the first electrode 6 could have a height typically between 20 and 35 lim, while the second electrode 4 would have a height between 40 and 70 Fm . Each of the electrodes could typically have a length of 500 gm (for the flexible part) and a width of 100 m. The overlap of the two electrodes would typically extend over a length of 40 μm and the space separating in height the two distal ends of the electrodes would typically be between 10 and 15 μm in the rest position, that is to say more precisely in the absence of magnetic field. The thickness of the beams 19 and 21 would also be between 10 and 15 Fm so as to allow a certain flexibility of these. The beams that we have just described therefore have the form of elongated and flexible rectangular blades which are arranged substantially in the extension of one another. In the presence of an external magnetic field oriented parallel to these blades, they will magnetize and an attractive magnetic force will appear between the two ends of the blades which are close to each other. The blades being thin compared to their length and therefore relatively flexible, the attractive force will bring the two ends into contact with each other. Under these conditions the two metallization areas 12 and 13 will be electrically connected to each other and the contactor will therefore be closed.

 The mixture of iron and nickel used for the production of the electrodes preferably has a weak magnetic hysteresis so that when the external magnetic field disappears, the magnetization of the two beams 19,21 also disappears and their two distal ends cease to attract each other. Under these conditions, the elasticity of the metal recalls the beams towards their rest position in which the two metallization areas 12, 13 are no longer electrically connected.

 We will now complete the description of this embodiment of the contactor according to the invention by describing a particular mode of implementation of the method of the present invention which makes it possible to advantageously make the "reed" contactor which has just been described. in relation to Figures 1 and 2.

 As we have already said, the present mode of implementation of the method allows the manufacture of a "reed" contactor in batches or "batch" on a silicon wafer which is finally cut to separate the contactors produced from each other.

 Figures 3 to 18 which describe the stages of the manufacturing process each show only one contactor, but it is obvious that these figures are in fact partial views of a wafer on which numerous contactors are arranged one at a time. next to the others, only one of them being visible in the partial view.

 A layer of silicon dioxide 10 and first of all created on the surface of the silicon by oxidation of the wafer in an oven in the presence of oxygen. This first operation provides an insulating substrate on which we will then create during a second step, separate conductive zones 12, 13 isolated from each other. This operation is carried out by creating on the oxidized silicon metallization areas 12, 13 in accordance with what is shown in FIG. 3. A thin layer, called titanium bonding layer 12a, 13a is first deposited on the entire surface of the wafer by thermal evaporation. The use of titanium is particularly advantageous because this metal adheres well to silicon dioxide. A gold metallization 12b, 13b is preferably then deposited on the titanium to improve the efficiency of the galvanic deposition. These metallization layers deposited by thermal evaporation are approximately extremely thin (0.25 Rm). The two metallization layers thus produced are then etched according to a conventional technique to produce a network of conductive pads which are preferably connected on the insulating plate. In fact, to allow subsequent galvanic deposition to occur, each of the metallization areas must be kept under tension using a power supply. The fact of providing a configuration in which the metallization areas are connected therefore has the advantage of making it possible to place them all simultaneously under voltage using the same electrical contactor. At this point in the manufacturing process, the contactor is similar to what is shown in FIG. 3.

 After the metallization configuration operation, a thick photoresist layer 23 (FIG. 4) is deposited on the surface of the wafer. This thick photoresist is preferably deposited by centrifugation. The photoresist is then configured using a second mask (not shown) to clear above the metallization areas 12, 13 of the openings 25, 26, 27 and 28 known as mold holes in the places where growth will take place later. galvanic. The metal blocks which will thus be formed in the molding holes 25 and 28 will constitute the two contact pads 56 making it possible to connect the "reed" contactor to an external electronic circuit, while the two metal blocks which will be formed in the molding holes 26,27 will respectively constitute the base of foot 15 and foot 17.

 Once the thick photoresist has been configured, the future contactor is similar to what is shown in FIG. 4. Metal blocks 31, 32, 33 and 34 are then grown in an alloy of iron and nickel or in gold for example, by galvanic deposition in the orifices 25, 26, 27 and 28 formed in the thick photoresist. The photoresist therefore plays the role of mold at this stage. At the end of the galvanic growth, the substrate is similar to what is shown in Figure 5.

 A new extremely thin double metallization 36a, 36b formed of a titanium bonding layer covered by a layer of gold is then deposited by thermal evaporation over the entire surface of the wafer in accordance with what is shown in FIG. 6.

 A new thick photoresist 38 is then deposited on the second metallization and configured to form a mold intended to receive a second galvanic deposit. The second configured thick photoresist layer is shown in Figure 7. The mold holes 40,41 formed in the thick photoresist 38 do not extend exactly vertically to the metal blocks 32,33 which have been formed in the first layer photoresist 23.

It will be noted more particularly that the molding hole 41 extends far beyond the metal block 33, the first photoresist layer 23 therefore now plays the role of sacrificial layer allowing the production of suspended structures.

 Galvanic deposition of a ferromagnetic material of iron-nickel is then carried out, for example, in the orifices 40, 41 made in the second photoresist layer 38 to constitute, on the one hand, a second stage 43 for the base 15 of the second electrode 4 and on the other hand the beam forming the first electrode 6. The galvanic growth is stopped before the iron-nickel alloy has reached the level of the surface of the photoresist 38. FIG. 8 represents the wafer at this stage of the process. Note that, quite generally, according to the method of the present invention, it is not necessary that there is even partial superposition between the metal blocks formed by galvanic growth in a layer of photoresist and those formed in the next layer. The metal structures formed in the upper layer can, if necessary, be entirely suspended or free.

 Then deposited on iron-nickel, always by galvanic growth, a gold fattening 45 provided to improve the electrical contact between the two electrodes 4,6 when their distal parts 19,21 touch each other during operation of the contactor. We see in Figure 9 the state of the contactor once the deposit of the gold fattening is complete.

 A third layer of thick photoresist 47 is then produced over the entire surface of the wafer.

The thickness of the photoresist deposited during this step is equal to the difference separating the two beams 19, 21 in the direction of the height in the finished contactor 1. This third layer of photoresist 47 is also configured to produce a molding hole 48 provided for receiving in the next step the third level of the foot 15 of the second electrode 4. The third level of photoresist 47 configured is visible in FIG. 10.

 A third galvanic deposition of iron-nickel or gold, for example, is then carried out inside the orifice made in the photoresist in accordance with what is shown in FIG. 11.

 A new thin double layer of titanium and gold 50a, 50b is then produced over the entire extent of the wafer by thermal evaporation in accordance with what is shown in FIG. 12.

 A fourth layer of thick photoresist 52 is then produced over the entire extent of the wafer. This photoresist is then configured to produce an orifice 54 for molding the beam 19 forming the second electrode 4 which will also be produced by galvanic deposition in a subsequent step. The fourth configured photoresist layer 52 is also visible in FIG. 13.

 A layer of gold 53 is first formed by galvanic deposition in the bottom of the molding hole 54 made in the photoresist 52. This layer of gold 53 constitutes a fattening on the beam 19 forming the second electrode 4 which will serve as the 'gold fattening 45 which was previously carried out on the beam 21 forming the first electrode 6 to promote electrical contact therebetween. Figure 14 represents the at the end of the stage of depositing this second fattening of gold.

 The iron-nickel core of the beam 19 forming the second electrode 4 is then produced by a final galvanic deposition. FIG. 15 represents the "reed" contactor 1 according to the invention once all the steps of galvanic deposition have been completed. The spacing between the two electrodes 4,6 being in this embodiment determined only by the thickness of the third layer of thick photoresist, it will be possible to produce contactors with extremely fine tolerances in the positioning of the electrodes.

 The contactor is then subjected to an etching reagent to release either in a single operation or in stages, the two electrodes 4,6 and therefore eliminate both the photoresist layers 23,38,47 and 52 as well as the gold metallizations and of titanium 36 and 50. FIG. 16 shows the two electrodes 4,6 of the contactor finished. Since the electrodes 4,6 are essentially made of an iron-nickel alloy, they are ferromagnetic and therefore highly magnetizable. The production of the beams 19, 21 according to the method by successive layers which has just been described makes it possible to give the latter a determined thickness. This thickness being chosen so as to provide the flexibility necessary to allow the distal ends of the two beams to come into contact in the presence of a relatively weak magnetic field. To avoid any risk of oxidation of the iron-nickel alloy, the electrodes are then placed in a hermetic enclosure filled with an inert gas. To this end, a honeycomb cover 8 made for example from micro-machined glass is glued onto the wafer which, once glued to the wafer, will enclose each pair of electrodes 4,6 in an individual cell (the thickness of adhesive joining the cover to the substrate is referenced 60 in Figure 17). FIG. 17 partially shows the plate on which the honeycomb cover 8 has been glued.

 At this stage of production, a single plate covered with a cover 8 comprising a multiplicity of cells (only a fragment of this assembly being shown in FIG. 17) groups together a whole batch of contactors. The assembly of the plate and the cover therefore defines a multiplicity of cavities of which approximately half encloses the pair of electrodes of a contactor, while the other cavities enclose contact pads 56.

 The exact distribution of contactors and contact pads 56 in the different cells naturally depends on the particular shape of the masks used for the configuration of the photoresist layers.

The assembly of assembled contactors must now be sawn to separate the contactors from one another.

 This operation is preferably carried out in two stages. In a first step, the cover material is cut to a depth sufficient to make it easily breakable. This operation produces the notches 58 visible in FIG. 18. Once this operation has been carried out, the wafer is, in a second step, cut to separate all of the individual contactors from one another. Once the different contactors have been separated, it is easy to break the cover fragments which are located above each of the contact pads 56 since these fragments have already been cut during the first sawing step. Once these cover fragments have been removed, the contact pads 56 are easily accessible for making the connections of the finished contactor with an external electrical circuit.

The cutting step which has just been described typically provides several thousand contactors from a 10 cm diameter wafer.

 Figures 19 and 20 show a "reed" contactor according to a second embodiment of the present invention. In this second embodiment, the beam 121 forming the first electrode 106 is fixed directly to the conductive pad 113 of the substrate 102, unlike the second electrode 104, the electrode 106 therefore does not include a foot. In this embodiment, only three layers of thick photoresist are used for the manufacture of the contactor instead of four in the embodiment described above. Since the first electrode 106 is obviously not flexible in the present embodiment, the beam 119 of the second electrode 104 alone will have to show sufficient deflection in the presence of an external magnetic field to close the contactor.

 Finally, remember that the "reed" contactor according to the invention can obviously be produced according to a micro-machining method different from that of the present invention, and that conversely the method of the present invention is not limited to the production of contactors "reed". The scope of the present invention is defined by the two independent claims appended.

Claims (5)

 1. Contactor comprising two conductive beams (19,21; 119,121) connected respectively to electrical connection means (56), said beams each comprising a distal end, said distal ends being close to each other and movable one with respect to the other between a first so-called open position in which they are separated from each other by a space, and a second so-called closed position in which they are in contact with each other, said contactor further comprising return means for returning said distal ends to said open position, said beams (19,21; 119,121) being furthermore made at least partially in a magnetizable material so that when they are subjected to a field magnetic induction of sufficient intensity, said distal ends are brought into said closed position, said contactor being characterized in that it comprises nd a base plane (2,102) comprising two separate electrically conductive zones (12,13; 112,113), in that said beams (19,21; 119,121) are integral with said base plane at the level of said two conductive zones, and in that at least one (19,21; 119) of said beams is supported on said base plane by means of support means (15,17; 115) to which it is suspended.  2. Contactor according to claim 1, characterized in that it comprises a cap (8; 108) which covers the beams (19,21; 119,121) and which is fixed on said base plane (2; 102) to form with the latter an enclosed enclosure, and in that said connection means (56) are disposed outside of said enclosure.  3. Method for manufacturing, by galvanic method, three-dimensional metallic microstructures (4,6; 104) suspended from a substrate, characterized in that it comprises the steps of:  -a) creating a first layer (23) of photoresist on one face of said substrate;  -b) configuring said first photoresist layer so as to produce in the thickness thereof at least one free growth space (25,27,28,29) revealing said face of the substrate;  -c) growing by galvanic deposition a metal pad (31, 32, 33, 34) from inside said free space, until the metal is flush with the surface of the photoresist;  -d) creating a metallization level (36) on the surface of said first photoresist layer;  -e) depositing a new layer of photoresist on said metallization level;  -f) configuring said new photoresist layer so as to produce in the thickness thereof at least one free growth space (40,41) revealing said metallization level;  -g) growing by galvanic deposition a metal pad (43,21) inside said free space produced in the new photoresist layer;  h) eliminating the photoresist layers and the non-functional parts of the metallization levels.  4. Method according to claim 3, characterized in that the initial step comprises the following operations:  -creating a bonding metallization layer (12a, 13a; 112a, 113a) on specific areas of the substrate;  -creating a secondary metallization layer (12b, 13b; 112b, 113b) in a non-oxidizable metal on said bonding metallization layer.  5. Method according to claim 3 or 4, characterized in that it also comprises, between step g and step h, the successive additional operations of:  -i) create a new level of metallization on the surface of the new photoresist layer;  -j) perform steps e, f and g again in order.